diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,35035 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9997037037037035,
+  "eval_steps": 500,
+  "global_step": 5061,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005925925925925926,
+      "grad_norm": 11.294554710388184,
+      "learning_rate": 1.9968385694526775e-05,
+      "loss": 4.9696,
+      "step": 10
+    },
+    {
+      "epoch": 0.011851851851851851,
+      "grad_norm": 4.096276760101318,
+      "learning_rate": 1.992886781268524e-05,
+      "loss": 1.0703,
+      "step": 20
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 3.3549132347106934,
+      "learning_rate": 1.988934993084371e-05,
+      "loss": 0.8739,
+      "step": 30
+    },
+    {
+      "epoch": 0.023703703703703703,
+      "grad_norm": 3.6204795837402344,
+      "learning_rate": 1.9849832049002176e-05,
+      "loss": 0.7627,
+      "step": 40
+    },
+    {
+      "epoch": 0.02962962962962963,
+      "grad_norm": 4.910303115844727,
+      "learning_rate": 1.981031416716064e-05,
+      "loss": 0.6878,
+      "step": 50
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 1.72483229637146,
+      "learning_rate": 1.977079628531911e-05,
+      "loss": 0.6754,
+      "step": 60
+    },
+    {
+      "epoch": 0.04148148148148148,
+      "grad_norm": 2.7193186283111572,
+      "learning_rate": 1.9731278403477577e-05,
+      "loss": 0.7195,
+      "step": 70
+    },
+    {
+      "epoch": 0.042074074074074076,
+      "grad_norm": 2.352292060852051,
+      "learning_rate": 1.972732661529342e-05,
+      "loss": 0.7267,
+      "step": 71
+    },
+    {
+      "epoch": 0.042666666666666665,
+      "grad_norm": 2.714202642440796,
+      "learning_rate": 1.972337482710927e-05,
+      "loss": 0.7515,
+      "step": 72
+    },
+    {
+      "epoch": 0.04325925925925926,
+      "grad_norm": 1.9589886665344238,
+      "learning_rate": 1.9719423038925117e-05,
+      "loss": 0.6682,
+      "step": 73
+    },
+    {
+      "epoch": 0.04385185185185185,
+      "grad_norm": 1.7612636089324951,
+      "learning_rate": 1.971547125074096e-05,
+      "loss": 0.5736,
+      "step": 74
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 3.771888017654419,
+      "learning_rate": 1.971151946255681e-05,
+      "loss": 0.6679,
+      "step": 75
+    },
+    {
+      "epoch": 0.045037037037037035,
+      "grad_norm": 3.2177622318267822,
+      "learning_rate": 1.9707567674372657e-05,
+      "loss": 0.6023,
+      "step": 76
+    },
+    {
+      "epoch": 0.04562962962962963,
+      "grad_norm": 3.4658498764038086,
+      "learning_rate": 1.9703615886188502e-05,
+      "loss": 0.6937,
+      "step": 77
+    },
+    {
+      "epoch": 0.04622222222222222,
+      "grad_norm": 3.8744027614593506,
+      "learning_rate": 1.969966409800435e-05,
+      "loss": 0.6304,
+      "step": 78
+    },
+    {
+      "epoch": 0.046814814814814816,
+      "grad_norm": 2.4407026767730713,
+      "learning_rate": 1.9695712309820198e-05,
+      "loss": 0.7268,
+      "step": 79
+    },
+    {
+      "epoch": 0.047407407407407405,
+      "grad_norm": 3.1892712116241455,
+      "learning_rate": 1.9691760521636042e-05,
+      "loss": 0.7308,
+      "step": 80
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 2.119270086288452,
+      "learning_rate": 1.968780873345189e-05,
+      "loss": 0.7266,
+      "step": 81
+    },
+    {
+      "epoch": 0.04859259259259259,
+      "grad_norm": 2.088822841644287,
+      "learning_rate": 1.9683856945267735e-05,
+      "loss": 0.6167,
+      "step": 82
+    },
+    {
+      "epoch": 0.049185185185185186,
+      "grad_norm": 2.152202606201172,
+      "learning_rate": 1.967990515708358e-05,
+      "loss": 0.6197,
+      "step": 83
+    },
+    {
+      "epoch": 0.049777777777777775,
+      "grad_norm": 2.0184404850006104,
+      "learning_rate": 1.9675953368899427e-05,
+      "loss": 0.6649,
+      "step": 84
+    },
+    {
+      "epoch": 0.05037037037037037,
+      "grad_norm": 1.606665015220642,
+      "learning_rate": 1.9672001580715275e-05,
+      "loss": 0.5462,
+      "step": 85
+    },
+    {
+      "epoch": 0.05096296296296296,
+      "grad_norm": 1.6500372886657715,
+      "learning_rate": 1.966804979253112e-05,
+      "loss": 0.5823,
+      "step": 86
+    },
+    {
+      "epoch": 0.051555555555555556,
+      "grad_norm": 3.070535898208618,
+      "learning_rate": 1.9664098004346968e-05,
+      "loss": 0.7119,
+      "step": 87
+    },
+    {
+      "epoch": 0.052148148148148145,
+      "grad_norm": 2.029632806777954,
+      "learning_rate": 1.9660146216162816e-05,
+      "loss": 0.6446,
+      "step": 88
+    },
+    {
+      "epoch": 0.05274074074074074,
+      "grad_norm": 2.297585964202881,
+      "learning_rate": 1.965619442797866e-05,
+      "loss": 0.6023,
+      "step": 89
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 2.2734296321868896,
+      "learning_rate": 1.9652242639794508e-05,
+      "loss": 0.7264,
+      "step": 90
+    },
+    {
+      "epoch": 0.053925925925925926,
+      "grad_norm": 1.9662164449691772,
+      "learning_rate": 1.9648290851610353e-05,
+      "loss": 0.6378,
+      "step": 91
+    },
+    {
+      "epoch": 0.05451851851851852,
+      "grad_norm": 2.248361349105835,
+      "learning_rate": 1.96443390634262e-05,
+      "loss": 0.5925,
+      "step": 92
+    },
+    {
+      "epoch": 0.05511111111111111,
+      "grad_norm": 1.6683038473129272,
+      "learning_rate": 1.964038727524205e-05,
+      "loss": 0.5358,
+      "step": 93
+    },
+    {
+      "epoch": 0.05570370370370371,
+      "grad_norm": 2.3114562034606934,
+      "learning_rate": 1.9636435487057893e-05,
+      "loss": 0.5771,
+      "step": 94
+    },
+    {
+      "epoch": 0.056296296296296296,
+      "grad_norm": 2.5473990440368652,
+      "learning_rate": 1.963248369887374e-05,
+      "loss": 0.7825,
+      "step": 95
+    },
+    {
+      "epoch": 0.05688888888888889,
+      "grad_norm": 5.387284755706787,
+      "learning_rate": 1.962853191068959e-05,
+      "loss": 0.6484,
+      "step": 96
+    },
+    {
+      "epoch": 0.05748148148148148,
+      "grad_norm": 6.117368221282959,
+      "learning_rate": 1.9624580122505434e-05,
+      "loss": 0.634,
+      "step": 97
+    },
+    {
+      "epoch": 0.05807407407407408,
+      "grad_norm": 3.1351606845855713,
+      "learning_rate": 1.962062833432128e-05,
+      "loss": 0.7081,
+      "step": 98
+    },
+    {
+      "epoch": 0.058666666666666666,
+      "grad_norm": 1.7522640228271484,
+      "learning_rate": 1.9616676546137126e-05,
+      "loss": 0.5208,
+      "step": 99
+    },
+    {
+      "epoch": 0.05925925925925926,
+      "grad_norm": 5.310774326324463,
+      "learning_rate": 1.9612724757952974e-05,
+      "loss": 0.6814,
+      "step": 100
+    },
+    {
+      "epoch": 0.05985185185185185,
+      "grad_norm": 4.405309677124023,
+      "learning_rate": 1.9608772969768822e-05,
+      "loss": 0.6156,
+      "step": 101
+    },
+    {
+      "epoch": 0.060444444444444446,
+      "grad_norm": 6.509561538696289,
+      "learning_rate": 1.9604821181584667e-05,
+      "loss": 0.7136,
+      "step": 102
+    },
+    {
+      "epoch": 0.061037037037037035,
+      "grad_norm": 2.8198821544647217,
+      "learning_rate": 1.9600869393400515e-05,
+      "loss": 0.6814,
+      "step": 103
+    },
+    {
+      "epoch": 0.06162962962962963,
+      "grad_norm": 3.36189341545105,
+      "learning_rate": 1.9596917605216362e-05,
+      "loss": 0.685,
+      "step": 104
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 4.839139461517334,
+      "learning_rate": 1.9592965817032207e-05,
+      "loss": 0.527,
+      "step": 105
+    },
+    {
+      "epoch": 0.06281481481481481,
+      "grad_norm": 2.5125792026519775,
+      "learning_rate": 1.9589014028848055e-05,
+      "loss": 0.7053,
+      "step": 106
+    },
+    {
+      "epoch": 0.06340740740740741,
+      "grad_norm": 2.988523244857788,
+      "learning_rate": 1.9585062240663903e-05,
+      "loss": 0.5708,
+      "step": 107
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 2.7243008613586426,
+      "learning_rate": 1.9581110452479747e-05,
+      "loss": 0.594,
+      "step": 108
+    },
+    {
+      "epoch": 0.06459259259259259,
+      "grad_norm": 1.8424056768417358,
+      "learning_rate": 1.9577158664295595e-05,
+      "loss": 0.5653,
+      "step": 109
+    },
+    {
+      "epoch": 0.06518518518518518,
+      "grad_norm": 4.343511581420898,
+      "learning_rate": 1.957320687611144e-05,
+      "loss": 0.7997,
+      "step": 110
+    },
+    {
+      "epoch": 0.06577777777777778,
+      "grad_norm": 2.5145394802093506,
+      "learning_rate": 1.9569255087927288e-05,
+      "loss": 0.7133,
+      "step": 111
+    },
+    {
+      "epoch": 0.06637037037037037,
+      "grad_norm": 2.634700298309326,
+      "learning_rate": 1.9565303299743136e-05,
+      "loss": 0.7139,
+      "step": 112
+    },
+    {
+      "epoch": 0.06696296296296296,
+      "grad_norm": 2.8199563026428223,
+      "learning_rate": 1.956135151155898e-05,
+      "loss": 0.6335,
+      "step": 113
+    },
+    {
+      "epoch": 0.06755555555555555,
+      "grad_norm": 1.6608991622924805,
+      "learning_rate": 1.9557399723374828e-05,
+      "loss": 0.5518,
+      "step": 114
+    },
+    {
+      "epoch": 0.06814814814814815,
+      "grad_norm": 4.260721206665039,
+      "learning_rate": 1.9553447935190676e-05,
+      "loss": 0.6553,
+      "step": 115
+    },
+    {
+      "epoch": 0.06874074074074074,
+      "grad_norm": 5.977010726928711,
+      "learning_rate": 1.954949614700652e-05,
+      "loss": 0.7438,
+      "step": 116
+    },
+    {
+      "epoch": 0.06933333333333333,
+      "grad_norm": 2.1591851711273193,
+      "learning_rate": 1.954554435882237e-05,
+      "loss": 0.5206,
+      "step": 117
+    },
+    {
+      "epoch": 0.06992592592592592,
+      "grad_norm": 1.8141409158706665,
+      "learning_rate": 1.9541592570638213e-05,
+      "loss": 0.5557,
+      "step": 118
+    },
+    {
+      "epoch": 0.07051851851851852,
+      "grad_norm": 2.5547099113464355,
+      "learning_rate": 1.953764078245406e-05,
+      "loss": 0.6226,
+      "step": 119
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 4.010224342346191,
+      "learning_rate": 1.953368899426991e-05,
+      "loss": 0.6361,
+      "step": 120
+    },
+    {
+      "epoch": 0.0717037037037037,
+      "grad_norm": 2.376124858856201,
+      "learning_rate": 1.9529737206085754e-05,
+      "loss": 0.6211,
+      "step": 121
+    },
+    {
+      "epoch": 0.0722962962962963,
+      "grad_norm": 1.4916654825210571,
+      "learning_rate": 1.95257854179016e-05,
+      "loss": 0.5552,
+      "step": 122
+    },
+    {
+      "epoch": 0.07288888888888889,
+      "grad_norm": 2.6941869258880615,
+      "learning_rate": 1.952183362971745e-05,
+      "loss": 0.5569,
+      "step": 123
+    },
+    {
+      "epoch": 0.07348148148148148,
+      "grad_norm": 1.8737499713897705,
+      "learning_rate": 1.9517881841533294e-05,
+      "loss": 0.5219,
+      "step": 124
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 2.4672229290008545,
+      "learning_rate": 1.9513930053349142e-05,
+      "loss": 0.5189,
+      "step": 125
+    },
+    {
+      "epoch": 0.07466666666666667,
+      "grad_norm": 2.1507484912872314,
+      "learning_rate": 1.950997826516499e-05,
+      "loss": 0.5576,
+      "step": 126
+    },
+    {
+      "epoch": 0.07525925925925926,
+      "grad_norm": 3.406736373901367,
+      "learning_rate": 1.9506026476980835e-05,
+      "loss": 0.6823,
+      "step": 127
+    },
+    {
+      "epoch": 0.07585185185185185,
+      "grad_norm": 2.9042787551879883,
+      "learning_rate": 1.9502074688796682e-05,
+      "loss": 0.5358,
+      "step": 128
+    },
+    {
+      "epoch": 0.07644444444444444,
+      "grad_norm": 2.409555196762085,
+      "learning_rate": 1.9498122900612527e-05,
+      "loss": 0.6153,
+      "step": 129
+    },
+    {
+      "epoch": 0.07703703703703704,
+      "grad_norm": 2.215803623199463,
+      "learning_rate": 1.9494171112428375e-05,
+      "loss": 0.6401,
+      "step": 130
+    },
+    {
+      "epoch": 0.07762962962962963,
+      "grad_norm": 4.093177795410156,
+      "learning_rate": 1.9490219324244223e-05,
+      "loss": 0.731,
+      "step": 131
+    },
+    {
+      "epoch": 0.07822222222222222,
+      "grad_norm": 1.7346997261047363,
+      "learning_rate": 1.9486267536060067e-05,
+      "loss": 0.5266,
+      "step": 132
+    },
+    {
+      "epoch": 0.07881481481481481,
+      "grad_norm": 2.3795764446258545,
+      "learning_rate": 1.9482315747875915e-05,
+      "loss": 0.4939,
+      "step": 133
+    },
+    {
+      "epoch": 0.07940740740740741,
+      "grad_norm": 1.7239115238189697,
+      "learning_rate": 1.9478363959691763e-05,
+      "loss": 0.5554,
+      "step": 134
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.055881500244141,
+      "learning_rate": 1.9474412171507608e-05,
+      "loss": 0.7579,
+      "step": 135
+    },
+    {
+      "epoch": 0.08059259259259259,
+      "grad_norm": 1.8134331703186035,
+      "learning_rate": 1.9470460383323456e-05,
+      "loss": 0.6168,
+      "step": 136
+    },
+    {
+      "epoch": 0.08118518518518518,
+      "grad_norm": 2.66621470451355,
+      "learning_rate": 1.94665085951393e-05,
+      "loss": 0.7431,
+      "step": 137
+    },
+    {
+      "epoch": 0.08177777777777778,
+      "grad_norm": 3.906780958175659,
+      "learning_rate": 1.9462556806955148e-05,
+      "loss": 0.6274,
+      "step": 138
+    },
+    {
+      "epoch": 0.08237037037037037,
+      "grad_norm": 3.205265998840332,
+      "learning_rate": 1.9458605018770996e-05,
+      "loss": 0.5445,
+      "step": 139
+    },
+    {
+      "epoch": 0.08296296296296296,
+      "grad_norm": 2.6175894737243652,
+      "learning_rate": 1.945465323058684e-05,
+      "loss": 0.5814,
+      "step": 140
+    },
+    {
+      "epoch": 0.08355555555555555,
+      "grad_norm": 2.8573319911956787,
+      "learning_rate": 1.945070144240269e-05,
+      "loss": 0.5853,
+      "step": 141
+    },
+    {
+      "epoch": 0.08414814814814815,
+      "grad_norm": 2.0878231525421143,
+      "learning_rate": 1.9446749654218537e-05,
+      "loss": 0.5922,
+      "step": 142
+    },
+    {
+      "epoch": 0.08474074074074074,
+      "grad_norm": 2.580152988433838,
+      "learning_rate": 1.944279786603438e-05,
+      "loss": 0.5123,
+      "step": 143
+    },
+    {
+      "epoch": 0.08533333333333333,
+      "grad_norm": 1.905293345451355,
+      "learning_rate": 1.943884607785023e-05,
+      "loss": 0.6146,
+      "step": 144
+    },
+    {
+      "epoch": 0.08592592592592592,
+      "grad_norm": 2.305149793624878,
+      "learning_rate": 1.9434894289666074e-05,
+      "loss": 0.725,
+      "step": 145
+    },
+    {
+      "epoch": 0.08651851851851852,
+      "grad_norm": 1.8791203498840332,
+      "learning_rate": 1.943094250148192e-05,
+      "loss": 0.5113,
+      "step": 146
+    },
+    {
+      "epoch": 0.08711111111111111,
+      "grad_norm": 3.2804646492004395,
+      "learning_rate": 1.942699071329777e-05,
+      "loss": 0.5662,
+      "step": 147
+    },
+    {
+      "epoch": 0.0877037037037037,
+      "grad_norm": 2.2309234142303467,
+      "learning_rate": 1.9423038925113614e-05,
+      "loss": 0.6014,
+      "step": 148
+    },
+    {
+      "epoch": 0.08829629629629629,
+      "grad_norm": 1.8793226480484009,
+      "learning_rate": 1.9419087136929462e-05,
+      "loss": 0.5555,
+      "step": 149
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.6368279457092285,
+      "learning_rate": 1.941513534874531e-05,
+      "loss": 0.5625,
+      "step": 150
+    },
+    {
+      "epoch": 0.08948148148148148,
+      "grad_norm": 2.7983596324920654,
+      "learning_rate": 1.9411183560561155e-05,
+      "loss": 0.575,
+      "step": 151
+    },
+    {
+      "epoch": 0.09007407407407407,
+      "grad_norm": 2.232161283493042,
+      "learning_rate": 1.9407231772377002e-05,
+      "loss": 0.6644,
+      "step": 152
+    },
+    {
+      "epoch": 0.09066666666666667,
+      "grad_norm": 1.6262363195419312,
+      "learning_rate": 1.940327998419285e-05,
+      "loss": 0.4941,
+      "step": 153
+    },
+    {
+      "epoch": 0.09125925925925926,
+      "grad_norm": 2.2929208278656006,
+      "learning_rate": 1.9399328196008695e-05,
+      "loss": 0.6568,
+      "step": 154
+    },
+    {
+      "epoch": 0.09185185185185185,
+      "grad_norm": 2.0559520721435547,
+      "learning_rate": 1.9395376407824543e-05,
+      "loss": 0.7173,
+      "step": 155
+    },
+    {
+      "epoch": 0.09244444444444444,
+      "grad_norm": 2.1673786640167236,
+      "learning_rate": 1.9391424619640387e-05,
+      "loss": 0.5892,
+      "step": 156
+    },
+    {
+      "epoch": 0.09303703703703704,
+      "grad_norm": 2.2163350582122803,
+      "learning_rate": 1.9387472831456235e-05,
+      "loss": 0.6314,
+      "step": 157
+    },
+    {
+      "epoch": 0.09362962962962963,
+      "grad_norm": 1.9577163457870483,
+      "learning_rate": 1.9383521043272083e-05,
+      "loss": 0.5708,
+      "step": 158
+    },
+    {
+      "epoch": 0.09422222222222222,
+      "grad_norm": 2.8221118450164795,
+      "learning_rate": 1.9379569255087928e-05,
+      "loss": 0.5451,
+      "step": 159
+    },
+    {
+      "epoch": 0.09481481481481481,
+      "grad_norm": 2.987020969390869,
+      "learning_rate": 1.9375617466903776e-05,
+      "loss": 0.6101,
+      "step": 160
+    },
+    {
+      "epoch": 0.09540740740740741,
+      "grad_norm": 2.1372368335723877,
+      "learning_rate": 1.9371665678719624e-05,
+      "loss": 0.6425,
+      "step": 161
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 2.7152347564697266,
+      "learning_rate": 1.9367713890535468e-05,
+      "loss": 0.6866,
+      "step": 162
+    },
+    {
+      "epoch": 0.09659259259259259,
+      "grad_norm": 3.6000654697418213,
+      "learning_rate": 1.9363762102351316e-05,
+      "loss": 0.7783,
+      "step": 163
+    },
+    {
+      "epoch": 0.09718518518518518,
+      "grad_norm": 3.9324700832366943,
+      "learning_rate": 1.935981031416716e-05,
+      "loss": 0.5843,
+      "step": 164
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 1.5524108409881592,
+      "learning_rate": 1.935585852598301e-05,
+      "loss": 0.5835,
+      "step": 165
+    },
+    {
+      "epoch": 0.09837037037037037,
+      "grad_norm": 4.644564151763916,
+      "learning_rate": 1.9351906737798857e-05,
+      "loss": 0.8458,
+      "step": 166
+    },
+    {
+      "epoch": 0.09896296296296296,
+      "grad_norm": 2.487111806869507,
+      "learning_rate": 1.93479549496147e-05,
+      "loss": 0.6279,
+      "step": 167
+    },
+    {
+      "epoch": 0.09955555555555555,
+      "grad_norm": 2.3678605556488037,
+      "learning_rate": 1.934400316143055e-05,
+      "loss": 0.5933,
+      "step": 168
+    },
+    {
+      "epoch": 0.10014814814814815,
+      "grad_norm": 3.075354814529419,
+      "learning_rate": 1.9340051373246397e-05,
+      "loss": 0.6175,
+      "step": 169
+    },
+    {
+      "epoch": 0.10074074074074074,
+      "grad_norm": 1.726096510887146,
+      "learning_rate": 1.933609958506224e-05,
+      "loss": 0.4846,
+      "step": 170
+    },
+    {
+      "epoch": 0.10133333333333333,
+      "grad_norm": 1.982561707496643,
+      "learning_rate": 1.933214779687809e-05,
+      "loss": 0.4975,
+      "step": 171
+    },
+    {
+      "epoch": 0.10192592592592592,
+      "grad_norm": 2.1716926097869873,
+      "learning_rate": 1.9328196008693934e-05,
+      "loss": 0.6438,
+      "step": 172
+    },
+    {
+      "epoch": 0.10251851851851852,
+      "grad_norm": 2.1267306804656982,
+      "learning_rate": 1.9324244220509782e-05,
+      "loss": 0.5818,
+      "step": 173
+    },
+    {
+      "epoch": 0.10311111111111111,
+      "grad_norm": 2.707035541534424,
+      "learning_rate": 1.932029243232563e-05,
+      "loss": 0.6564,
+      "step": 174
+    },
+    {
+      "epoch": 0.1037037037037037,
+      "grad_norm": 3.397979736328125,
+      "learning_rate": 1.9316340644141474e-05,
+      "loss": 0.5196,
+      "step": 175
+    },
+    {
+      "epoch": 0.10429629629629629,
+      "grad_norm": 2.0516886711120605,
+      "learning_rate": 1.9312388855957322e-05,
+      "loss": 0.6255,
+      "step": 176
+    },
+    {
+      "epoch": 0.10488888888888889,
+      "grad_norm": 1.8699017763137817,
+      "learning_rate": 1.930843706777317e-05,
+      "loss": 0.5397,
+      "step": 177
+    },
+    {
+      "epoch": 0.10548148148148148,
+      "grad_norm": 2.8243205547332764,
+      "learning_rate": 1.9304485279589015e-05,
+      "loss": 0.6771,
+      "step": 178
+    },
+    {
+      "epoch": 0.10607407407407407,
+      "grad_norm": 3.450754165649414,
+      "learning_rate": 1.9300533491404863e-05,
+      "loss": 0.5661,
+      "step": 179
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1.701911449432373,
+      "learning_rate": 1.929658170322071e-05,
+      "loss": 0.6017,
+      "step": 180
+    },
+    {
+      "epoch": 0.10725925925925926,
+      "grad_norm": 1.4007916450500488,
+      "learning_rate": 1.9292629915036555e-05,
+      "loss": 0.4606,
+      "step": 181
+    },
+    {
+      "epoch": 0.10785185185185185,
+      "grad_norm": 1.709399700164795,
+      "learning_rate": 1.9288678126852403e-05,
+      "loss": 0.5051,
+      "step": 182
+    },
+    {
+      "epoch": 0.10844444444444444,
+      "grad_norm": 1.3734759092330933,
+      "learning_rate": 1.9284726338668248e-05,
+      "loss": 0.4719,
+      "step": 183
+    },
+    {
+      "epoch": 0.10903703703703704,
+      "grad_norm": 2.2830238342285156,
+      "learning_rate": 1.9280774550484096e-05,
+      "loss": 0.5859,
+      "step": 184
+    },
+    {
+      "epoch": 0.10962962962962963,
+      "grad_norm": 6.9427618980407715,
+      "learning_rate": 1.9276822762299944e-05,
+      "loss": 0.9656,
+      "step": 185
+    },
+    {
+      "epoch": 0.11022222222222222,
+      "grad_norm": 5.653810977935791,
+      "learning_rate": 1.9272870974115788e-05,
+      "loss": 0.56,
+      "step": 186
+    },
+    {
+      "epoch": 0.11081481481481481,
+      "grad_norm": 3.9420740604400635,
+      "learning_rate": 1.9268919185931636e-05,
+      "loss": 0.7358,
+      "step": 187
+    },
+    {
+      "epoch": 0.11140740740740741,
+      "grad_norm": 3.624258279800415,
+      "learning_rate": 1.9264967397747484e-05,
+      "loss": 0.6322,
+      "step": 188
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 1.9406553506851196,
+      "learning_rate": 1.926101560956333e-05,
+      "loss": 0.6185,
+      "step": 189
+    },
+    {
+      "epoch": 0.11259259259259259,
+      "grad_norm": 2.7644498348236084,
+      "learning_rate": 1.9257063821379177e-05,
+      "loss": 0.543,
+      "step": 190
+    },
+    {
+      "epoch": 0.11318518518518518,
+      "grad_norm": 3.4620232582092285,
+      "learning_rate": 1.925311203319502e-05,
+      "loss": 0.5843,
+      "step": 191
+    },
+    {
+      "epoch": 0.11377777777777778,
+      "grad_norm": 3.375847101211548,
+      "learning_rate": 1.924916024501087e-05,
+      "loss": 0.6561,
+      "step": 192
+    },
+    {
+      "epoch": 0.11437037037037037,
+      "grad_norm": 1.6364002227783203,
+      "learning_rate": 1.9245208456826717e-05,
+      "loss": 0.5243,
+      "step": 193
+    },
+    {
+      "epoch": 0.11496296296296296,
+      "grad_norm": 1.7472093105316162,
+      "learning_rate": 1.924125666864256e-05,
+      "loss": 0.6591,
+      "step": 194
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 3.657224655151367,
+      "learning_rate": 1.923730488045841e-05,
+      "loss": 0.4968,
+      "step": 195
+    },
+    {
+      "epoch": 0.11614814814814815,
+      "grad_norm": 3.3418705463409424,
+      "learning_rate": 1.9233353092274257e-05,
+      "loss": 0.4229,
+      "step": 196
+    },
+    {
+      "epoch": 0.11674074074074074,
+      "grad_norm": 2.039853572845459,
+      "learning_rate": 1.9229401304090102e-05,
+      "loss": 0.4748,
+      "step": 197
+    },
+    {
+      "epoch": 0.11733333333333333,
+      "grad_norm": 4.553790092468262,
+      "learning_rate": 1.922544951590595e-05,
+      "loss": 0.6237,
+      "step": 198
+    },
+    {
+      "epoch": 0.11792592592592592,
+      "grad_norm": 3.3934576511383057,
+      "learning_rate": 1.9221497727721798e-05,
+      "loss": 0.5671,
+      "step": 199
+    },
+    {
+      "epoch": 0.11851851851851852,
+      "grad_norm": 2.483027696609497,
+      "learning_rate": 1.9217545939537642e-05,
+      "loss": 0.5466,
+      "step": 200
+    },
+    {
+      "epoch": 0.11911111111111111,
+      "grad_norm": 1.9673043489456177,
+      "learning_rate": 1.921359415135349e-05,
+      "loss": 0.5293,
+      "step": 201
+    },
+    {
+      "epoch": 0.1197037037037037,
+      "grad_norm": 2.3617494106292725,
+      "learning_rate": 1.9209642363169335e-05,
+      "loss": 0.6673,
+      "step": 202
+    },
+    {
+      "epoch": 0.12029629629629629,
+      "grad_norm": 6.1691460609436035,
+      "learning_rate": 1.9205690574985183e-05,
+      "loss": 0.905,
+      "step": 203
+    },
+    {
+      "epoch": 0.12088888888888889,
+      "grad_norm": 1.925883412361145,
+      "learning_rate": 1.920173878680103e-05,
+      "loss": 0.4993,
+      "step": 204
+    },
+    {
+      "epoch": 0.12148148148148148,
+      "grad_norm": 2.4107139110565186,
+      "learning_rate": 1.9197786998616875e-05,
+      "loss": 0.5981,
+      "step": 205
+    },
+    {
+      "epoch": 0.12207407407407407,
+      "grad_norm": 1.9846134185791016,
+      "learning_rate": 1.9193835210432723e-05,
+      "loss": 0.4634,
+      "step": 206
+    },
+    {
+      "epoch": 0.12266666666666666,
+      "grad_norm": 1.979257345199585,
+      "learning_rate": 1.918988342224857e-05,
+      "loss": 0.4279,
+      "step": 207
+    },
+    {
+      "epoch": 0.12325925925925926,
+      "grad_norm": 2.2600274085998535,
+      "learning_rate": 1.9185931634064416e-05,
+      "loss": 0.4911,
+      "step": 208
+    },
+    {
+      "epoch": 0.12385185185185185,
+      "grad_norm": 3.020496368408203,
+      "learning_rate": 1.9181979845880264e-05,
+      "loss": 0.5045,
+      "step": 209
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 1.4864962100982666,
+      "learning_rate": 1.9178028057696108e-05,
+      "loss": 0.471,
+      "step": 210
+    },
+    {
+      "epoch": 0.12503703703703703,
+      "grad_norm": 2.4236865043640137,
+      "learning_rate": 1.9174076269511956e-05,
+      "loss": 0.5277,
+      "step": 211
+    },
+    {
+      "epoch": 0.12562962962962962,
+      "grad_norm": 1.7135413885116577,
+      "learning_rate": 1.9170124481327804e-05,
+      "loss": 0.3967,
+      "step": 212
+    },
+    {
+      "epoch": 0.12622222222222224,
+      "grad_norm": 2.0406389236450195,
+      "learning_rate": 1.916617269314365e-05,
+      "loss": 0.5422,
+      "step": 213
+    },
+    {
+      "epoch": 0.12681481481481482,
+      "grad_norm": 1.9436522722244263,
+      "learning_rate": 1.9162220904959497e-05,
+      "loss": 0.4753,
+      "step": 214
+    },
+    {
+      "epoch": 0.1274074074074074,
+      "grad_norm": 2.451998233795166,
+      "learning_rate": 1.9158269116775345e-05,
+      "loss": 0.5285,
+      "step": 215
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 1.6436266899108887,
+      "learning_rate": 1.915431732859119e-05,
+      "loss": 0.5418,
+      "step": 216
+    },
+    {
+      "epoch": 0.1285925925925926,
+      "grad_norm": 1.8988094329833984,
+      "learning_rate": 1.9150365540407037e-05,
+      "loss": 0.4922,
+      "step": 217
+    },
+    {
+      "epoch": 0.12918518518518518,
+      "grad_norm": 3.117997884750366,
+      "learning_rate": 1.914641375222288e-05,
+      "loss": 0.6478,
+      "step": 218
+    },
+    {
+      "epoch": 0.12977777777777777,
+      "grad_norm": 2.286196231842041,
+      "learning_rate": 1.914246196403873e-05,
+      "loss": 0.4664,
+      "step": 219
+    },
+    {
+      "epoch": 0.13037037037037036,
+      "grad_norm": 2.0614893436431885,
+      "learning_rate": 1.9138510175854577e-05,
+      "loss": 0.5363,
+      "step": 220
+    },
+    {
+      "epoch": 0.13096296296296298,
+      "grad_norm": 3.1992673873901367,
+      "learning_rate": 1.9134558387670422e-05,
+      "loss": 0.4796,
+      "step": 221
+    },
+    {
+      "epoch": 0.13155555555555556,
+      "grad_norm": 3.5712876319885254,
+      "learning_rate": 1.913060659948627e-05,
+      "loss": 0.5867,
+      "step": 222
+    },
+    {
+      "epoch": 0.13214814814814815,
+      "grad_norm": 2.360849618911743,
+      "learning_rate": 1.9126654811302118e-05,
+      "loss": 0.6046,
+      "step": 223
+    },
+    {
+      "epoch": 0.13274074074074074,
+      "grad_norm": 2.7265260219573975,
+      "learning_rate": 1.9122703023117962e-05,
+      "loss": 0.5203,
+      "step": 224
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 2.297422409057617,
+      "learning_rate": 1.911875123493381e-05,
+      "loss": 0.6715,
+      "step": 225
+    },
+    {
+      "epoch": 0.13392592592592592,
+      "grad_norm": 5.845449447631836,
+      "learning_rate": 1.9114799446749658e-05,
+      "loss": 0.6042,
+      "step": 226
+    },
+    {
+      "epoch": 0.1345185185185185,
+      "grad_norm": 3.849013328552246,
+      "learning_rate": 1.9110847658565503e-05,
+      "loss": 0.4762,
+      "step": 227
+    },
+    {
+      "epoch": 0.1351111111111111,
+      "grad_norm": 3.0813441276550293,
+      "learning_rate": 1.910689587038135e-05,
+      "loss": 0.4128,
+      "step": 228
+    },
+    {
+      "epoch": 0.13570370370370372,
+      "grad_norm": 2.1398019790649414,
+      "learning_rate": 1.9102944082197195e-05,
+      "loss": 0.5229,
+      "step": 229
+    },
+    {
+      "epoch": 0.1362962962962963,
+      "grad_norm": 3.056514263153076,
+      "learning_rate": 1.9098992294013043e-05,
+      "loss": 0.6832,
+      "step": 230
+    },
+    {
+      "epoch": 0.1368888888888889,
+      "grad_norm": 3.3704869747161865,
+      "learning_rate": 1.909504050582889e-05,
+      "loss": 0.6267,
+      "step": 231
+    },
+    {
+      "epoch": 0.13748148148148148,
+      "grad_norm": 2.5823540687561035,
+      "learning_rate": 1.9091088717644736e-05,
+      "loss": 0.6731,
+      "step": 232
+    },
+    {
+      "epoch": 0.13807407407407407,
+      "grad_norm": 2.7428412437438965,
+      "learning_rate": 1.9087136929460584e-05,
+      "loss": 0.5544,
+      "step": 233
+    },
+    {
+      "epoch": 0.13866666666666666,
+      "grad_norm": 3.3956573009490967,
+      "learning_rate": 1.908318514127643e-05,
+      "loss": 0.6719,
+      "step": 234
+    },
+    {
+      "epoch": 0.13925925925925925,
+      "grad_norm": 4.516423225402832,
+      "learning_rate": 1.9079233353092276e-05,
+      "loss": 0.5946,
+      "step": 235
+    },
+    {
+      "epoch": 0.13985185185185184,
+      "grad_norm": 2.4637858867645264,
+      "learning_rate": 1.9075281564908124e-05,
+      "loss": 0.4903,
+      "step": 236
+    },
+    {
+      "epoch": 0.14044444444444446,
+      "grad_norm": 2.062366247177124,
+      "learning_rate": 1.907132977672397e-05,
+      "loss": 0.5584,
+      "step": 237
+    },
+    {
+      "epoch": 0.14103703703703704,
+      "grad_norm": 2.2562131881713867,
+      "learning_rate": 1.9067377988539817e-05,
+      "loss": 0.5722,
+      "step": 238
+    },
+    {
+      "epoch": 0.14162962962962963,
+      "grad_norm": 1.4968137741088867,
+      "learning_rate": 1.9063426200355665e-05,
+      "loss": 0.4809,
+      "step": 239
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 1.887015461921692,
+      "learning_rate": 1.905947441217151e-05,
+      "loss": 0.6553,
+      "step": 240
+    },
+    {
+      "epoch": 0.1428148148148148,
+      "grad_norm": 2.252770185470581,
+      "learning_rate": 1.9055522623987354e-05,
+      "loss": 0.4543,
+      "step": 241
+    },
+    {
+      "epoch": 0.1434074074074074,
+      "grad_norm": 2.006333589553833,
+      "learning_rate": 1.90515708358032e-05,
+      "loss": 0.6074,
+      "step": 242
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 2.9486825466156006,
+      "learning_rate": 1.904761904761905e-05,
+      "loss": 0.5421,
+      "step": 243
+    },
+    {
+      "epoch": 0.1445925925925926,
+      "grad_norm": 1.623424768447876,
+      "learning_rate": 1.9043667259434894e-05,
+      "loss": 0.4642,
+      "step": 244
+    },
+    {
+      "epoch": 0.1451851851851852,
+      "grad_norm": 1.8616788387298584,
+      "learning_rate": 1.9039715471250742e-05,
+      "loss": 0.5112,
+      "step": 245
+    },
+    {
+      "epoch": 0.14577777777777778,
+      "grad_norm": 3.8365442752838135,
+      "learning_rate": 1.9035763683066587e-05,
+      "loss": 0.688,
+      "step": 246
+    },
+    {
+      "epoch": 0.14637037037037037,
+      "grad_norm": 2.452410936355591,
+      "learning_rate": 1.9031811894882434e-05,
+      "loss": 0.5267,
+      "step": 247
+    },
+    {
+      "epoch": 0.14696296296296296,
+      "grad_norm": 2.7336888313293457,
+      "learning_rate": 1.9027860106698282e-05,
+      "loss": 0.5428,
+      "step": 248
+    },
+    {
+      "epoch": 0.14755555555555555,
+      "grad_norm": 1.9628325700759888,
+      "learning_rate": 1.9023908318514127e-05,
+      "loss": 0.5495,
+      "step": 249
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 2.648228645324707,
+      "learning_rate": 1.9019956530329975e-05,
+      "loss": 0.5809,
+      "step": 250
+    },
+    {
+      "epoch": 0.14874074074074073,
+      "grad_norm": 5.814465045928955,
+      "learning_rate": 1.9016004742145823e-05,
+      "loss": 0.4962,
+      "step": 251
+    },
+    {
+      "epoch": 0.14933333333333335,
+      "grad_norm": 3.9342072010040283,
+      "learning_rate": 1.9012052953961667e-05,
+      "loss": 0.4765,
+      "step": 252
+    },
+    {
+      "epoch": 0.14992592592592593,
+      "grad_norm": 1.7547112703323364,
+      "learning_rate": 1.9008101165777515e-05,
+      "loss": 0.4955,
+      "step": 253
+    },
+    {
+      "epoch": 0.15051851851851852,
+      "grad_norm": 3.030758857727051,
+      "learning_rate": 1.9004149377593363e-05,
+      "loss": 0.6107,
+      "step": 254
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 2.652066230773926,
+      "learning_rate": 1.9000197589409208e-05,
+      "loss": 0.4748,
+      "step": 255
+    },
+    {
+      "epoch": 0.1517037037037037,
+      "grad_norm": 3.6256392002105713,
+      "learning_rate": 1.8996245801225056e-05,
+      "loss": 0.6961,
+      "step": 256
+    },
+    {
+      "epoch": 0.1522962962962963,
+      "grad_norm": 2.67743182182312,
+      "learning_rate": 1.89922940130409e-05,
+      "loss": 0.6202,
+      "step": 257
+    },
+    {
+      "epoch": 0.15288888888888888,
+      "grad_norm": 3.7790815830230713,
+      "learning_rate": 1.8988342224856748e-05,
+      "loss": 0.6534,
+      "step": 258
+    },
+    {
+      "epoch": 0.15348148148148147,
+      "grad_norm": 4.67446756362915,
+      "learning_rate": 1.8984390436672596e-05,
+      "loss": 0.5707,
+      "step": 259
+    },
+    {
+      "epoch": 0.15407407407407409,
+      "grad_norm": 5.310025215148926,
+      "learning_rate": 1.898043864848844e-05,
+      "loss": 0.5379,
+      "step": 260
+    },
+    {
+      "epoch": 0.15466666666666667,
+      "grad_norm": 2.649723529815674,
+      "learning_rate": 1.897648686030429e-05,
+      "loss": 0.646,
+      "step": 261
+    },
+    {
+      "epoch": 0.15525925925925926,
+      "grad_norm": 1.9800117015838623,
+      "learning_rate": 1.8972535072120137e-05,
+      "loss": 0.5216,
+      "step": 262
+    },
+    {
+      "epoch": 0.15585185185185185,
+      "grad_norm": 7.9760565757751465,
+      "learning_rate": 1.896858328393598e-05,
+      "loss": 1.1217,
+      "step": 263
+    },
+    {
+      "epoch": 0.15644444444444444,
+      "grad_norm": 3.295161724090576,
+      "learning_rate": 1.896463149575183e-05,
+      "loss": 0.5431,
+      "step": 264
+    },
+    {
+      "epoch": 0.15703703703703703,
+      "grad_norm": 2.2408595085144043,
+      "learning_rate": 1.8960679707567674e-05,
+      "loss": 0.4222,
+      "step": 265
+    },
+    {
+      "epoch": 0.15762962962962962,
+      "grad_norm": 1.9301613569259644,
+      "learning_rate": 1.895672791938352e-05,
+      "loss": 0.5403,
+      "step": 266
+    },
+    {
+      "epoch": 0.1582222222222222,
+      "grad_norm": 6.026066303253174,
+      "learning_rate": 1.895277613119937e-05,
+      "loss": 0.6248,
+      "step": 267
+    },
+    {
+      "epoch": 0.15881481481481483,
+      "grad_norm": 2.730334758758545,
+      "learning_rate": 1.8948824343015214e-05,
+      "loss": 0.5794,
+      "step": 268
+    },
+    {
+      "epoch": 0.15940740740740741,
+      "grad_norm": 5.820464134216309,
+      "learning_rate": 1.8944872554831062e-05,
+      "loss": 0.6291,
+      "step": 269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1990089416503906,
+      "learning_rate": 1.894092076664691e-05,
+      "loss": 0.4902,
+      "step": 270
+    },
+    {
+      "epoch": 0.1605925925925926,
+      "grad_norm": 2.2797467708587646,
+      "learning_rate": 1.8936968978462754e-05,
+      "loss": 0.647,
+      "step": 271
+    },
+    {
+      "epoch": 0.16118518518518518,
+      "grad_norm": 2.643035411834717,
+      "learning_rate": 1.8933017190278602e-05,
+      "loss": 0.5365,
+      "step": 272
+    },
+    {
+      "epoch": 0.16177777777777777,
+      "grad_norm": 4.068871974945068,
+      "learning_rate": 1.892906540209445e-05,
+      "loss": 0.5481,
+      "step": 273
+    },
+    {
+      "epoch": 0.16237037037037036,
+      "grad_norm": 3.189082145690918,
+      "learning_rate": 1.8925113613910295e-05,
+      "loss": 0.5343,
+      "step": 274
+    },
+    {
+      "epoch": 0.16296296296296298,
+      "grad_norm": 3.156517505645752,
+      "learning_rate": 1.8921161825726143e-05,
+      "loss": 0.4889,
+      "step": 275
+    },
+    {
+      "epoch": 0.16355555555555557,
+      "grad_norm": 2.42551326751709,
+      "learning_rate": 1.8917210037541987e-05,
+      "loss": 0.6257,
+      "step": 276
+    },
+    {
+      "epoch": 0.16414814814814815,
+      "grad_norm": 2.5290143489837646,
+      "learning_rate": 1.8913258249357835e-05,
+      "loss": 0.4745,
+      "step": 277
+    },
+    {
+      "epoch": 0.16474074074074074,
+      "grad_norm": 3.8375189304351807,
+      "learning_rate": 1.8909306461173683e-05,
+      "loss": 0.4826,
+      "step": 278
+    },
+    {
+      "epoch": 0.16533333333333333,
+      "grad_norm": 4.4902448654174805,
+      "learning_rate": 1.8905354672989528e-05,
+      "loss": 0.5547,
+      "step": 279
+    },
+    {
+      "epoch": 0.16592592592592592,
+      "grad_norm": 3.0676541328430176,
+      "learning_rate": 1.8901402884805376e-05,
+      "loss": 0.524,
+      "step": 280
+    },
+    {
+      "epoch": 0.1665185185185185,
+      "grad_norm": 1.6182653903961182,
+      "learning_rate": 1.8897451096621224e-05,
+      "loss": 0.499,
+      "step": 281
+    },
+    {
+      "epoch": 0.1671111111111111,
+      "grad_norm": 1.7028684616088867,
+      "learning_rate": 1.8893499308437068e-05,
+      "loss": 0.3703,
+      "step": 282
+    },
+    {
+      "epoch": 0.16770370370370372,
+      "grad_norm": 3.7343387603759766,
+      "learning_rate": 1.8889547520252916e-05,
+      "loss": 0.5549,
+      "step": 283
+    },
+    {
+      "epoch": 0.1682962962962963,
+      "grad_norm": 2.643249273300171,
+      "learning_rate": 1.888559573206876e-05,
+      "loss": 0.4946,
+      "step": 284
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 1.745198369026184,
+      "learning_rate": 1.888164394388461e-05,
+      "loss": 0.5112,
+      "step": 285
+    },
+    {
+      "epoch": 0.16948148148148148,
+      "grad_norm": 1.8202131986618042,
+      "learning_rate": 1.8877692155700457e-05,
+      "loss": 0.5174,
+      "step": 286
+    },
+    {
+      "epoch": 0.17007407407407407,
+      "grad_norm": 3.363424777984619,
+      "learning_rate": 1.88737403675163e-05,
+      "loss": 0.6087,
+      "step": 287
+    },
+    {
+      "epoch": 0.17066666666666666,
+      "grad_norm": 3.8631231784820557,
+      "learning_rate": 1.886978857933215e-05,
+      "loss": 0.5144,
+      "step": 288
+    },
+    {
+      "epoch": 0.17125925925925925,
+      "grad_norm": 2.59785532951355,
+      "learning_rate": 1.8865836791147997e-05,
+      "loss": 0.6792,
+      "step": 289
+    },
+    {
+      "epoch": 0.17185185185185184,
+      "grad_norm": 2.8715460300445557,
+      "learning_rate": 1.886188500296384e-05,
+      "loss": 0.4834,
+      "step": 290
+    },
+    {
+      "epoch": 0.17244444444444446,
+      "grad_norm": 3.737656354904175,
+      "learning_rate": 1.885793321477969e-05,
+      "loss": 0.6847,
+      "step": 291
+    },
+    {
+      "epoch": 0.17303703703703704,
+      "grad_norm": 1.793757438659668,
+      "learning_rate": 1.8853981426595534e-05,
+      "loss": 0.5149,
+      "step": 292
+    },
+    {
+      "epoch": 0.17362962962962963,
+      "grad_norm": 2.800292491912842,
+      "learning_rate": 1.8850029638411382e-05,
+      "loss": 0.6077,
+      "step": 293
+    },
+    {
+      "epoch": 0.17422222222222222,
+      "grad_norm": 3.2902846336364746,
+      "learning_rate": 1.884607785022723e-05,
+      "loss": 0.6812,
+      "step": 294
+    },
+    {
+      "epoch": 0.1748148148148148,
+      "grad_norm": 2.829030990600586,
+      "learning_rate": 1.8842126062043074e-05,
+      "loss": 0.7321,
+      "step": 295
+    },
+    {
+      "epoch": 0.1754074074074074,
+      "grad_norm": 4.255771636962891,
+      "learning_rate": 1.8838174273858922e-05,
+      "loss": 0.5393,
+      "step": 296
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 6.132540702819824,
+      "learning_rate": 1.883422248567477e-05,
+      "loss": 0.5467,
+      "step": 297
+    },
+    {
+      "epoch": 0.17659259259259258,
+      "grad_norm": 4.5477705001831055,
+      "learning_rate": 1.8830270697490615e-05,
+      "loss": 0.4571,
+      "step": 298
+    },
+    {
+      "epoch": 0.1771851851851852,
+      "grad_norm": 5.270909309387207,
+      "learning_rate": 1.8826318909306463e-05,
+      "loss": 0.5269,
+      "step": 299
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 1.7171128988265991,
+      "learning_rate": 1.882236712112231e-05,
+      "loss": 0.4693,
+      "step": 300
+    },
+    {
+      "epoch": 0.17837037037037037,
+      "grad_norm": 2.0390264987945557,
+      "learning_rate": 1.8818415332938155e-05,
+      "loss": 0.587,
+      "step": 301
+    },
+    {
+      "epoch": 0.17896296296296296,
+      "grad_norm": 3.10963773727417,
+      "learning_rate": 1.8814463544754003e-05,
+      "loss": 0.6624,
+      "step": 302
+    },
+    {
+      "epoch": 0.17955555555555555,
+      "grad_norm": 2.171713352203369,
+      "learning_rate": 1.8810511756569848e-05,
+      "loss": 0.4656,
+      "step": 303
+    },
+    {
+      "epoch": 0.18014814814814814,
+      "grad_norm": 2.354914665222168,
+      "learning_rate": 1.8806559968385696e-05,
+      "loss": 0.5866,
+      "step": 304
+    },
+    {
+      "epoch": 0.18074074074074073,
+      "grad_norm": 5.202226161956787,
+      "learning_rate": 1.8802608180201544e-05,
+      "loss": 0.6902,
+      "step": 305
+    },
+    {
+      "epoch": 0.18133333333333335,
+      "grad_norm": 2.505213975906372,
+      "learning_rate": 1.8798656392017388e-05,
+      "loss": 0.4944,
+      "step": 306
+    },
+    {
+      "epoch": 0.18192592592592594,
+      "grad_norm": 4.339182376861572,
+      "learning_rate": 1.8794704603833236e-05,
+      "loss": 0.661,
+      "step": 307
+    },
+    {
+      "epoch": 0.18251851851851852,
+      "grad_norm": 3.9119484424591064,
+      "learning_rate": 1.8790752815649084e-05,
+      "loss": 0.428,
+      "step": 308
+    },
+    {
+      "epoch": 0.1831111111111111,
+      "grad_norm": 1.6751925945281982,
+      "learning_rate": 1.878680102746493e-05,
+      "loss": 0.5096,
+      "step": 309
+    },
+    {
+      "epoch": 0.1837037037037037,
+      "grad_norm": 1.928195834159851,
+      "learning_rate": 1.8782849239280777e-05,
+      "loss": 0.5217,
+      "step": 310
+    },
+    {
+      "epoch": 0.1842962962962963,
+      "grad_norm": 2.79538893699646,
+      "learning_rate": 1.877889745109662e-05,
+      "loss": 0.5535,
+      "step": 311
+    },
+    {
+      "epoch": 0.18488888888888888,
+      "grad_norm": 3.399937152862549,
+      "learning_rate": 1.877494566291247e-05,
+      "loss": 0.6239,
+      "step": 312
+    },
+    {
+      "epoch": 0.18548148148148147,
+      "grad_norm": 3.6707208156585693,
+      "learning_rate": 1.8770993874728317e-05,
+      "loss": 0.5001,
+      "step": 313
+    },
+    {
+      "epoch": 0.1860740740740741,
+      "grad_norm": 2.1471364498138428,
+      "learning_rate": 1.876704208654416e-05,
+      "loss": 0.4507,
+      "step": 314
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 1.9385000467300415,
+      "learning_rate": 1.876309029836001e-05,
+      "loss": 0.5707,
+      "step": 315
+    },
+    {
+      "epoch": 0.18725925925925926,
+      "grad_norm": 2.880038261413574,
+      "learning_rate": 1.8759138510175857e-05,
+      "loss": 0.5364,
+      "step": 316
+    },
+    {
+      "epoch": 0.18785185185185185,
+      "grad_norm": 2.605193614959717,
+      "learning_rate": 1.8755186721991702e-05,
+      "loss": 0.5582,
+      "step": 317
+    },
+    {
+      "epoch": 0.18844444444444444,
+      "grad_norm": 2.5736889839172363,
+      "learning_rate": 1.875123493380755e-05,
+      "loss": 0.5031,
+      "step": 318
+    },
+    {
+      "epoch": 0.18903703703703703,
+      "grad_norm": 2.3642890453338623,
+      "learning_rate": 1.8747283145623394e-05,
+      "loss": 0.5284,
+      "step": 319
+    },
+    {
+      "epoch": 0.18962962962962962,
+      "grad_norm": 1.7185070514678955,
+      "learning_rate": 1.8743331357439242e-05,
+      "loss": 0.5418,
+      "step": 320
+    },
+    {
+      "epoch": 0.1902222222222222,
+      "grad_norm": 3.7714345455169678,
+      "learning_rate": 1.873937956925509e-05,
+      "loss": 0.6059,
+      "step": 321
+    },
+    {
+      "epoch": 0.19081481481481483,
+      "grad_norm": 2.400514841079712,
+      "learning_rate": 1.8735427781070935e-05,
+      "loss": 0.5655,
+      "step": 322
+    },
+    {
+      "epoch": 0.19140740740740741,
+      "grad_norm": 2.254185438156128,
+      "learning_rate": 1.8731475992886783e-05,
+      "loss": 0.543,
+      "step": 323
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 2.881833076477051,
+      "learning_rate": 1.872752420470263e-05,
+      "loss": 0.5523,
+      "step": 324
+    },
+    {
+      "epoch": 0.1925925925925926,
+      "grad_norm": 3.209739923477173,
+      "learning_rate": 1.8723572416518475e-05,
+      "loss": 0.4653,
+      "step": 325
+    },
+    {
+      "epoch": 0.19318518518518518,
+      "grad_norm": 3.9774537086486816,
+      "learning_rate": 1.8719620628334323e-05,
+      "loss": 0.5369,
+      "step": 326
+    },
+    {
+      "epoch": 0.19377777777777777,
+      "grad_norm": 2.5234620571136475,
+      "learning_rate": 1.871566884015017e-05,
+      "loss": 0.5561,
+      "step": 327
+    },
+    {
+      "epoch": 0.19437037037037036,
+      "grad_norm": 2.214995861053467,
+      "learning_rate": 1.8711717051966016e-05,
+      "loss": 0.4046,
+      "step": 328
+    },
+    {
+      "epoch": 0.19496296296296298,
+      "grad_norm": 1.8406845331192017,
+      "learning_rate": 1.8707765263781864e-05,
+      "loss": 0.4458,
+      "step": 329
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 8.582117080688477,
+      "learning_rate": 1.8703813475597708e-05,
+      "loss": 0.9309,
+      "step": 330
+    },
+    {
+      "epoch": 0.19614814814814815,
+      "grad_norm": 4.1338019371032715,
+      "learning_rate": 1.8699861687413556e-05,
+      "loss": 0.5031,
+      "step": 331
+    },
+    {
+      "epoch": 0.19674074074074074,
+      "grad_norm": 4.926871299743652,
+      "learning_rate": 1.8695909899229404e-05,
+      "loss": 0.6799,
+      "step": 332
+    },
+    {
+      "epoch": 0.19733333333333333,
+      "grad_norm": 2.475904703140259,
+      "learning_rate": 1.869195811104525e-05,
+      "loss": 0.6227,
+      "step": 333
+    },
+    {
+      "epoch": 0.19792592592592592,
+      "grad_norm": 3.296563148498535,
+      "learning_rate": 1.8688006322861097e-05,
+      "loss": 0.6584,
+      "step": 334
+    },
+    {
+      "epoch": 0.1985185185185185,
+      "grad_norm": 1.9598593711853027,
+      "learning_rate": 1.8684054534676944e-05,
+      "loss": 0.5306,
+      "step": 335
+    },
+    {
+      "epoch": 0.1991111111111111,
+      "grad_norm": 3.59101939201355,
+      "learning_rate": 1.868010274649279e-05,
+      "loss": 0.6287,
+      "step": 336
+    },
+    {
+      "epoch": 0.19970370370370372,
+      "grad_norm": 4.404383659362793,
+      "learning_rate": 1.8676150958308637e-05,
+      "loss": 0.5458,
+      "step": 337
+    },
+    {
+      "epoch": 0.2002962962962963,
+      "grad_norm": 3.640699863433838,
+      "learning_rate": 1.867219917012448e-05,
+      "loss": 0.5965,
+      "step": 338
+    },
+    {
+      "epoch": 0.2008888888888889,
+      "grad_norm": 2.7888059616088867,
+      "learning_rate": 1.866824738194033e-05,
+      "loss": 0.4657,
+      "step": 339
+    },
+    {
+      "epoch": 0.20148148148148148,
+      "grad_norm": 1.6770066022872925,
+      "learning_rate": 1.8664295593756177e-05,
+      "loss": 0.537,
+      "step": 340
+    },
+    {
+      "epoch": 0.20207407407407407,
+      "grad_norm": 2.7858057022094727,
+      "learning_rate": 1.8660343805572022e-05,
+      "loss": 0.6477,
+      "step": 341
+    },
+    {
+      "epoch": 0.20266666666666666,
+      "grad_norm": 2.479156732559204,
+      "learning_rate": 1.865639201738787e-05,
+      "loss": 0.4862,
+      "step": 342
+    },
+    {
+      "epoch": 0.20325925925925925,
+      "grad_norm": 2.6876354217529297,
+      "learning_rate": 1.8652440229203718e-05,
+      "loss": 0.4982,
+      "step": 343
+    },
+    {
+      "epoch": 0.20385185185185184,
+      "grad_norm": 3.623163938522339,
+      "learning_rate": 1.8648488441019562e-05,
+      "loss": 0.5596,
+      "step": 344
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 1.599784016609192,
+      "learning_rate": 1.864453665283541e-05,
+      "loss": 0.458,
+      "step": 345
+    },
+    {
+      "epoch": 0.20503703703703705,
+      "grad_norm": 1.7964986562728882,
+      "learning_rate": 1.8640584864651258e-05,
+      "loss": 0.3593,
+      "step": 346
+    },
+    {
+      "epoch": 0.20562962962962963,
+      "grad_norm": 2.628377914428711,
+      "learning_rate": 1.8636633076467103e-05,
+      "loss": 0.649,
+      "step": 347
+    },
+    {
+      "epoch": 0.20622222222222222,
+      "grad_norm": 2.1253912448883057,
+      "learning_rate": 1.863268128828295e-05,
+      "loss": 0.4166,
+      "step": 348
+    },
+    {
+      "epoch": 0.2068148148148148,
+      "grad_norm": 2.1227195262908936,
+      "learning_rate": 1.8628729500098795e-05,
+      "loss": 0.624,
+      "step": 349
+    },
+    {
+      "epoch": 0.2074074074074074,
+      "grad_norm": 3.80692458152771,
+      "learning_rate": 1.8624777711914643e-05,
+      "loss": 0.5464,
+      "step": 350
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 1.859786868095398,
+      "learning_rate": 1.862082592373049e-05,
+      "loss": 0.454,
+      "step": 351
+    },
+    {
+      "epoch": 0.20859259259259258,
+      "grad_norm": 1.8824098110198975,
+      "learning_rate": 1.8616874135546336e-05,
+      "loss": 0.5325,
+      "step": 352
+    },
+    {
+      "epoch": 0.2091851851851852,
+      "grad_norm": 2.3211655616760254,
+      "learning_rate": 1.8612922347362184e-05,
+      "loss": 0.4375,
+      "step": 353
+    },
+    {
+      "epoch": 0.20977777777777779,
+      "grad_norm": 1.5811023712158203,
+      "learning_rate": 1.860897055917803e-05,
+      "loss": 0.4778,
+      "step": 354
+    },
+    {
+      "epoch": 0.21037037037037037,
+      "grad_norm": 2.0082366466522217,
+      "learning_rate": 1.8605018770993876e-05,
+      "loss": 0.6782,
+      "step": 355
+    },
+    {
+      "epoch": 0.21096296296296296,
+      "grad_norm": 1.4607501029968262,
+      "learning_rate": 1.8601066982809724e-05,
+      "loss": 0.4879,
+      "step": 356
+    },
+    {
+      "epoch": 0.21155555555555555,
+      "grad_norm": 2.297304391860962,
+      "learning_rate": 1.859711519462557e-05,
+      "loss": 0.5225,
+      "step": 357
+    },
+    {
+      "epoch": 0.21214814814814814,
+      "grad_norm": 1.9646096229553223,
+      "learning_rate": 1.8593163406441417e-05,
+      "loss": 0.3864,
+      "step": 358
+    },
+    {
+      "epoch": 0.21274074074074073,
+      "grad_norm": 4.1259565353393555,
+      "learning_rate": 1.8589211618257264e-05,
+      "loss": 0.6917,
+      "step": 359
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 1.7865204811096191,
+      "learning_rate": 1.858525983007311e-05,
+      "loss": 0.4305,
+      "step": 360
+    },
+    {
+      "epoch": 0.21392592592592594,
+      "grad_norm": 2.1542136669158936,
+      "learning_rate": 1.8581308041888957e-05,
+      "loss": 0.4651,
+      "step": 361
+    },
+    {
+      "epoch": 0.21451851851851853,
+      "grad_norm": 2.8617169857025146,
+      "learning_rate": 1.8577356253704805e-05,
+      "loss": 0.4368,
+      "step": 362
+    },
+    {
+      "epoch": 0.21511111111111111,
+      "grad_norm": 2.323131561279297,
+      "learning_rate": 1.857340446552065e-05,
+      "loss": 0.5405,
+      "step": 363
+    },
+    {
+      "epoch": 0.2157037037037037,
+      "grad_norm": 5.647866249084473,
+      "learning_rate": 1.8569452677336497e-05,
+      "loss": 0.7005,
+      "step": 364
+    },
+    {
+      "epoch": 0.2162962962962963,
+      "grad_norm": 2.8186733722686768,
+      "learning_rate": 1.8565500889152342e-05,
+      "loss": 0.6498,
+      "step": 365
+    },
+    {
+      "epoch": 0.21688888888888888,
+      "grad_norm": 2.097808599472046,
+      "learning_rate": 1.856154910096819e-05,
+      "loss": 0.5573,
+      "step": 366
+    },
+    {
+      "epoch": 0.21748148148148147,
+      "grad_norm": 2.234680414199829,
+      "learning_rate": 1.8557597312784038e-05,
+      "loss": 0.472,
+      "step": 367
+    },
+    {
+      "epoch": 0.2180740740740741,
+      "grad_norm": 2.291813611984253,
+      "learning_rate": 1.8553645524599882e-05,
+      "loss": 0.5557,
+      "step": 368
+    },
+    {
+      "epoch": 0.21866666666666668,
+      "grad_norm": 1.680774450302124,
+      "learning_rate": 1.854969373641573e-05,
+      "loss": 0.3836,
+      "step": 369
+    },
+    {
+      "epoch": 0.21925925925925926,
+      "grad_norm": 2.1022019386291504,
+      "learning_rate": 1.8545741948231578e-05,
+      "loss": 0.5396,
+      "step": 370
+    },
+    {
+      "epoch": 0.21985185185185185,
+      "grad_norm": 3.3733205795288086,
+      "learning_rate": 1.8541790160047423e-05,
+      "loss": 0.5308,
+      "step": 371
+    },
+    {
+      "epoch": 0.22044444444444444,
+      "grad_norm": 3.2941739559173584,
+      "learning_rate": 1.853783837186327e-05,
+      "loss": 0.591,
+      "step": 372
+    },
+    {
+      "epoch": 0.22103703703703703,
+      "grad_norm": 1.8328680992126465,
+      "learning_rate": 1.853388658367912e-05,
+      "loss": 0.4564,
+      "step": 373
+    },
+    {
+      "epoch": 0.22162962962962962,
+      "grad_norm": 1.91807222366333,
+      "learning_rate": 1.8529934795494963e-05,
+      "loss": 0.4161,
+      "step": 374
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 3.105701446533203,
+      "learning_rate": 1.852598300731081e-05,
+      "loss": 0.4633,
+      "step": 375
+    },
+    {
+      "epoch": 0.22281481481481483,
+      "grad_norm": 4.3208465576171875,
+      "learning_rate": 1.8522031219126656e-05,
+      "loss": 0.801,
+      "step": 376
+    },
+    {
+      "epoch": 0.22340740740740742,
+      "grad_norm": 3.7107362747192383,
+      "learning_rate": 1.8518079430942504e-05,
+      "loss": 0.5507,
+      "step": 377
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 2.348876953125,
+      "learning_rate": 1.851412764275835e-05,
+      "loss": 0.5133,
+      "step": 378
+    },
+    {
+      "epoch": 0.2245925925925926,
+      "grad_norm": 2.2960000038146973,
+      "learning_rate": 1.8510175854574196e-05,
+      "loss": 0.5765,
+      "step": 379
+    },
+    {
+      "epoch": 0.22518518518518518,
+      "grad_norm": 1.7323541641235352,
+      "learning_rate": 1.8506224066390044e-05,
+      "loss": 0.4271,
+      "step": 380
+    },
+    {
+      "epoch": 0.22577777777777777,
+      "grad_norm": 1.8389445543289185,
+      "learning_rate": 1.8502272278205892e-05,
+      "loss": 0.4786,
+      "step": 381
+    },
+    {
+      "epoch": 0.22637037037037036,
+      "grad_norm": 1.883070468902588,
+      "learning_rate": 1.8498320490021737e-05,
+      "loss": 0.4201,
+      "step": 382
+    },
+    {
+      "epoch": 0.22696296296296295,
+      "grad_norm": 2.320633888244629,
+      "learning_rate": 1.8494368701837584e-05,
+      "loss": 0.4817,
+      "step": 383
+    },
+    {
+      "epoch": 0.22755555555555557,
+      "grad_norm": 2.2077395915985107,
+      "learning_rate": 1.849041691365343e-05,
+      "loss": 0.518,
+      "step": 384
+    },
+    {
+      "epoch": 0.22814814814814816,
+      "grad_norm": 1.8962241411209106,
+      "learning_rate": 1.8486465125469277e-05,
+      "loss": 0.4782,
+      "step": 385
+    },
+    {
+      "epoch": 0.22874074074074074,
+      "grad_norm": 2.3410565853118896,
+      "learning_rate": 1.8482513337285125e-05,
+      "loss": 0.5012,
+      "step": 386
+    },
+    {
+      "epoch": 0.22933333333333333,
+      "grad_norm": 2.7748098373413086,
+      "learning_rate": 1.847856154910097e-05,
+      "loss": 0.5763,
+      "step": 387
+    },
+    {
+      "epoch": 0.22992592592592592,
+      "grad_norm": 4.009687423706055,
+      "learning_rate": 1.8474609760916817e-05,
+      "loss": 0.4468,
+      "step": 388
+    },
+    {
+      "epoch": 0.2305185185185185,
+      "grad_norm": 2.4330830574035645,
+      "learning_rate": 1.8470657972732665e-05,
+      "loss": 0.6003,
+      "step": 389
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 4.087474346160889,
+      "learning_rate": 1.846670618454851e-05,
+      "loss": 0.7499,
+      "step": 390
+    },
+    {
+      "epoch": 0.23170370370370372,
+      "grad_norm": 3.378931760787964,
+      "learning_rate": 1.8462754396364358e-05,
+      "loss": 0.6191,
+      "step": 391
+    },
+    {
+      "epoch": 0.2322962962962963,
+      "grad_norm": 2.1903679370880127,
+      "learning_rate": 1.8458802608180206e-05,
+      "loss": 0.4693,
+      "step": 392
+    },
+    {
+      "epoch": 0.2328888888888889,
+      "grad_norm": 2.904266595840454,
+      "learning_rate": 1.845485081999605e-05,
+      "loss": 0.5632,
+      "step": 393
+    },
+    {
+      "epoch": 0.23348148148148148,
+      "grad_norm": 2.3514697551727295,
+      "learning_rate": 1.8450899031811898e-05,
+      "loss": 0.4953,
+      "step": 394
+    },
+    {
+      "epoch": 0.23407407407407407,
+      "grad_norm": 2.72554874420166,
+      "learning_rate": 1.8446947243627743e-05,
+      "loss": 0.4678,
+      "step": 395
+    },
+    {
+      "epoch": 0.23466666666666666,
+      "grad_norm": 4.713881015777588,
+      "learning_rate": 1.844299545544359e-05,
+      "loss": 0.6228,
+      "step": 396
+    },
+    {
+      "epoch": 0.23525925925925925,
+      "grad_norm": 3.264920473098755,
+      "learning_rate": 1.843904366725944e-05,
+      "loss": 0.441,
+      "step": 397
+    },
+    {
+      "epoch": 0.23585185185185184,
+      "grad_norm": 2.2753679752349854,
+      "learning_rate": 1.8435091879075283e-05,
+      "loss": 0.5503,
+      "step": 398
+    },
+    {
+      "epoch": 0.23644444444444446,
+      "grad_norm": 1.5795868635177612,
+      "learning_rate": 1.8431140090891128e-05,
+      "loss": 0.4101,
+      "step": 399
+    },
+    {
+      "epoch": 0.23703703703703705,
+      "grad_norm": 2.929638624191284,
+      "learning_rate": 1.8427188302706976e-05,
+      "loss": 0.6192,
+      "step": 400
+    },
+    {
+      "epoch": 0.23762962962962964,
+      "grad_norm": 3.5968801975250244,
+      "learning_rate": 1.8423236514522824e-05,
+      "loss": 0.6679,
+      "step": 401
+    },
+    {
+      "epoch": 0.23822222222222222,
+      "grad_norm": 1.9605062007904053,
+      "learning_rate": 1.8419284726338668e-05,
+      "loss": 0.4834,
+      "step": 402
+    },
+    {
+      "epoch": 0.2388148148148148,
+      "grad_norm": 1.5270830392837524,
+      "learning_rate": 1.8415332938154516e-05,
+      "loss": 0.4982,
+      "step": 403
+    },
+    {
+      "epoch": 0.2394074074074074,
+      "grad_norm": 1.716869592666626,
+      "learning_rate": 1.841138114997036e-05,
+      "loss": 0.446,
+      "step": 404
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4413650035858154,
+      "learning_rate": 1.840742936178621e-05,
+      "loss": 0.4163,
+      "step": 405
+    },
+    {
+      "epoch": 0.24059259259259258,
+      "grad_norm": 1.606567144393921,
+      "learning_rate": 1.8403477573602057e-05,
+      "loss": 0.4925,
+      "step": 406
+    },
+    {
+      "epoch": 0.2411851851851852,
+      "grad_norm": 1.3578121662139893,
+      "learning_rate": 1.83995257854179e-05,
+      "loss": 0.4775,
+      "step": 407
+    },
+    {
+      "epoch": 0.24177777777777779,
+      "grad_norm": 2.5981223583221436,
+      "learning_rate": 1.839557399723375e-05,
+      "loss": 0.55,
+      "step": 408
+    },
+    {
+      "epoch": 0.24237037037037037,
+      "grad_norm": 1.5464465618133545,
+      "learning_rate": 1.8391622209049597e-05,
+      "loss": 0.4725,
+      "step": 409
+    },
+    {
+      "epoch": 0.24296296296296296,
+      "grad_norm": 4.732463359832764,
+      "learning_rate": 1.838767042086544e-05,
+      "loss": 0.6061,
+      "step": 410
+    },
+    {
+      "epoch": 0.24355555555555555,
+      "grad_norm": 2.8851571083068848,
+      "learning_rate": 1.838371863268129e-05,
+      "loss": 0.4257,
+      "step": 411
+    },
+    {
+      "epoch": 0.24414814814814814,
+      "grad_norm": 2.9463212490081787,
+      "learning_rate": 1.8379766844497134e-05,
+      "loss": 0.6483,
+      "step": 412
+    },
+    {
+      "epoch": 0.24474074074074073,
+      "grad_norm": 1.683834433555603,
+      "learning_rate": 1.8375815056312982e-05,
+      "loss": 0.5151,
+      "step": 413
+    },
+    {
+      "epoch": 0.24533333333333332,
+      "grad_norm": 2.3084797859191895,
+      "learning_rate": 1.837186326812883e-05,
+      "loss": 0.4943,
+      "step": 414
+    },
+    {
+      "epoch": 0.24592592592592594,
+      "grad_norm": 2.9474170207977295,
+      "learning_rate": 1.8367911479944674e-05,
+      "loss": 0.6064,
+      "step": 415
+    },
+    {
+      "epoch": 0.24651851851851853,
+      "grad_norm": 2.0990829467773438,
+      "learning_rate": 1.8363959691760522e-05,
+      "loss": 0.484,
+      "step": 416
+    },
+    {
+      "epoch": 0.24711111111111111,
+      "grad_norm": 3.2794294357299805,
+      "learning_rate": 1.836000790357637e-05,
+      "loss": 0.5391,
+      "step": 417
+    },
+    {
+      "epoch": 0.2477037037037037,
+      "grad_norm": 3.0696778297424316,
+      "learning_rate": 1.8356056115392215e-05,
+      "loss": 0.446,
+      "step": 418
+    },
+    {
+      "epoch": 0.2482962962962963,
+      "grad_norm": 2.972454071044922,
+      "learning_rate": 1.8352104327208063e-05,
+      "loss": 0.5963,
+      "step": 419
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 2.7924375534057617,
+      "learning_rate": 1.834815253902391e-05,
+      "loss": 0.6128,
+      "step": 420
+    },
+    {
+      "epoch": 0.24948148148148147,
+      "grad_norm": 2.886996030807495,
+      "learning_rate": 1.8344200750839755e-05,
+      "loss": 0.5706,
+      "step": 421
+    },
+    {
+      "epoch": 0.25007407407407406,
+      "grad_norm": 2.2148282527923584,
+      "learning_rate": 1.8340248962655603e-05,
+      "loss": 0.4677,
+      "step": 422
+    },
+    {
+      "epoch": 0.25066666666666665,
+      "grad_norm": 1.8105928897857666,
+      "learning_rate": 1.8336297174471448e-05,
+      "loss": 0.5709,
+      "step": 423
+    },
+    {
+      "epoch": 0.25125925925925924,
+      "grad_norm": 1.9336295127868652,
+      "learning_rate": 1.8332345386287296e-05,
+      "loss": 0.4911,
+      "step": 424
+    },
+    {
+      "epoch": 0.2518518518518518,
+      "grad_norm": 1.3766508102416992,
+      "learning_rate": 1.8328393598103144e-05,
+      "loss": 0.4303,
+      "step": 425
+    },
+    {
+      "epoch": 0.25244444444444447,
+      "grad_norm": 1.1964658498764038,
+      "learning_rate": 1.8324441809918988e-05,
+      "loss": 0.2973,
+      "step": 426
+    },
+    {
+      "epoch": 0.25303703703703706,
+      "grad_norm": 1.4528887271881104,
+      "learning_rate": 1.8320490021734836e-05,
+      "loss": 0.4305,
+      "step": 427
+    },
+    {
+      "epoch": 0.25362962962962965,
+      "grad_norm": 3.0747618675231934,
+      "learning_rate": 1.8316538233550684e-05,
+      "loss": 0.5927,
+      "step": 428
+    },
+    {
+      "epoch": 0.25422222222222224,
+      "grad_norm": 1.9362715482711792,
+      "learning_rate": 1.831258644536653e-05,
+      "loss": 0.4507,
+      "step": 429
+    },
+    {
+      "epoch": 0.2548148148148148,
+      "grad_norm": 2.7456164360046387,
+      "learning_rate": 1.8308634657182377e-05,
+      "loss": 0.5125,
+      "step": 430
+    },
+    {
+      "epoch": 0.2554074074074074,
+      "grad_norm": 2.7088091373443604,
+      "learning_rate": 1.830468286899822e-05,
+      "loss": 0.7237,
+      "step": 431
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.784407615661621,
+      "learning_rate": 1.830073108081407e-05,
+      "loss": 0.4684,
+      "step": 432
+    },
+    {
+      "epoch": 0.2565925925925926,
+      "grad_norm": 2.26371431350708,
+      "learning_rate": 1.8296779292629917e-05,
+      "loss": 0.516,
+      "step": 433
+    },
+    {
+      "epoch": 0.2571851851851852,
+      "grad_norm": 4.512001037597656,
+      "learning_rate": 1.829282750444576e-05,
+      "loss": 0.5454,
+      "step": 434
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 2.8425097465515137,
+      "learning_rate": 1.828887571626161e-05,
+      "loss": 0.5813,
+      "step": 435
+    },
+    {
+      "epoch": 0.25837037037037036,
+      "grad_norm": 2.115699052810669,
+      "learning_rate": 1.8284923928077457e-05,
+      "loss": 0.5377,
+      "step": 436
+    },
+    {
+      "epoch": 0.25896296296296295,
+      "grad_norm": 1.617918610572815,
+      "learning_rate": 1.8280972139893302e-05,
+      "loss": 0.43,
+      "step": 437
+    },
+    {
+      "epoch": 0.25955555555555554,
+      "grad_norm": 2.1052541732788086,
+      "learning_rate": 1.827702035170915e-05,
+      "loss": 0.4954,
+      "step": 438
+    },
+    {
+      "epoch": 0.26014814814814813,
+      "grad_norm": 1.7497344017028809,
+      "learning_rate": 1.8273068563524994e-05,
+      "loss": 0.4452,
+      "step": 439
+    },
+    {
+      "epoch": 0.2607407407407407,
+      "grad_norm": 2.779019832611084,
+      "learning_rate": 1.8269116775340842e-05,
+      "loss": 0.5012,
+      "step": 440
+    },
+    {
+      "epoch": 0.2613333333333333,
+      "grad_norm": 1.9839228391647339,
+      "learning_rate": 1.826516498715669e-05,
+      "loss": 0.4263,
+      "step": 441
+    },
+    {
+      "epoch": 0.26192592592592595,
+      "grad_norm": 2.2786829471588135,
+      "learning_rate": 1.8261213198972535e-05,
+      "loss": 0.3591,
+      "step": 442
+    },
+    {
+      "epoch": 0.26251851851851854,
+      "grad_norm": 1.2465208768844604,
+      "learning_rate": 1.8257261410788383e-05,
+      "loss": 0.3972,
+      "step": 443
+    },
+    {
+      "epoch": 0.26311111111111113,
+      "grad_norm": 2.9657158851623535,
+      "learning_rate": 1.825330962260423e-05,
+      "loss": 0.5099,
+      "step": 444
+    },
+    {
+      "epoch": 0.2637037037037037,
+      "grad_norm": 3.691671371459961,
+      "learning_rate": 1.8249357834420075e-05,
+      "loss": 0.5844,
+      "step": 445
+    },
+    {
+      "epoch": 0.2642962962962963,
+      "grad_norm": 1.3381620645523071,
+      "learning_rate": 1.8245406046235923e-05,
+      "loss": 0.4228,
+      "step": 446
+    },
+    {
+      "epoch": 0.2648888888888889,
+      "grad_norm": 3.584442138671875,
+      "learning_rate": 1.824145425805177e-05,
+      "loss": 0.6497,
+      "step": 447
+    },
+    {
+      "epoch": 0.2654814814814815,
+      "grad_norm": 3.4696805477142334,
+      "learning_rate": 1.8237502469867616e-05,
+      "loss": 0.47,
+      "step": 448
+    },
+    {
+      "epoch": 0.2660740740740741,
+      "grad_norm": 1.9171570539474487,
+      "learning_rate": 1.8233550681683464e-05,
+      "loss": 0.4831,
+      "step": 449
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 2.6452527046203613,
+      "learning_rate": 1.8229598893499308e-05,
+      "loss": 0.5479,
+      "step": 450
+    },
+    {
+      "epoch": 0.26725925925925925,
+      "grad_norm": 1.554205060005188,
+      "learning_rate": 1.8225647105315156e-05,
+      "loss": 0.3973,
+      "step": 451
+    },
+    {
+      "epoch": 0.26785185185185184,
+      "grad_norm": 2.029510021209717,
+      "learning_rate": 1.8221695317131004e-05,
+      "loss": 0.4833,
+      "step": 452
+    },
+    {
+      "epoch": 0.26844444444444443,
+      "grad_norm": 1.4451189041137695,
+      "learning_rate": 1.821774352894685e-05,
+      "loss": 0.4716,
+      "step": 453
+    },
+    {
+      "epoch": 0.269037037037037,
+      "grad_norm": 6.500146865844727,
+      "learning_rate": 1.8213791740762697e-05,
+      "loss": 0.6825,
+      "step": 454
+    },
+    {
+      "epoch": 0.2696296296296296,
+      "grad_norm": 3.8132314682006836,
+      "learning_rate": 1.8209839952578544e-05,
+      "loss": 0.5502,
+      "step": 455
+    },
+    {
+      "epoch": 0.2702222222222222,
+      "grad_norm": 2.2256932258605957,
+      "learning_rate": 1.820588816439439e-05,
+      "loss": 0.4736,
+      "step": 456
+    },
+    {
+      "epoch": 0.27081481481481484,
+      "grad_norm": 3.1733031272888184,
+      "learning_rate": 1.8201936376210237e-05,
+      "loss": 0.5009,
+      "step": 457
+    },
+    {
+      "epoch": 0.27140740740740743,
+      "grad_norm": 3.608124017715454,
+      "learning_rate": 1.819798458802608e-05,
+      "loss": 0.4402,
+      "step": 458
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 1.8988388776779175,
+      "learning_rate": 1.819403279984193e-05,
+      "loss": 0.4948,
+      "step": 459
+    },
+    {
+      "epoch": 0.2725925925925926,
+      "grad_norm": 4.190664291381836,
+      "learning_rate": 1.8190081011657777e-05,
+      "loss": 0.4602,
+      "step": 460
+    },
+    {
+      "epoch": 0.2731851851851852,
+      "grad_norm": 2.4542839527130127,
+      "learning_rate": 1.8186129223473622e-05,
+      "loss": 0.4975,
+      "step": 461
+    },
+    {
+      "epoch": 0.2737777777777778,
+      "grad_norm": 1.7694164514541626,
+      "learning_rate": 1.818217743528947e-05,
+      "loss": 0.4789,
+      "step": 462
+    },
+    {
+      "epoch": 0.2743703703703704,
+      "grad_norm": 1.7848409414291382,
+      "learning_rate": 1.8178225647105318e-05,
+      "loss": 0.3999,
+      "step": 463
+    },
+    {
+      "epoch": 0.27496296296296296,
+      "grad_norm": 4.788225173950195,
+      "learning_rate": 1.8174273858921162e-05,
+      "loss": 0.6271,
+      "step": 464
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 2.7896904945373535,
+      "learning_rate": 1.817032207073701e-05,
+      "loss": 0.4903,
+      "step": 465
+    },
+    {
+      "epoch": 0.27614814814814814,
+      "grad_norm": 2.406104564666748,
+      "learning_rate": 1.8166370282552855e-05,
+      "loss": 0.598,
+      "step": 466
+    },
+    {
+      "epoch": 0.27674074074074073,
+      "grad_norm": 2.756197929382324,
+      "learning_rate": 1.8162418494368703e-05,
+      "loss": 0.4021,
+      "step": 467
+    },
+    {
+      "epoch": 0.2773333333333333,
+      "grad_norm": 2.4852449893951416,
+      "learning_rate": 1.815846670618455e-05,
+      "loss": 0.4975,
+      "step": 468
+    },
+    {
+      "epoch": 0.2779259259259259,
+      "grad_norm": 2.1789650917053223,
+      "learning_rate": 1.8154514918000395e-05,
+      "loss": 0.4406,
+      "step": 469
+    },
+    {
+      "epoch": 0.2785185185185185,
+      "grad_norm": 2.3887224197387695,
+      "learning_rate": 1.8150563129816243e-05,
+      "loss": 0.5471,
+      "step": 470
+    },
+    {
+      "epoch": 0.2791111111111111,
+      "grad_norm": 2.2593002319335938,
+      "learning_rate": 1.814661134163209e-05,
+      "loss": 0.498,
+      "step": 471
+    },
+    {
+      "epoch": 0.2797037037037037,
+      "grad_norm": 3.4869823455810547,
+      "learning_rate": 1.8142659553447936e-05,
+      "loss": 0.5418,
+      "step": 472
+    },
+    {
+      "epoch": 0.2802962962962963,
+      "grad_norm": 2.9183638095855713,
+      "learning_rate": 1.8138707765263784e-05,
+      "loss": 0.451,
+      "step": 473
+    },
+    {
+      "epoch": 0.2808888888888889,
+      "grad_norm": 2.2074339389801025,
+      "learning_rate": 1.813475597707963e-05,
+      "loss": 0.4392,
+      "step": 474
+    },
+    {
+      "epoch": 0.2814814814814815,
+      "grad_norm": 1.8651514053344727,
+      "learning_rate": 1.8130804188895476e-05,
+      "loss": 0.451,
+      "step": 475
+    },
+    {
+      "epoch": 0.2820740740740741,
+      "grad_norm": 1.8148307800292969,
+      "learning_rate": 1.8126852400711324e-05,
+      "loss": 0.477,
+      "step": 476
+    },
+    {
+      "epoch": 0.2826666666666667,
+      "grad_norm": 1.716787338256836,
+      "learning_rate": 1.812290061252717e-05,
+      "loss": 0.3797,
+      "step": 477
+    },
+    {
+      "epoch": 0.28325925925925927,
+      "grad_norm": 1.7248690128326416,
+      "learning_rate": 1.8118948824343017e-05,
+      "loss": 0.4898,
+      "step": 478
+    },
+    {
+      "epoch": 0.28385185185185186,
+      "grad_norm": 2.2242603302001953,
+      "learning_rate": 1.8114997036158864e-05,
+      "loss": 0.5388,
+      "step": 479
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 2.110924482345581,
+      "learning_rate": 1.811104524797471e-05,
+      "loss": 0.4047,
+      "step": 480
+    },
+    {
+      "epoch": 0.28503703703703703,
+      "grad_norm": 1.409021258354187,
+      "learning_rate": 1.8107093459790557e-05,
+      "loss": 0.4442,
+      "step": 481
+    },
+    {
+      "epoch": 0.2856296296296296,
+      "grad_norm": 2.648329257965088,
+      "learning_rate": 1.8103141671606405e-05,
+      "loss": 0.5797,
+      "step": 482
+    },
+    {
+      "epoch": 0.2862222222222222,
+      "grad_norm": 2.1465585231781006,
+      "learning_rate": 1.809918988342225e-05,
+      "loss": 0.5105,
+      "step": 483
+    },
+    {
+      "epoch": 0.2868148148148148,
+      "grad_norm": 3.0427465438842773,
+      "learning_rate": 1.8095238095238097e-05,
+      "loss": 0.6048,
+      "step": 484
+    },
+    {
+      "epoch": 0.2874074074074074,
+      "grad_norm": 3.310042142868042,
+      "learning_rate": 1.8091286307053942e-05,
+      "loss": 0.5192,
+      "step": 485
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 4.687093257904053,
+      "learning_rate": 1.808733451886979e-05,
+      "loss": 0.6189,
+      "step": 486
+    },
+    {
+      "epoch": 0.28859259259259257,
+      "grad_norm": 3.0052645206451416,
+      "learning_rate": 1.8083382730685638e-05,
+      "loss": 0.501,
+      "step": 487
+    },
+    {
+      "epoch": 0.2891851851851852,
+      "grad_norm": 2.9586539268493652,
+      "learning_rate": 1.8079430942501482e-05,
+      "loss": 0.5351,
+      "step": 488
+    },
+    {
+      "epoch": 0.2897777777777778,
+      "grad_norm": 2.685770034790039,
+      "learning_rate": 1.807547915431733e-05,
+      "loss": 0.4578,
+      "step": 489
+    },
+    {
+      "epoch": 0.2903703703703704,
+      "grad_norm": 2.226623773574829,
+      "learning_rate": 1.8071527366133178e-05,
+      "loss": 0.4896,
+      "step": 490
+    },
+    {
+      "epoch": 0.290962962962963,
+      "grad_norm": 2.331883192062378,
+      "learning_rate": 1.8067575577949023e-05,
+      "loss": 0.4991,
+      "step": 491
+    },
+    {
+      "epoch": 0.29155555555555557,
+      "grad_norm": 2.8249287605285645,
+      "learning_rate": 1.806362378976487e-05,
+      "loss": 0.4366,
+      "step": 492
+    },
+    {
+      "epoch": 0.29214814814814816,
+      "grad_norm": 3.016862154006958,
+      "learning_rate": 1.805967200158072e-05,
+      "loss": 0.4403,
+      "step": 493
+    },
+    {
+      "epoch": 0.29274074074074075,
+      "grad_norm": 3.372351884841919,
+      "learning_rate": 1.8055720213396563e-05,
+      "loss": 0.5367,
+      "step": 494
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 1.8267637491226196,
+      "learning_rate": 1.805176842521241e-05,
+      "loss": 0.4152,
+      "step": 495
+    },
+    {
+      "epoch": 0.2939259259259259,
+      "grad_norm": 2.2756102085113525,
+      "learning_rate": 1.8047816637028256e-05,
+      "loss": 0.5168,
+      "step": 496
+    },
+    {
+      "epoch": 0.2945185185185185,
+      "grad_norm": 2.5963470935821533,
+      "learning_rate": 1.8043864848844104e-05,
+      "loss": 0.4845,
+      "step": 497
+    },
+    {
+      "epoch": 0.2951111111111111,
+      "grad_norm": 2.169074773788452,
+      "learning_rate": 1.803991306065995e-05,
+      "loss": 0.5859,
+      "step": 498
+    },
+    {
+      "epoch": 0.2957037037037037,
+      "grad_norm": 1.7161184549331665,
+      "learning_rate": 1.8035961272475796e-05,
+      "loss": 0.3991,
+      "step": 499
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 3.72123384475708,
+      "learning_rate": 1.8032009484291644e-05,
+      "loss": 0.6731,
+      "step": 500
+    },
+    {
+      "epoch": 0.29688888888888887,
+      "grad_norm": 2.6333744525909424,
+      "learning_rate": 1.8028057696107492e-05,
+      "loss": 0.4696,
+      "step": 501
+    },
+    {
+      "epoch": 0.29748148148148146,
+      "grad_norm": 2.2341761589050293,
+      "learning_rate": 1.8024105907923337e-05,
+      "loss": 0.5744,
+      "step": 502
+    },
+    {
+      "epoch": 0.29807407407407405,
+      "grad_norm": 2.4152097702026367,
+      "learning_rate": 1.8020154119739184e-05,
+      "loss": 0.4313,
+      "step": 503
+    },
+    {
+      "epoch": 0.2986666666666667,
+      "grad_norm": 1.737403392791748,
+      "learning_rate": 1.801620233155503e-05,
+      "loss": 0.5058,
+      "step": 504
+    },
+    {
+      "epoch": 0.2992592592592593,
+      "grad_norm": 2.153754949569702,
+      "learning_rate": 1.8012250543370877e-05,
+      "loss": 0.4041,
+      "step": 505
+    },
+    {
+      "epoch": 0.29985185185185187,
+      "grad_norm": 6.76655387878418,
+      "learning_rate": 1.8008298755186725e-05,
+      "loss": 0.7339,
+      "step": 506
+    },
+    {
+      "epoch": 0.30044444444444446,
+      "grad_norm": 1.9587161540985107,
+      "learning_rate": 1.800434696700257e-05,
+      "loss": 0.47,
+      "step": 507
+    },
+    {
+      "epoch": 0.30103703703703705,
+      "grad_norm": 1.7894585132598877,
+      "learning_rate": 1.8000395178818417e-05,
+      "loss": 0.4546,
+      "step": 508
+    },
+    {
+      "epoch": 0.30162962962962964,
+      "grad_norm": 2.0707993507385254,
+      "learning_rate": 1.7996443390634265e-05,
+      "loss": 0.5567,
+      "step": 509
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 2.9826173782348633,
+      "learning_rate": 1.799249160245011e-05,
+      "loss": 0.4581,
+      "step": 510
+    },
+    {
+      "epoch": 0.3028148148148148,
+      "grad_norm": 1.8411195278167725,
+      "learning_rate": 1.7988539814265958e-05,
+      "loss": 0.4169,
+      "step": 511
+    },
+    {
+      "epoch": 0.3034074074074074,
+      "grad_norm": 2.485487937927246,
+      "learning_rate": 1.7984588026081802e-05,
+      "loss": 0.6536,
+      "step": 512
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 1.8039320707321167,
+      "learning_rate": 1.798063623789765e-05,
+      "loss": 0.4378,
+      "step": 513
+    },
+    {
+      "epoch": 0.3045925925925926,
+      "grad_norm": 2.040472984313965,
+      "learning_rate": 1.7976684449713498e-05,
+      "loss": 0.4082,
+      "step": 514
+    },
+    {
+      "epoch": 0.30518518518518517,
+      "grad_norm": 6.858224391937256,
+      "learning_rate": 1.7972732661529343e-05,
+      "loss": 0.7139,
+      "step": 515
+    },
+    {
+      "epoch": 0.30577777777777776,
+      "grad_norm": 1.8847365379333496,
+      "learning_rate": 1.796878087334519e-05,
+      "loss": 0.3304,
+      "step": 516
+    },
+    {
+      "epoch": 0.30637037037037035,
+      "grad_norm": 2.2443008422851562,
+      "learning_rate": 1.796482908516104e-05,
+      "loss": 0.4778,
+      "step": 517
+    },
+    {
+      "epoch": 0.30696296296296294,
+      "grad_norm": 1.7769358158111572,
+      "learning_rate": 1.7960877296976883e-05,
+      "loss": 0.4803,
+      "step": 518
+    },
+    {
+      "epoch": 0.3075555555555556,
+      "grad_norm": 1.8991682529449463,
+      "learning_rate": 1.795692550879273e-05,
+      "loss": 0.4983,
+      "step": 519
+    },
+    {
+      "epoch": 0.30814814814814817,
+      "grad_norm": 2.591670274734497,
+      "learning_rate": 1.795297372060858e-05,
+      "loss": 0.5595,
+      "step": 520
+    },
+    {
+      "epoch": 0.30874074074074076,
+      "grad_norm": 2.552835702896118,
+      "learning_rate": 1.7949021932424424e-05,
+      "loss": 0.542,
+      "step": 521
+    },
+    {
+      "epoch": 0.30933333333333335,
+      "grad_norm": 2.5875978469848633,
+      "learning_rate": 1.794507014424027e-05,
+      "loss": 0.6322,
+      "step": 522
+    },
+    {
+      "epoch": 0.30992592592592594,
+      "grad_norm": 1.707445740699768,
+      "learning_rate": 1.7941118356056116e-05,
+      "loss": 0.4759,
+      "step": 523
+    },
+    {
+      "epoch": 0.3105185185185185,
+      "grad_norm": 2.8056752681732178,
+      "learning_rate": 1.7937166567871964e-05,
+      "loss": 0.551,
+      "step": 524
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 3.2457683086395264,
+      "learning_rate": 1.7933214779687812e-05,
+      "loss": 0.4624,
+      "step": 525
+    },
+    {
+      "epoch": 0.3117037037037037,
+      "grad_norm": 2.133510112762451,
+      "learning_rate": 1.7929262991503657e-05,
+      "loss": 0.4426,
+      "step": 526
+    },
+    {
+      "epoch": 0.3122962962962963,
+      "grad_norm": 1.925686240196228,
+      "learning_rate": 1.7925311203319504e-05,
+      "loss": 0.6337,
+      "step": 527
+    },
+    {
+      "epoch": 0.3128888888888889,
+      "grad_norm": 1.244136929512024,
+      "learning_rate": 1.7921359415135352e-05,
+      "loss": 0.3611,
+      "step": 528
+    },
+    {
+      "epoch": 0.31348148148148147,
+      "grad_norm": 1.768754005432129,
+      "learning_rate": 1.7917407626951197e-05,
+      "loss": 0.3699,
+      "step": 529
+    },
+    {
+      "epoch": 0.31407407407407406,
+      "grad_norm": 2.2116715908050537,
+      "learning_rate": 1.7913455838767045e-05,
+      "loss": 0.4844,
+      "step": 530
+    },
+    {
+      "epoch": 0.31466666666666665,
+      "grad_norm": 1.60231614112854,
+      "learning_rate": 1.790950405058289e-05,
+      "loss": 0.4064,
+      "step": 531
+    },
+    {
+      "epoch": 0.31525925925925924,
+      "grad_norm": 1.9998438358306885,
+      "learning_rate": 1.7905552262398737e-05,
+      "loss": 0.4366,
+      "step": 532
+    },
+    {
+      "epoch": 0.31585185185185183,
+      "grad_norm": 5.848787784576416,
+      "learning_rate": 1.7901600474214585e-05,
+      "loss": 0.8291,
+      "step": 533
+    },
+    {
+      "epoch": 0.3164444444444444,
+      "grad_norm": 2.385829448699951,
+      "learning_rate": 1.789764868603043e-05,
+      "loss": 0.5325,
+      "step": 534
+    },
+    {
+      "epoch": 0.31703703703703706,
+      "grad_norm": 3.0892837047576904,
+      "learning_rate": 1.7893696897846278e-05,
+      "loss": 0.3504,
+      "step": 535
+    },
+    {
+      "epoch": 0.31762962962962965,
+      "grad_norm": 2.588989496231079,
+      "learning_rate": 1.7889745109662126e-05,
+      "loss": 0.528,
+      "step": 536
+    },
+    {
+      "epoch": 0.31822222222222224,
+      "grad_norm": 1.693864345550537,
+      "learning_rate": 1.788579332147797e-05,
+      "loss": 0.4924,
+      "step": 537
+    },
+    {
+      "epoch": 0.31881481481481483,
+      "grad_norm": 2.4877865314483643,
+      "learning_rate": 1.7881841533293818e-05,
+      "loss": 0.528,
+      "step": 538
+    },
+    {
+      "epoch": 0.3194074074074074,
+      "grad_norm": 1.909454345703125,
+      "learning_rate": 1.7877889745109666e-05,
+      "loss": 0.4649,
+      "step": 539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0338573455810547,
+      "learning_rate": 1.787393795692551e-05,
+      "loss": 0.4898,
+      "step": 540
+    },
+    {
+      "epoch": 0.3205925925925926,
+      "grad_norm": 1.8881645202636719,
+      "learning_rate": 1.786998616874136e-05,
+      "loss": 0.4506,
+      "step": 541
+    },
+    {
+      "epoch": 0.3211851851851852,
+      "grad_norm": 2.184641122817993,
+      "learning_rate": 1.7866034380557203e-05,
+      "loss": 0.4777,
+      "step": 542
+    },
+    {
+      "epoch": 0.3217777777777778,
+      "grad_norm": 2.0071682929992676,
+      "learning_rate": 1.786208259237305e-05,
+      "loss": 0.5301,
+      "step": 543
+    },
+    {
+      "epoch": 0.32237037037037036,
+      "grad_norm": 2.365307569503784,
+      "learning_rate": 1.78581308041889e-05,
+      "loss": 0.3655,
+      "step": 544
+    },
+    {
+      "epoch": 0.32296296296296295,
+      "grad_norm": 2.359083414077759,
+      "learning_rate": 1.7854179016004744e-05,
+      "loss": 0.4558,
+      "step": 545
+    },
+    {
+      "epoch": 0.32355555555555554,
+      "grad_norm": 3.0280895233154297,
+      "learning_rate": 1.785022722782059e-05,
+      "loss": 0.458,
+      "step": 546
+    },
+    {
+      "epoch": 0.32414814814814813,
+      "grad_norm": 3.5337653160095215,
+      "learning_rate": 1.784627543963644e-05,
+      "loss": 0.4793,
+      "step": 547
+    },
+    {
+      "epoch": 0.3247407407407407,
+      "grad_norm": 2.0983779430389404,
+      "learning_rate": 1.7842323651452284e-05,
+      "loss": 0.3827,
+      "step": 548
+    },
+    {
+      "epoch": 0.3253333333333333,
+      "grad_norm": 1.9279870986938477,
+      "learning_rate": 1.7838371863268132e-05,
+      "loss": 0.4283,
+      "step": 549
+    },
+    {
+      "epoch": 0.32592592592592595,
+      "grad_norm": 1.8846718072891235,
+      "learning_rate": 1.7834420075083976e-05,
+      "loss": 0.4509,
+      "step": 550
+    },
+    {
+      "epoch": 0.32651851851851854,
+      "grad_norm": 1.640023946762085,
+      "learning_rate": 1.7830468286899824e-05,
+      "loss": 0.3749,
+      "step": 551
+    },
+    {
+      "epoch": 0.32711111111111113,
+      "grad_norm": 2.0699543952941895,
+      "learning_rate": 1.7826516498715672e-05,
+      "loss": 0.4092,
+      "step": 552
+    },
+    {
+      "epoch": 0.3277037037037037,
+      "grad_norm": 3.447312116622925,
+      "learning_rate": 1.7822564710531517e-05,
+      "loss": 0.6427,
+      "step": 553
+    },
+    {
+      "epoch": 0.3282962962962963,
+      "grad_norm": 2.0277581214904785,
+      "learning_rate": 1.7818612922347365e-05,
+      "loss": 0.4715,
+      "step": 554
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 2.0481624603271484,
+      "learning_rate": 1.7814661134163213e-05,
+      "loss": 0.5548,
+      "step": 555
+    },
+    {
+      "epoch": 0.3294814814814815,
+      "grad_norm": 1.574076771736145,
+      "learning_rate": 1.7810709345979057e-05,
+      "loss": 0.4632,
+      "step": 556
+    },
+    {
+      "epoch": 0.3300740740740741,
+      "grad_norm": 2.9811863899230957,
+      "learning_rate": 1.7806757557794902e-05,
+      "loss": 0.4743,
+      "step": 557
+    },
+    {
+      "epoch": 0.33066666666666666,
+      "grad_norm": 2.096325397491455,
+      "learning_rate": 1.780280576961075e-05,
+      "loss": 0.4075,
+      "step": 558
+    },
+    {
+      "epoch": 0.33125925925925925,
+      "grad_norm": 1.9038057327270508,
+      "learning_rate": 1.7798853981426594e-05,
+      "loss": 0.5256,
+      "step": 559
+    },
+    {
+      "epoch": 0.33185185185185184,
+      "grad_norm": 1.894836187362671,
+      "learning_rate": 1.7794902193242442e-05,
+      "loss": 0.3798,
+      "step": 560
+    },
+    {
+      "epoch": 0.33244444444444443,
+      "grad_norm": 2.256967544555664,
+      "learning_rate": 1.779095040505829e-05,
+      "loss": 0.5204,
+      "step": 561
+    },
+    {
+      "epoch": 0.333037037037037,
+      "grad_norm": 2.9190285205841064,
+      "learning_rate": 1.7786998616874135e-05,
+      "loss": 0.6275,
+      "step": 562
+    },
+    {
+      "epoch": 0.3336296296296296,
+      "grad_norm": 1.461643099784851,
+      "learning_rate": 1.7783046828689983e-05,
+      "loss": 0.3778,
+      "step": 563
+    },
+    {
+      "epoch": 0.3342222222222222,
+      "grad_norm": 1.9270161390304565,
+      "learning_rate": 1.777909504050583e-05,
+      "loss": 0.4912,
+      "step": 564
+    },
+    {
+      "epoch": 0.3348148148148148,
+      "grad_norm": 1.761772871017456,
+      "learning_rate": 1.7775143252321675e-05,
+      "loss": 0.4684,
+      "step": 565
+    },
+    {
+      "epoch": 0.33540740740740743,
+      "grad_norm": 1.8690305948257446,
+      "learning_rate": 1.7771191464137523e-05,
+      "loss": 0.5067,
+      "step": 566
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 1.8497658967971802,
+      "learning_rate": 1.776723967595337e-05,
+      "loss": 0.3896,
+      "step": 567
+    },
+    {
+      "epoch": 0.3365925925925926,
+      "grad_norm": 4.557904243469238,
+      "learning_rate": 1.7763287887769216e-05,
+      "loss": 0.7556,
+      "step": 568
+    },
+    {
+      "epoch": 0.3371851851851852,
+      "grad_norm": 2.411327600479126,
+      "learning_rate": 1.7759336099585064e-05,
+      "loss": 0.5753,
+      "step": 569
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 2.100950002670288,
+      "learning_rate": 1.7755384311400908e-05,
+      "loss": 0.4579,
+      "step": 570
+    },
+    {
+      "epoch": 0.3383703703703704,
+      "grad_norm": 1.3860840797424316,
+      "learning_rate": 1.7751432523216756e-05,
+      "loss": 0.4235,
+      "step": 571
+    },
+    {
+      "epoch": 0.33896296296296297,
+      "grad_norm": 2.163808822631836,
+      "learning_rate": 1.7747480735032604e-05,
+      "loss": 0.4095,
+      "step": 572
+    },
+    {
+      "epoch": 0.33955555555555555,
+      "grad_norm": 1.7060058116912842,
+      "learning_rate": 1.774352894684845e-05,
+      "loss": 0.437,
+      "step": 573
+    },
+    {
+      "epoch": 0.34014814814814814,
+      "grad_norm": 3.6680541038513184,
+      "learning_rate": 1.7739577158664296e-05,
+      "loss": 0.7008,
+      "step": 574
+    },
+    {
+      "epoch": 0.34074074074074073,
+      "grad_norm": 1.631714940071106,
+      "learning_rate": 1.7735625370480144e-05,
+      "loss": 0.4541,
+      "step": 575
+    },
+    {
+      "epoch": 0.3413333333333333,
+      "grad_norm": 1.6932752132415771,
+      "learning_rate": 1.773167358229599e-05,
+      "loss": 0.4999,
+      "step": 576
+    },
+    {
+      "epoch": 0.3419259259259259,
+      "grad_norm": 2.7662551403045654,
+      "learning_rate": 1.7727721794111837e-05,
+      "loss": 0.5928,
+      "step": 577
+    },
+    {
+      "epoch": 0.3425185185185185,
+      "grad_norm": 1.9387743473052979,
+      "learning_rate": 1.772377000592768e-05,
+      "loss": 0.3855,
+      "step": 578
+    },
+    {
+      "epoch": 0.3431111111111111,
+      "grad_norm": 1.613596796989441,
+      "learning_rate": 1.771981821774353e-05,
+      "loss": 0.4542,
+      "step": 579
+    },
+    {
+      "epoch": 0.3437037037037037,
+      "grad_norm": 1.6049559116363525,
+      "learning_rate": 1.7715866429559377e-05,
+      "loss": 0.4695,
+      "step": 580
+    },
+    {
+      "epoch": 0.3442962962962963,
+      "grad_norm": 1.8409245014190674,
+      "learning_rate": 1.7711914641375222e-05,
+      "loss": 0.4671,
+      "step": 581
+    },
+    {
+      "epoch": 0.3448888888888889,
+      "grad_norm": 1.9014924764633179,
+      "learning_rate": 1.770796285319107e-05,
+      "loss": 0.4849,
+      "step": 582
+    },
+    {
+      "epoch": 0.3454814814814815,
+      "grad_norm": 3.3903725147247314,
+      "learning_rate": 1.7704011065006918e-05,
+      "loss": 0.5377,
+      "step": 583
+    },
+    {
+      "epoch": 0.3460740740740741,
+      "grad_norm": 2.126344680786133,
+      "learning_rate": 1.7700059276822762e-05,
+      "loss": 0.4532,
+      "step": 584
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 1.9364334344863892,
+      "learning_rate": 1.769610748863861e-05,
+      "loss": 0.4528,
+      "step": 585
+    },
+    {
+      "epoch": 0.34725925925925927,
+      "grad_norm": 2.107285499572754,
+      "learning_rate": 1.7692155700454455e-05,
+      "loss": 0.4576,
+      "step": 586
+    },
+    {
+      "epoch": 0.34785185185185186,
+      "grad_norm": 1.9335010051727295,
+      "learning_rate": 1.7688203912270303e-05,
+      "loss": 0.5767,
+      "step": 587
+    },
+    {
+      "epoch": 0.34844444444444445,
+      "grad_norm": 2.4319674968719482,
+      "learning_rate": 1.768425212408615e-05,
+      "loss": 0.5074,
+      "step": 588
+    },
+    {
+      "epoch": 0.34903703703703703,
+      "grad_norm": 1.7924549579620361,
+      "learning_rate": 1.7680300335901995e-05,
+      "loss": 0.3815,
+      "step": 589
+    },
+    {
+      "epoch": 0.3496296296296296,
+      "grad_norm": 1.7605587244033813,
+      "learning_rate": 1.7676348547717843e-05,
+      "loss": 0.4107,
+      "step": 590
+    },
+    {
+      "epoch": 0.3502222222222222,
+      "grad_norm": 1.3815611600875854,
+      "learning_rate": 1.767239675953369e-05,
+      "loss": 0.3622,
+      "step": 591
+    },
+    {
+      "epoch": 0.3508148148148148,
+      "grad_norm": 2.4648361206054688,
+      "learning_rate": 1.7668444971349536e-05,
+      "loss": 0.4662,
+      "step": 592
+    },
+    {
+      "epoch": 0.3514074074074074,
+      "grad_norm": 1.8347736597061157,
+      "learning_rate": 1.7664493183165384e-05,
+      "loss": 0.5298,
+      "step": 593
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 1.6538032293319702,
+      "learning_rate": 1.766054139498123e-05,
+      "loss": 0.4317,
+      "step": 594
+    },
+    {
+      "epoch": 0.35259259259259257,
+      "grad_norm": 2.03703236579895,
+      "learning_rate": 1.7656589606797076e-05,
+      "loss": 0.3902,
+      "step": 595
+    },
+    {
+      "epoch": 0.35318518518518516,
+      "grad_norm": 2.8834545612335205,
+      "learning_rate": 1.7652637818612924e-05,
+      "loss": 0.4845,
+      "step": 596
+    },
+    {
+      "epoch": 0.3537777777777778,
+      "grad_norm": 2.017702102661133,
+      "learning_rate": 1.764868603042877e-05,
+      "loss": 0.4704,
+      "step": 597
+    },
+    {
+      "epoch": 0.3543703703703704,
+      "grad_norm": 1.7782742977142334,
+      "learning_rate": 1.7644734242244616e-05,
+      "loss": 0.4374,
+      "step": 598
+    },
+    {
+      "epoch": 0.354962962962963,
+      "grad_norm": 2.168747663497925,
+      "learning_rate": 1.7640782454060464e-05,
+      "loss": 0.5434,
+      "step": 599
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 1.9563325643539429,
+      "learning_rate": 1.763683066587631e-05,
+      "loss": 0.5634,
+      "step": 600
+    },
+    {
+      "epoch": 0.35614814814814816,
+      "grad_norm": 2.0246036052703857,
+      "learning_rate": 1.7632878877692157e-05,
+      "loss": 0.3731,
+      "step": 601
+    },
+    {
+      "epoch": 0.35674074074074075,
+      "grad_norm": 2.7511146068573,
+      "learning_rate": 1.7628927089508005e-05,
+      "loss": 0.5918,
+      "step": 602
+    },
+    {
+      "epoch": 0.35733333333333334,
+      "grad_norm": 2.3714897632598877,
+      "learning_rate": 1.762497530132385e-05,
+      "loss": 0.5415,
+      "step": 603
+    },
+    {
+      "epoch": 0.3579259259259259,
+      "grad_norm": 2.091686487197876,
+      "learning_rate": 1.7621023513139697e-05,
+      "loss": 0.4914,
+      "step": 604
+    },
+    {
+      "epoch": 0.3585185185185185,
+      "grad_norm": 1.372531533241272,
+      "learning_rate": 1.7617071724955542e-05,
+      "loss": 0.4495,
+      "step": 605
+    },
+    {
+      "epoch": 0.3591111111111111,
+      "grad_norm": 2.6099765300750732,
+      "learning_rate": 1.761311993677139e-05,
+      "loss": 0.5671,
+      "step": 606
+    },
+    {
+      "epoch": 0.3597037037037037,
+      "grad_norm": 2.2447385787963867,
+      "learning_rate": 1.7609168148587238e-05,
+      "loss": 0.3964,
+      "step": 607
+    },
+    {
+      "epoch": 0.3602962962962963,
+      "grad_norm": 2.5114998817443848,
+      "learning_rate": 1.7605216360403082e-05,
+      "loss": 0.4225,
+      "step": 608
+    },
+    {
+      "epoch": 0.36088888888888887,
+      "grad_norm": 1.560344934463501,
+      "learning_rate": 1.760126457221893e-05,
+      "loss": 0.3483,
+      "step": 609
+    },
+    {
+      "epoch": 0.36148148148148146,
+      "grad_norm": 1.716403603553772,
+      "learning_rate": 1.7597312784034778e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.36207407407407405,
+      "grad_norm": 3.6560916900634766,
+      "learning_rate": 1.7593360995850623e-05,
+      "loss": 0.7983,
+      "step": 611
+    },
+    {
+      "epoch": 0.3626666666666667,
+      "grad_norm": 2.56884765625,
+      "learning_rate": 1.758940920766647e-05,
+      "loss": 0.492,
+      "step": 612
+    },
+    {
+      "epoch": 0.3632592592592593,
+      "grad_norm": 4.894947052001953,
+      "learning_rate": 1.7585457419482315e-05,
+      "loss": 0.6476,
+      "step": 613
+    },
+    {
+      "epoch": 0.36385185185185187,
+      "grad_norm": 1.6462690830230713,
+      "learning_rate": 1.7581505631298163e-05,
+      "loss": 0.4268,
+      "step": 614
+    },
+    {
+      "epoch": 0.36444444444444446,
+      "grad_norm": 1.755002737045288,
+      "learning_rate": 1.757755384311401e-05,
+      "loss": 0.4682,
+      "step": 615
+    },
+    {
+      "epoch": 0.36503703703703705,
+      "grad_norm": 1.9155497550964355,
+      "learning_rate": 1.7573602054929856e-05,
+      "loss": 0.4528,
+      "step": 616
+    },
+    {
+      "epoch": 0.36562962962962964,
+      "grad_norm": 2.0476064682006836,
+      "learning_rate": 1.7569650266745704e-05,
+      "loss": 0.3593,
+      "step": 617
+    },
+    {
+      "epoch": 0.3662222222222222,
+      "grad_norm": 2.4431557655334473,
+      "learning_rate": 1.756569847856155e-05,
+      "loss": 0.4587,
+      "step": 618
+    },
+    {
+      "epoch": 0.3668148148148148,
+      "grad_norm": 2.559109687805176,
+      "learning_rate": 1.7561746690377396e-05,
+      "loss": 0.3825,
+      "step": 619
+    },
+    {
+      "epoch": 0.3674074074074074,
+      "grad_norm": 2.3843023777008057,
+      "learning_rate": 1.7557794902193244e-05,
+      "loss": 0.4671,
+      "step": 620
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 2.7746808528900146,
+      "learning_rate": 1.7553843114009092e-05,
+      "loss": 0.537,
+      "step": 621
+    },
+    {
+      "epoch": 0.3685925925925926,
+      "grad_norm": 2.0530636310577393,
+      "learning_rate": 1.7549891325824936e-05,
+      "loss": 0.5109,
+      "step": 622
+    },
+    {
+      "epoch": 0.36918518518518517,
+      "grad_norm": 2.9484496116638184,
+      "learning_rate": 1.7545939537640784e-05,
+      "loss": 0.5099,
+      "step": 623
+    },
+    {
+      "epoch": 0.36977777777777776,
+      "grad_norm": 2.8568387031555176,
+      "learning_rate": 1.754198774945663e-05,
+      "loss": 0.4449,
+      "step": 624
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 2.3108699321746826,
+      "learning_rate": 1.7538035961272477e-05,
+      "loss": 0.4965,
+      "step": 625
+    },
+    {
+      "epoch": 0.37096296296296294,
+      "grad_norm": 2.982693910598755,
+      "learning_rate": 1.7534084173088325e-05,
+      "loss": 0.4668,
+      "step": 626
+    },
+    {
+      "epoch": 0.37155555555555553,
+      "grad_norm": 4.840386867523193,
+      "learning_rate": 1.753013238490417e-05,
+      "loss": 0.4947,
+      "step": 627
+    },
+    {
+      "epoch": 0.3721481481481482,
+      "grad_norm": 4.448367118835449,
+      "learning_rate": 1.7526180596720017e-05,
+      "loss": 0.4846,
+      "step": 628
+    },
+    {
+      "epoch": 0.37274074074074076,
+      "grad_norm": 2.632222890853882,
+      "learning_rate": 1.7522228808535865e-05,
+      "loss": 0.493,
+      "step": 629
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 2.5487730503082275,
+      "learning_rate": 1.751827702035171e-05,
+      "loss": 0.3982,
+      "step": 630
+    },
+    {
+      "epoch": 0.37392592592592594,
+      "grad_norm": 1.3909459114074707,
+      "learning_rate": 1.7514325232167558e-05,
+      "loss": 0.371,
+      "step": 631
+    },
+    {
+      "epoch": 0.37451851851851853,
+      "grad_norm": 3.138512134552002,
+      "learning_rate": 1.7510373443983402e-05,
+      "loss": 0.4142,
+      "step": 632
+    },
+    {
+      "epoch": 0.3751111111111111,
+      "grad_norm": 3.331690549850464,
+      "learning_rate": 1.750642165579925e-05,
+      "loss": 0.4561,
+      "step": 633
+    },
+    {
+      "epoch": 0.3757037037037037,
+      "grad_norm": 4.277185440063477,
+      "learning_rate": 1.7502469867615098e-05,
+      "loss": 0.5056,
+      "step": 634
+    },
+    {
+      "epoch": 0.3762962962962963,
+      "grad_norm": 1.9603720903396606,
+      "learning_rate": 1.7498518079430943e-05,
+      "loss": 0.318,
+      "step": 635
+    },
+    {
+      "epoch": 0.3768888888888889,
+      "grad_norm": 2.760451555252075,
+      "learning_rate": 1.749456629124679e-05,
+      "loss": 0.5165,
+      "step": 636
+    },
+    {
+      "epoch": 0.3774814814814815,
+      "grad_norm": 1.588133454322815,
+      "learning_rate": 1.749061450306264e-05,
+      "loss": 0.4165,
+      "step": 637
+    },
+    {
+      "epoch": 0.37807407407407406,
+      "grad_norm": 4.102574825286865,
+      "learning_rate": 1.7486662714878483e-05,
+      "loss": 0.5701,
+      "step": 638
+    },
+    {
+      "epoch": 0.37866666666666665,
+      "grad_norm": 2.8866281509399414,
+      "learning_rate": 1.748271092669433e-05,
+      "loss": 0.3949,
+      "step": 639
+    },
+    {
+      "epoch": 0.37925925925925924,
+      "grad_norm": 2.061077356338501,
+      "learning_rate": 1.747875913851018e-05,
+      "loss": 0.3985,
+      "step": 640
+    },
+    {
+      "epoch": 0.37985185185185183,
+      "grad_norm": 2.801758289337158,
+      "learning_rate": 1.7474807350326024e-05,
+      "loss": 0.5972,
+      "step": 641
+    },
+    {
+      "epoch": 0.3804444444444444,
+      "grad_norm": 2.083134174346924,
+      "learning_rate": 1.747085556214187e-05,
+      "loss": 0.4221,
+      "step": 642
+    },
+    {
+      "epoch": 0.38103703703703706,
+      "grad_norm": 2.5239572525024414,
+      "learning_rate": 1.7466903773957716e-05,
+      "loss": 0.5471,
+      "step": 643
+    },
+    {
+      "epoch": 0.38162962962962965,
+      "grad_norm": 4.911279678344727,
+      "learning_rate": 1.7462951985773564e-05,
+      "loss": 0.6518,
+      "step": 644
+    },
+    {
+      "epoch": 0.38222222222222224,
+      "grad_norm": 2.1469693183898926,
+      "learning_rate": 1.7459000197589412e-05,
+      "loss": 0.4383,
+      "step": 645
+    },
+    {
+      "epoch": 0.38281481481481483,
+      "grad_norm": 1.703818917274475,
+      "learning_rate": 1.7455048409405256e-05,
+      "loss": 0.3851,
+      "step": 646
+    },
+    {
+      "epoch": 0.3834074074074074,
+      "grad_norm": 2.8534135818481445,
+      "learning_rate": 1.7451096621221104e-05,
+      "loss": 0.4981,
+      "step": 647
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 1.8709168434143066,
+      "learning_rate": 1.7447144833036952e-05,
+      "loss": 0.3979,
+      "step": 648
+    },
+    {
+      "epoch": 0.3845925925925926,
+      "grad_norm": 2.8124961853027344,
+      "learning_rate": 1.7443193044852797e-05,
+      "loss": 0.3693,
+      "step": 649
+    },
+    {
+      "epoch": 0.3851851851851852,
+      "grad_norm": 1.993384838104248,
+      "learning_rate": 1.7439241256668645e-05,
+      "loss": 0.4073,
+      "step": 650
+    },
+    {
+      "epoch": 0.3857777777777778,
+      "grad_norm": 1.8268855810165405,
+      "learning_rate": 1.743528946848449e-05,
+      "loss": 0.4085,
+      "step": 651
+    },
+    {
+      "epoch": 0.38637037037037036,
+      "grad_norm": 2.0153682231903076,
+      "learning_rate": 1.7431337680300337e-05,
+      "loss": 0.3893,
+      "step": 652
+    },
+    {
+      "epoch": 0.38696296296296295,
+      "grad_norm": 1.8038874864578247,
+      "learning_rate": 1.7427385892116185e-05,
+      "loss": 0.3848,
+      "step": 653
+    },
+    {
+      "epoch": 0.38755555555555554,
+      "grad_norm": 2.423920154571533,
+      "learning_rate": 1.742343410393203e-05,
+      "loss": 0.4203,
+      "step": 654
+    },
+    {
+      "epoch": 0.38814814814814813,
+      "grad_norm": 2.744053840637207,
+      "learning_rate": 1.7419482315747878e-05,
+      "loss": 0.4921,
+      "step": 655
+    },
+    {
+      "epoch": 0.3887407407407407,
+      "grad_norm": 1.7281179428100586,
+      "learning_rate": 1.7415530527563726e-05,
+      "loss": 0.4987,
+      "step": 656
+    },
+    {
+      "epoch": 0.3893333333333333,
+      "grad_norm": 1.5895280838012695,
+      "learning_rate": 1.741157873937957e-05,
+      "loss": 0.3695,
+      "step": 657
+    },
+    {
+      "epoch": 0.38992592592592595,
+      "grad_norm": 1.7054800987243652,
+      "learning_rate": 1.7407626951195418e-05,
+      "loss": 0.3699,
+      "step": 658
+    },
+    {
+      "epoch": 0.39051851851851854,
+      "grad_norm": 2.1619656085968018,
+      "learning_rate": 1.7403675163011263e-05,
+      "loss": 0.4175,
+      "step": 659
+    },
+    {
+      "epoch": 0.39111111111111113,
+      "grad_norm": 2.146920680999756,
+      "learning_rate": 1.739972337482711e-05,
+      "loss": 0.4306,
+      "step": 660
+    },
+    {
+      "epoch": 0.3917037037037037,
+      "grad_norm": 3.42474365234375,
+      "learning_rate": 1.739577158664296e-05,
+      "loss": 0.5367,
+      "step": 661
+    },
+    {
+      "epoch": 0.3922962962962963,
+      "grad_norm": 3.4475483894348145,
+      "learning_rate": 1.7391819798458803e-05,
+      "loss": 0.5108,
+      "step": 662
+    },
+    {
+      "epoch": 0.3928888888888889,
+      "grad_norm": 2.5728862285614014,
+      "learning_rate": 1.738786801027465e-05,
+      "loss": 0.4898,
+      "step": 663
+    },
+    {
+      "epoch": 0.3934814814814815,
+      "grad_norm": 2.306109666824341,
+      "learning_rate": 1.73839162220905e-05,
+      "loss": 0.4789,
+      "step": 664
+    },
+    {
+      "epoch": 0.3940740740740741,
+      "grad_norm": 2.1014938354492188,
+      "learning_rate": 1.7379964433906344e-05,
+      "loss": 0.5479,
+      "step": 665
+    },
+    {
+      "epoch": 0.39466666666666667,
+      "grad_norm": 4.2000203132629395,
+      "learning_rate": 1.737601264572219e-05,
+      "loss": 0.5086,
+      "step": 666
+    },
+    {
+      "epoch": 0.39525925925925925,
+      "grad_norm": 3.8940694332122803,
+      "learning_rate": 1.737206085753804e-05,
+      "loss": 0.5285,
+      "step": 667
+    },
+    {
+      "epoch": 0.39585185185185184,
+      "grad_norm": 5.133601665496826,
+      "learning_rate": 1.7368109069353884e-05,
+      "loss": 0.6168,
+      "step": 668
+    },
+    {
+      "epoch": 0.39644444444444443,
+      "grad_norm": 2.3495290279388428,
+      "learning_rate": 1.7364157281169732e-05,
+      "loss": 0.4344,
+      "step": 669
+    },
+    {
+      "epoch": 0.397037037037037,
+      "grad_norm": 1.6571722030639648,
+      "learning_rate": 1.7360205492985576e-05,
+      "loss": 0.4727,
+      "step": 670
+    },
+    {
+      "epoch": 0.3976296296296296,
+      "grad_norm": 1.6108791828155518,
+      "learning_rate": 1.7356253704801424e-05,
+      "loss": 0.439,
+      "step": 671
+    },
+    {
+      "epoch": 0.3982222222222222,
+      "grad_norm": 5.217345237731934,
+      "learning_rate": 1.7352301916617272e-05,
+      "loss": 0.6918,
+      "step": 672
+    },
+    {
+      "epoch": 0.3988148148148148,
+      "grad_norm": 2.1959116458892822,
+      "learning_rate": 1.7348350128433117e-05,
+      "loss": 0.4956,
+      "step": 673
+    },
+    {
+      "epoch": 0.39940740740740743,
+      "grad_norm": 2.198957920074463,
+      "learning_rate": 1.7344398340248965e-05,
+      "loss": 0.3916,
+      "step": 674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.944075345993042,
+      "learning_rate": 1.7340446552064813e-05,
+      "loss": 0.5828,
+      "step": 675
+    },
+    {
+      "epoch": 0.4005925925925926,
+      "grad_norm": 3.1505320072174072,
+      "learning_rate": 1.7336494763880657e-05,
+      "loss": 0.3915,
+      "step": 676
+    },
+    {
+      "epoch": 0.4011851851851852,
+      "grad_norm": 2.9767041206359863,
+      "learning_rate": 1.7332542975696505e-05,
+      "loss": 0.4784,
+      "step": 677
+    },
+    {
+      "epoch": 0.4017777777777778,
+      "grad_norm": 2.7852866649627686,
+      "learning_rate": 1.732859118751235e-05,
+      "loss": 0.3328,
+      "step": 678
+    },
+    {
+      "epoch": 0.4023703703703704,
+      "grad_norm": 2.1984496116638184,
+      "learning_rate": 1.7324639399328198e-05,
+      "loss": 0.4984,
+      "step": 679
+    },
+    {
+      "epoch": 0.40296296296296297,
+      "grad_norm": 1.5234653949737549,
+      "learning_rate": 1.7320687611144046e-05,
+      "loss": 0.4553,
+      "step": 680
+    },
+    {
+      "epoch": 0.40355555555555556,
+      "grad_norm": 2.4370908737182617,
+      "learning_rate": 1.731673582295989e-05,
+      "loss": 0.4678,
+      "step": 681
+    },
+    {
+      "epoch": 0.40414814814814815,
+      "grad_norm": 1.6188708543777466,
+      "learning_rate": 1.7312784034775738e-05,
+      "loss": 0.3489,
+      "step": 682
+    },
+    {
+      "epoch": 0.40474074074074073,
+      "grad_norm": 1.5459580421447754,
+      "learning_rate": 1.7308832246591586e-05,
+      "loss": 0.3994,
+      "step": 683
+    },
+    {
+      "epoch": 0.4053333333333333,
+      "grad_norm": 4.605407238006592,
+      "learning_rate": 1.730488045840743e-05,
+      "loss": 0.7398,
+      "step": 684
+    },
+    {
+      "epoch": 0.4059259259259259,
+      "grad_norm": 2.255581855773926,
+      "learning_rate": 1.730092867022328e-05,
+      "loss": 0.4301,
+      "step": 685
+    },
+    {
+      "epoch": 0.4065185185185185,
+      "grad_norm": 1.5918084383010864,
+      "learning_rate": 1.7296976882039127e-05,
+      "loss": 0.3467,
+      "step": 686
+    },
+    {
+      "epoch": 0.4071111111111111,
+      "grad_norm": 2.140010356903076,
+      "learning_rate": 1.729302509385497e-05,
+      "loss": 0.3636,
+      "step": 687
+    },
+    {
+      "epoch": 0.4077037037037037,
+      "grad_norm": 2.485490083694458,
+      "learning_rate": 1.728907330567082e-05,
+      "loss": 0.4062,
+      "step": 688
+    },
+    {
+      "epoch": 0.4082962962962963,
+      "grad_norm": 2.160299777984619,
+      "learning_rate": 1.7285121517486664e-05,
+      "loss": 0.5892,
+      "step": 689
+    },
+    {
+      "epoch": 0.4088888888888889,
+      "grad_norm": 2.920334577560425,
+      "learning_rate": 1.728116972930251e-05,
+      "loss": 0.4564,
+      "step": 690
+    },
+    {
+      "epoch": 0.4094814814814815,
+      "grad_norm": 3.1677141189575195,
+      "learning_rate": 1.727721794111836e-05,
+      "loss": 0.503,
+      "step": 691
+    },
+    {
+      "epoch": 0.4100740740740741,
+      "grad_norm": 2.0015292167663574,
+      "learning_rate": 1.7273266152934204e-05,
+      "loss": 0.3727,
+      "step": 692
+    },
+    {
+      "epoch": 0.4106666666666667,
+      "grad_norm": 1.5431313514709473,
+      "learning_rate": 1.7269314364750052e-05,
+      "loss": 0.3897,
+      "step": 693
+    },
+    {
+      "epoch": 0.41125925925925927,
+      "grad_norm": 1.559937596321106,
+      "learning_rate": 1.72653625765659e-05,
+      "loss": 0.4198,
+      "step": 694
+    },
+    {
+      "epoch": 0.41185185185185186,
+      "grad_norm": 2.3283064365386963,
+      "learning_rate": 1.7261410788381744e-05,
+      "loss": 0.4374,
+      "step": 695
+    },
+    {
+      "epoch": 0.41244444444444445,
+      "grad_norm": 1.6182314157485962,
+      "learning_rate": 1.7257459000197592e-05,
+      "loss": 0.4345,
+      "step": 696
+    },
+    {
+      "epoch": 0.41303703703703704,
+      "grad_norm": 2.362454414367676,
+      "learning_rate": 1.7253507212013437e-05,
+      "loss": 0.4878,
+      "step": 697
+    },
+    {
+      "epoch": 0.4136296296296296,
+      "grad_norm": 1.7828232049942017,
+      "learning_rate": 1.7249555423829285e-05,
+      "loss": 0.3607,
+      "step": 698
+    },
+    {
+      "epoch": 0.4142222222222222,
+      "grad_norm": 2.4924163818359375,
+      "learning_rate": 1.7245603635645133e-05,
+      "loss": 0.4836,
+      "step": 699
+    },
+    {
+      "epoch": 0.4148148148148148,
+      "grad_norm": 3.8239803314208984,
+      "learning_rate": 1.7241651847460977e-05,
+      "loss": 0.5245,
+      "step": 700
+    },
+    {
+      "epoch": 0.4154074074074074,
+      "grad_norm": 4.042765140533447,
+      "learning_rate": 1.7237700059276825e-05,
+      "loss": 0.4798,
+      "step": 701
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 4.468754768371582,
+      "learning_rate": 1.7233748271092673e-05,
+      "loss": 0.6322,
+      "step": 702
+    },
+    {
+      "epoch": 0.41659259259259257,
+      "grad_norm": 2.8157455921173096,
+      "learning_rate": 1.7229796482908518e-05,
+      "loss": 0.3272,
+      "step": 703
+    },
+    {
+      "epoch": 0.41718518518518516,
+      "grad_norm": 7.073381423950195,
+      "learning_rate": 1.7225844694724366e-05,
+      "loss": 0.6554,
+      "step": 704
+    },
+    {
+      "epoch": 0.4177777777777778,
+      "grad_norm": 1.7138015031814575,
+      "learning_rate": 1.722189290654021e-05,
+      "loss": 0.4317,
+      "step": 705
+    },
+    {
+      "epoch": 0.4183703703703704,
+      "grad_norm": 3.400871515274048,
+      "learning_rate": 1.7217941118356058e-05,
+      "loss": 0.407,
+      "step": 706
+    },
+    {
+      "epoch": 0.418962962962963,
+      "grad_norm": 2.9848408699035645,
+      "learning_rate": 1.7213989330171906e-05,
+      "loss": 0.3922,
+      "step": 707
+    },
+    {
+      "epoch": 0.41955555555555557,
+      "grad_norm": 3.3588242530822754,
+      "learning_rate": 1.721003754198775e-05,
+      "loss": 0.5256,
+      "step": 708
+    },
+    {
+      "epoch": 0.42014814814814816,
+      "grad_norm": 1.6815155744552612,
+      "learning_rate": 1.72060857538036e-05,
+      "loss": 0.4466,
+      "step": 709
+    },
+    {
+      "epoch": 0.42074074074074075,
+      "grad_norm": 1.8659682273864746,
+      "learning_rate": 1.7202133965619447e-05,
+      "loss": 0.3424,
+      "step": 710
+    },
+    {
+      "epoch": 0.42133333333333334,
+      "grad_norm": 2.181671142578125,
+      "learning_rate": 1.719818217743529e-05,
+      "loss": 0.3343,
+      "step": 711
+    },
+    {
+      "epoch": 0.4219259259259259,
+      "grad_norm": 2.5460152626037598,
+      "learning_rate": 1.719423038925114e-05,
+      "loss": 0.3688,
+      "step": 712
+    },
+    {
+      "epoch": 0.4225185185185185,
+      "grad_norm": 1.8101012706756592,
+      "learning_rate": 1.7190278601066987e-05,
+      "loss": 0.3309,
+      "step": 713
+    },
+    {
+      "epoch": 0.4231111111111111,
+      "grad_norm": 3.7047224044799805,
+      "learning_rate": 1.718632681288283e-05,
+      "loss": 0.5216,
+      "step": 714
+    },
+    {
+      "epoch": 0.4237037037037037,
+      "grad_norm": 1.2390779256820679,
+      "learning_rate": 1.7182375024698676e-05,
+      "loss": 0.3197,
+      "step": 715
+    },
+    {
+      "epoch": 0.4242962962962963,
+      "grad_norm": 1.883718729019165,
+      "learning_rate": 1.7178423236514524e-05,
+      "loss": 0.4439,
+      "step": 716
+    },
+    {
+      "epoch": 0.42488888888888887,
+      "grad_norm": 2.6261446475982666,
+      "learning_rate": 1.717447144833037e-05,
+      "loss": 0.3773,
+      "step": 717
+    },
+    {
+      "epoch": 0.42548148148148146,
+      "grad_norm": 3.37211537361145,
+      "learning_rate": 1.7170519660146216e-05,
+      "loss": 0.4223,
+      "step": 718
+    },
+    {
+      "epoch": 0.42607407407407405,
+      "grad_norm": 3.1350882053375244,
+      "learning_rate": 1.7166567871962064e-05,
+      "loss": 0.5476,
+      "step": 719
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 2.0384840965270996,
+      "learning_rate": 1.716261608377791e-05,
+      "loss": 0.5152,
+      "step": 720
+    },
+    {
+      "epoch": 0.4272592592592593,
+      "grad_norm": 2.0325429439544678,
+      "learning_rate": 1.7158664295593757e-05,
+      "loss": 0.3711,
+      "step": 721
+    },
+    {
+      "epoch": 0.42785185185185187,
+      "grad_norm": 2.5946013927459717,
+      "learning_rate": 1.7154712507409605e-05,
+      "loss": 0.5286,
+      "step": 722
+    },
+    {
+      "epoch": 0.42844444444444446,
+      "grad_norm": 1.7490061521530151,
+      "learning_rate": 1.715076071922545e-05,
+      "loss": 0.3495,
+      "step": 723
+    },
+    {
+      "epoch": 0.42903703703703705,
+      "grad_norm": 2.927640438079834,
+      "learning_rate": 1.7146808931041297e-05,
+      "loss": 0.5296,
+      "step": 724
+    },
+    {
+      "epoch": 0.42962962962962964,
+      "grad_norm": 3.1339566707611084,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.3342,
+      "step": 725
+    },
+    {
+      "epoch": 0.43022222222222223,
+      "grad_norm": 3.994549036026001,
+      "learning_rate": 1.713890535467299e-05,
+      "loss": 0.7666,
+      "step": 726
+    },
+    {
+      "epoch": 0.4308148148148148,
+      "grad_norm": 1.9822049140930176,
+      "learning_rate": 1.7134953566488838e-05,
+      "loss": 0.3624,
+      "step": 727
+    },
+    {
+      "epoch": 0.4314074074074074,
+      "grad_norm": 2.5283572673797607,
+      "learning_rate": 1.7131001778304682e-05,
+      "loss": 0.4574,
+      "step": 728
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 1.8353033065795898,
+      "learning_rate": 1.712704999012053e-05,
+      "loss": 0.4293,
+      "step": 729
+    },
+    {
+      "epoch": 0.4325925925925926,
+      "grad_norm": 1.8253928422927856,
+      "learning_rate": 1.7123098201936378e-05,
+      "loss": 0.3923,
+      "step": 730
+    },
+    {
+      "epoch": 0.4331851851851852,
+      "grad_norm": 1.4956635236740112,
+      "learning_rate": 1.7119146413752223e-05,
+      "loss": 0.4004,
+      "step": 731
+    },
+    {
+      "epoch": 0.43377777777777776,
+      "grad_norm": 1.9412473440170288,
+      "learning_rate": 1.711519462556807e-05,
+      "loss": 0.328,
+      "step": 732
+    },
+    {
+      "epoch": 0.43437037037037035,
+      "grad_norm": 2.7054789066314697,
+      "learning_rate": 1.7111242837383915e-05,
+      "loss": 0.4876,
+      "step": 733
+    },
+    {
+      "epoch": 0.43496296296296294,
+      "grad_norm": 2.321749210357666,
+      "learning_rate": 1.7107291049199763e-05,
+      "loss": 0.5035,
+      "step": 734
+    },
+    {
+      "epoch": 0.43555555555555553,
+      "grad_norm": 1.9178348779678345,
+      "learning_rate": 1.710333926101561e-05,
+      "loss": 0.4513,
+      "step": 735
+    },
+    {
+      "epoch": 0.4361481481481482,
+      "grad_norm": 2.0546038150787354,
+      "learning_rate": 1.7099387472831456e-05,
+      "loss": 0.4995,
+      "step": 736
+    },
+    {
+      "epoch": 0.43674074074074076,
+      "grad_norm": 1.5794297456741333,
+      "learning_rate": 1.7095435684647304e-05,
+      "loss": 0.3643,
+      "step": 737
+    },
+    {
+      "epoch": 0.43733333333333335,
+      "grad_norm": 1.919447660446167,
+      "learning_rate": 1.709148389646315e-05,
+      "loss": 0.3417,
+      "step": 738
+    },
+    {
+      "epoch": 0.43792592592592594,
+      "grad_norm": 1.440281629562378,
+      "learning_rate": 1.7087532108278996e-05,
+      "loss": 0.3751,
+      "step": 739
+    },
+    {
+      "epoch": 0.43851851851851853,
+      "grad_norm": 1.5072587728500366,
+      "learning_rate": 1.7083580320094844e-05,
+      "loss": 0.3882,
+      "step": 740
+    },
+    {
+      "epoch": 0.4391111111111111,
+      "grad_norm": 3.3286335468292236,
+      "learning_rate": 1.7079628531910692e-05,
+      "loss": 0.4404,
+      "step": 741
+    },
+    {
+      "epoch": 0.4397037037037037,
+      "grad_norm": 1.6914730072021484,
+      "learning_rate": 1.7075676743726536e-05,
+      "loss": 0.384,
+      "step": 742
+    },
+    {
+      "epoch": 0.4402962962962963,
+      "grad_norm": 2.079453229904175,
+      "learning_rate": 1.7071724955542384e-05,
+      "loss": 0.404,
+      "step": 743
+    },
+    {
+      "epoch": 0.4408888888888889,
+      "grad_norm": 1.5200729370117188,
+      "learning_rate": 1.706777316735823e-05,
+      "loss": 0.3424,
+      "step": 744
+    },
+    {
+      "epoch": 0.4414814814814815,
+      "grad_norm": 2.0190014839172363,
+      "learning_rate": 1.7063821379174077e-05,
+      "loss": 0.4098,
+      "step": 745
+    },
+    {
+      "epoch": 0.44207407407407406,
+      "grad_norm": 1.6551612615585327,
+      "learning_rate": 1.7059869590989925e-05,
+      "loss": 0.4428,
+      "step": 746
+    },
+    {
+      "epoch": 0.44266666666666665,
+      "grad_norm": 2.087196111679077,
+      "learning_rate": 1.705591780280577e-05,
+      "loss": 0.4525,
+      "step": 747
+    },
+    {
+      "epoch": 0.44325925925925924,
+      "grad_norm": 2.1805005073547363,
+      "learning_rate": 1.7051966014621617e-05,
+      "loss": 0.4749,
+      "step": 748
+    },
+    {
+      "epoch": 0.44385185185185183,
+      "grad_norm": 2.3415818214416504,
+      "learning_rate": 1.7048014226437465e-05,
+      "loss": 0.5174,
+      "step": 749
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.7995414733886719,
+      "learning_rate": 1.704406243825331e-05,
+      "loss": 0.3263,
+      "step": 750
+    },
+    {
+      "epoch": 0.44503703703703706,
+      "grad_norm": 2.973446846008301,
+      "learning_rate": 1.7040110650069158e-05,
+      "loss": 0.5417,
+      "step": 751
+    },
+    {
+      "epoch": 0.44562962962962965,
+      "grad_norm": 2.4484903812408447,
+      "learning_rate": 1.7036158861885002e-05,
+      "loss": 0.3912,
+      "step": 752
+    },
+    {
+      "epoch": 0.44622222222222224,
+      "grad_norm": 2.092816114425659,
+      "learning_rate": 1.703220707370085e-05,
+      "loss": 0.3285,
+      "step": 753
+    },
+    {
+      "epoch": 0.44681481481481483,
+      "grad_norm": 1.7917747497558594,
+      "learning_rate": 1.7028255285516698e-05,
+      "loss": 0.3726,
+      "step": 754
+    },
+    {
+      "epoch": 0.4474074074074074,
+      "grad_norm": 1.6348568201065063,
+      "learning_rate": 1.7024303497332543e-05,
+      "loss": 0.5264,
+      "step": 755
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.9282933473587036,
+      "learning_rate": 1.702035170914839e-05,
+      "loss": 0.3641,
+      "step": 756
+    },
+    {
+      "epoch": 0.4485925925925926,
+      "grad_norm": 3.362813949584961,
+      "learning_rate": 1.701639992096424e-05,
+      "loss": 0.5195,
+      "step": 757
+    },
+    {
+      "epoch": 0.4491851851851852,
+      "grad_norm": 3.403738021850586,
+      "learning_rate": 1.7012448132780083e-05,
+      "loss": 0.6189,
+      "step": 758
+    },
+    {
+      "epoch": 0.4497777777777778,
+      "grad_norm": 1.7579808235168457,
+      "learning_rate": 1.700849634459593e-05,
+      "loss": 0.439,
+      "step": 759
+    },
+    {
+      "epoch": 0.45037037037037037,
+      "grad_norm": 2.405736207962036,
+      "learning_rate": 1.7004544556411776e-05,
+      "loss": 0.4584,
+      "step": 760
+    },
+    {
+      "epoch": 0.45096296296296295,
+      "grad_norm": 2.802668333053589,
+      "learning_rate": 1.7000592768227624e-05,
+      "loss": 0.5502,
+      "step": 761
+    },
+    {
+      "epoch": 0.45155555555555554,
+      "grad_norm": 2.443613290786743,
+      "learning_rate": 1.699664098004347e-05,
+      "loss": 0.4755,
+      "step": 762
+    },
+    {
+      "epoch": 0.45214814814814813,
+      "grad_norm": 2.0970232486724854,
+      "learning_rate": 1.6992689191859316e-05,
+      "loss": 0.3699,
+      "step": 763
+    },
+    {
+      "epoch": 0.4527407407407407,
+      "grad_norm": 2.144465923309326,
+      "learning_rate": 1.6988737403675164e-05,
+      "loss": 0.4057,
+      "step": 764
+    },
+    {
+      "epoch": 0.4533333333333333,
+      "grad_norm": 3.0747861862182617,
+      "learning_rate": 1.6984785615491012e-05,
+      "loss": 0.436,
+      "step": 765
+    },
+    {
+      "epoch": 0.4539259259259259,
+      "grad_norm": 2.672030448913574,
+      "learning_rate": 1.6980833827306856e-05,
+      "loss": 0.4222,
+      "step": 766
+    },
+    {
+      "epoch": 0.45451851851851854,
+      "grad_norm": 2.3740336894989014,
+      "learning_rate": 1.6976882039122704e-05,
+      "loss": 0.5197,
+      "step": 767
+    },
+    {
+      "epoch": 0.45511111111111113,
+      "grad_norm": 3.370637893676758,
+      "learning_rate": 1.6972930250938552e-05,
+      "loss": 0.5817,
+      "step": 768
+    },
+    {
+      "epoch": 0.4557037037037037,
+      "grad_norm": 1.4780560731887817,
+      "learning_rate": 1.6968978462754397e-05,
+      "loss": 0.4291,
+      "step": 769
+    },
+    {
+      "epoch": 0.4562962962962963,
+      "grad_norm": 1.5800895690917969,
+      "learning_rate": 1.6965026674570245e-05,
+      "loss": 0.4808,
+      "step": 770
+    },
+    {
+      "epoch": 0.4568888888888889,
+      "grad_norm": 1.7177149057388306,
+      "learning_rate": 1.696107488638609e-05,
+      "loss": 0.3755,
+      "step": 771
+    },
+    {
+      "epoch": 0.4574814814814815,
+      "grad_norm": 1.9221620559692383,
+      "learning_rate": 1.6957123098201937e-05,
+      "loss": 0.5025,
+      "step": 772
+    },
+    {
+      "epoch": 0.4580740740740741,
+      "grad_norm": 3.466310501098633,
+      "learning_rate": 1.6953171310017785e-05,
+      "loss": 0.4771,
+      "step": 773
+    },
+    {
+      "epoch": 0.45866666666666667,
+      "grad_norm": 2.286259412765503,
+      "learning_rate": 1.694921952183363e-05,
+      "loss": 0.3152,
+      "step": 774
+    },
+    {
+      "epoch": 0.45925925925925926,
+      "grad_norm": 2.420823335647583,
+      "learning_rate": 1.6945267733649478e-05,
+      "loss": 0.3499,
+      "step": 775
+    },
+    {
+      "epoch": 0.45985185185185184,
+      "grad_norm": 1.7428709268569946,
+      "learning_rate": 1.6941315945465326e-05,
+      "loss": 0.44,
+      "step": 776
+    },
+    {
+      "epoch": 0.46044444444444443,
+      "grad_norm": 2.029711961746216,
+      "learning_rate": 1.693736415728117e-05,
+      "loss": 0.4027,
+      "step": 777
+    },
+    {
+      "epoch": 0.461037037037037,
+      "grad_norm": 2.3087120056152344,
+      "learning_rate": 1.6933412369097018e-05,
+      "loss": 0.418,
+      "step": 778
+    },
+    {
+      "epoch": 0.4616296296296296,
+      "grad_norm": 2.248068332672119,
+      "learning_rate": 1.6929460580912863e-05,
+      "loss": 0.4229,
+      "step": 779
+    },
+    {
+      "epoch": 0.4622222222222222,
+      "grad_norm": 1.937864899635315,
+      "learning_rate": 1.692550879272871e-05,
+      "loss": 0.4662,
+      "step": 780
+    },
+    {
+      "epoch": 0.4628148148148148,
+      "grad_norm": 3.1207239627838135,
+      "learning_rate": 1.692155700454456e-05,
+      "loss": 0.515,
+      "step": 781
+    },
+    {
+      "epoch": 0.46340740740740743,
+      "grad_norm": 2.6201674938201904,
+      "learning_rate": 1.6917605216360403e-05,
+      "loss": 0.4325,
+      "step": 782
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 2.7524359226226807,
+      "learning_rate": 1.691365342817625e-05,
+      "loss": 0.3577,
+      "step": 783
+    },
+    {
+      "epoch": 0.4645925925925926,
+      "grad_norm": 2.435903787612915,
+      "learning_rate": 1.69097016399921e-05,
+      "loss": 0.528,
+      "step": 784
+    },
+    {
+      "epoch": 0.4651851851851852,
+      "grad_norm": 2.0309510231018066,
+      "learning_rate": 1.6905749851807944e-05,
+      "loss": 0.5139,
+      "step": 785
+    },
+    {
+      "epoch": 0.4657777777777778,
+      "grad_norm": 2.859248161315918,
+      "learning_rate": 1.690179806362379e-05,
+      "loss": 0.4774,
+      "step": 786
+    },
+    {
+      "epoch": 0.4663703703703704,
+      "grad_norm": 1.8914960622787476,
+      "learning_rate": 1.689784627543964e-05,
+      "loss": 0.3826,
+      "step": 787
+    },
+    {
+      "epoch": 0.46696296296296297,
+      "grad_norm": 1.6827701330184937,
+      "learning_rate": 1.6893894487255484e-05,
+      "loss": 0.3763,
+      "step": 788
+    },
+    {
+      "epoch": 0.46755555555555556,
+      "grad_norm": 1.9770413637161255,
+      "learning_rate": 1.6889942699071332e-05,
+      "loss": 0.4589,
+      "step": 789
+    },
+    {
+      "epoch": 0.46814814814814815,
+      "grad_norm": 1.8565627336502075,
+      "learning_rate": 1.6885990910887176e-05,
+      "loss": 0.4016,
+      "step": 790
+    },
+    {
+      "epoch": 0.46874074074074074,
+      "grad_norm": 3.2250149250030518,
+      "learning_rate": 1.6882039122703024e-05,
+      "loss": 0.3617,
+      "step": 791
+    },
+    {
+      "epoch": 0.4693333333333333,
+      "grad_norm": 3.2301926612854004,
+      "learning_rate": 1.6878087334518872e-05,
+      "loss": 0.4438,
+      "step": 792
+    },
+    {
+      "epoch": 0.4699259259259259,
+      "grad_norm": 2.7895421981811523,
+      "learning_rate": 1.6874135546334717e-05,
+      "loss": 0.4196,
+      "step": 793
+    },
+    {
+      "epoch": 0.4705185185185185,
+      "grad_norm": 2.311990976333618,
+      "learning_rate": 1.6870183758150565e-05,
+      "loss": 0.3821,
+      "step": 794
+    },
+    {
+      "epoch": 0.4711111111111111,
+      "grad_norm": 2.3196358680725098,
+      "learning_rate": 1.6866231969966413e-05,
+      "loss": 0.3802,
+      "step": 795
+    },
+    {
+      "epoch": 0.4717037037037037,
+      "grad_norm": 3.462026596069336,
+      "learning_rate": 1.6862280181782257e-05,
+      "loss": 0.557,
+      "step": 796
+    },
+    {
+      "epoch": 0.47229629629629627,
+      "grad_norm": 3.843899965286255,
+      "learning_rate": 1.6858328393598105e-05,
+      "loss": 0.4843,
+      "step": 797
+    },
+    {
+      "epoch": 0.4728888888888889,
+      "grad_norm": 2.7698090076446533,
+      "learning_rate": 1.685437660541395e-05,
+      "loss": 0.4862,
+      "step": 798
+    },
+    {
+      "epoch": 0.4734814814814815,
+      "grad_norm": 1.7282465696334839,
+      "learning_rate": 1.6850424817229798e-05,
+      "loss": 0.5397,
+      "step": 799
+    },
+    {
+      "epoch": 0.4740740740740741,
+      "grad_norm": 2.73073673248291,
+      "learning_rate": 1.6846473029045646e-05,
+      "loss": 0.4459,
+      "step": 800
+    },
+    {
+      "epoch": 0.4746666666666667,
+      "grad_norm": 1.7496604919433594,
+      "learning_rate": 1.684252124086149e-05,
+      "loss": 0.4833,
+      "step": 801
+    },
+    {
+      "epoch": 0.47525925925925927,
+      "grad_norm": 4.049375534057617,
+      "learning_rate": 1.6838569452677338e-05,
+      "loss": 0.6647,
+      "step": 802
+    },
+    {
+      "epoch": 0.47585185185185186,
+      "grad_norm": 2.253788948059082,
+      "learning_rate": 1.6834617664493186e-05,
+      "loss": 0.4326,
+      "step": 803
+    },
+    {
+      "epoch": 0.47644444444444445,
+      "grad_norm": 1.5722308158874512,
+      "learning_rate": 1.683066587630903e-05,
+      "loss": 0.4566,
+      "step": 804
+    },
+    {
+      "epoch": 0.47703703703703704,
+      "grad_norm": 1.8845884799957275,
+      "learning_rate": 1.682671408812488e-05,
+      "loss": 0.5138,
+      "step": 805
+    },
+    {
+      "epoch": 0.4776296296296296,
+      "grad_norm": 1.9207592010498047,
+      "learning_rate": 1.6822762299940723e-05,
+      "loss": 0.47,
+      "step": 806
+    },
+    {
+      "epoch": 0.4782222222222222,
+      "grad_norm": 3.50961971282959,
+      "learning_rate": 1.681881051175657e-05,
+      "loss": 0.4661,
+      "step": 807
+    },
+    {
+      "epoch": 0.4788148148148148,
+      "grad_norm": 6.050410270690918,
+      "learning_rate": 1.681485872357242e-05,
+      "loss": 0.6105,
+      "step": 808
+    },
+    {
+      "epoch": 0.4794074074074074,
+      "grad_norm": 2.432921886444092,
+      "learning_rate": 1.6810906935388264e-05,
+      "loss": 0.3745,
+      "step": 809
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9763994216918945,
+      "learning_rate": 1.680695514720411e-05,
+      "loss": 0.4823,
+      "step": 810
+    },
+    {
+      "epoch": 0.48059259259259257,
+      "grad_norm": 3.9631431102752686,
+      "learning_rate": 1.680300335901996e-05,
+      "loss": 0.5348,
+      "step": 811
+    },
+    {
+      "epoch": 0.48118518518518516,
+      "grad_norm": 1.6850072145462036,
+      "learning_rate": 1.6799051570835804e-05,
+      "loss": 0.468,
+      "step": 812
+    },
+    {
+      "epoch": 0.4817777777777778,
+      "grad_norm": 2.571612596511841,
+      "learning_rate": 1.6795099782651652e-05,
+      "loss": 0.4137,
+      "step": 813
+    },
+    {
+      "epoch": 0.4823703703703704,
+      "grad_norm": 3.943154811859131,
+      "learning_rate": 1.67911479944675e-05,
+      "loss": 0.5786,
+      "step": 814
+    },
+    {
+      "epoch": 0.482962962962963,
+      "grad_norm": 6.187957763671875,
+      "learning_rate": 1.6787196206283344e-05,
+      "loss": 0.4541,
+      "step": 815
+    },
+    {
+      "epoch": 0.48355555555555557,
+      "grad_norm": 1.5988909006118774,
+      "learning_rate": 1.6783244418099192e-05,
+      "loss": 0.3643,
+      "step": 816
+    },
+    {
+      "epoch": 0.48414814814814816,
+      "grad_norm": 1.505283236503601,
+      "learning_rate": 1.6779292629915037e-05,
+      "loss": 0.3265,
+      "step": 817
+    },
+    {
+      "epoch": 0.48474074074074075,
+      "grad_norm": 1.8044296503067017,
+      "learning_rate": 1.6775340841730885e-05,
+      "loss": 0.4347,
+      "step": 818
+    },
+    {
+      "epoch": 0.48533333333333334,
+      "grad_norm": 2.5203919410705566,
+      "learning_rate": 1.6771389053546733e-05,
+      "loss": 0.5117,
+      "step": 819
+    },
+    {
+      "epoch": 0.48592592592592593,
+      "grad_norm": 1.6357861757278442,
+      "learning_rate": 1.6767437265362577e-05,
+      "loss": 0.4199,
+      "step": 820
+    },
+    {
+      "epoch": 0.4865185185185185,
+      "grad_norm": 2.35353684425354,
+      "learning_rate": 1.6763485477178425e-05,
+      "loss": 0.4384,
+      "step": 821
+    },
+    {
+      "epoch": 0.4871111111111111,
+      "grad_norm": 2.345845937728882,
+      "learning_rate": 1.6759533688994273e-05,
+      "loss": 0.3418,
+      "step": 822
+    },
+    {
+      "epoch": 0.4877037037037037,
+      "grad_norm": 1.7866135835647583,
+      "learning_rate": 1.6755581900810118e-05,
+      "loss": 0.3309,
+      "step": 823
+    },
+    {
+      "epoch": 0.4882962962962963,
+      "grad_norm": 1.885056495666504,
+      "learning_rate": 1.6751630112625966e-05,
+      "loss": 0.4258,
+      "step": 824
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 4.475062847137451,
+      "learning_rate": 1.674767832444181e-05,
+      "loss": 0.5035,
+      "step": 825
+    },
+    {
+      "epoch": 0.48948148148148146,
+      "grad_norm": 1.7886329889297485,
+      "learning_rate": 1.6743726536257658e-05,
+      "loss": 0.501,
+      "step": 826
+    },
+    {
+      "epoch": 0.49007407407407405,
+      "grad_norm": 3.058565616607666,
+      "learning_rate": 1.6739774748073506e-05,
+      "loss": 0.4883,
+      "step": 827
+    },
+    {
+      "epoch": 0.49066666666666664,
+      "grad_norm": 1.5245200395584106,
+      "learning_rate": 1.673582295988935e-05,
+      "loss": 0.453,
+      "step": 828
+    },
+    {
+      "epoch": 0.4912592592592593,
+      "grad_norm": 1.4573155641555786,
+      "learning_rate": 1.67318711717052e-05,
+      "loss": 0.4035,
+      "step": 829
+    },
+    {
+      "epoch": 0.4918518518518519,
+      "grad_norm": 1.7327078580856323,
+      "learning_rate": 1.6727919383521046e-05,
+      "loss": 0.3943,
+      "step": 830
+    },
+    {
+      "epoch": 0.49244444444444446,
+      "grad_norm": 1.940186858177185,
+      "learning_rate": 1.672396759533689e-05,
+      "loss": 0.4808,
+      "step": 831
+    },
+    {
+      "epoch": 0.49303703703703705,
+      "grad_norm": 2.25527286529541,
+      "learning_rate": 1.672001580715274e-05,
+      "loss": 0.3518,
+      "step": 832
+    },
+    {
+      "epoch": 0.49362962962962964,
+      "grad_norm": 2.1204354763031006,
+      "learning_rate": 1.6716064018968587e-05,
+      "loss": 0.4222,
+      "step": 833
+    },
+    {
+      "epoch": 0.49422222222222223,
+      "grad_norm": 1.9461992979049683,
+      "learning_rate": 1.671211223078443e-05,
+      "loss": 0.4233,
+      "step": 834
+    },
+    {
+      "epoch": 0.4948148148148148,
+      "grad_norm": 1.930467963218689,
+      "learning_rate": 1.670816044260028e-05,
+      "loss": 0.4315,
+      "step": 835
+    },
+    {
+      "epoch": 0.4954074074074074,
+      "grad_norm": 1.6301329135894775,
+      "learning_rate": 1.6704208654416124e-05,
+      "loss": 0.4054,
+      "step": 836
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 1.883833885192871,
+      "learning_rate": 1.6700256866231972e-05,
+      "loss": 0.5408,
+      "step": 837
+    },
+    {
+      "epoch": 0.4965925925925926,
+      "grad_norm": 1.5842968225479126,
+      "learning_rate": 1.669630507804782e-05,
+      "loss": 0.4132,
+      "step": 838
+    },
+    {
+      "epoch": 0.4971851851851852,
+      "grad_norm": 1.5667299032211304,
+      "learning_rate": 1.6692353289863664e-05,
+      "loss": 0.3737,
+      "step": 839
+    },
+    {
+      "epoch": 0.49777777777777776,
+      "grad_norm": 7.742247104644775,
+      "learning_rate": 1.6688401501679512e-05,
+      "loss": 0.7435,
+      "step": 840
+    },
+    {
+      "epoch": 0.49837037037037035,
+      "grad_norm": 2.5636415481567383,
+      "learning_rate": 1.668444971349536e-05,
+      "loss": 0.4405,
+      "step": 841
+    },
+    {
+      "epoch": 0.49896296296296294,
+      "grad_norm": 1.382217526435852,
+      "learning_rate": 1.6680497925311205e-05,
+      "loss": 0.3467,
+      "step": 842
+    },
+    {
+      "epoch": 0.49955555555555553,
+      "grad_norm": 3.4491465091705322,
+      "learning_rate": 1.6676546137127053e-05,
+      "loss": 0.4103,
+      "step": 843
+    },
+    {
+      "epoch": 0.5001481481481481,
+      "grad_norm": 1.8836669921875,
+      "learning_rate": 1.6672594348942897e-05,
+      "loss": 0.4035,
+      "step": 844
+    },
+    {
+      "epoch": 0.5007407407407407,
+      "grad_norm": 2.271338939666748,
+      "learning_rate": 1.6668642560758745e-05,
+      "loss": 0.5201,
+      "step": 845
+    },
+    {
+      "epoch": 0.5013333333333333,
+      "grad_norm": 1.9047560691833496,
+      "learning_rate": 1.6664690772574593e-05,
+      "loss": 0.3839,
+      "step": 846
+    },
+    {
+      "epoch": 0.5019259259259259,
+      "grad_norm": 2.0606637001037598,
+      "learning_rate": 1.6660738984390438e-05,
+      "loss": 0.5109,
+      "step": 847
+    },
+    {
+      "epoch": 0.5025185185185185,
+      "grad_norm": 1.7151490449905396,
+      "learning_rate": 1.6656787196206286e-05,
+      "loss": 0.3551,
+      "step": 848
+    },
+    {
+      "epoch": 0.5031111111111111,
+      "grad_norm": 1.941512107849121,
+      "learning_rate": 1.6652835408022134e-05,
+      "loss": 0.4066,
+      "step": 849
+    },
+    {
+      "epoch": 0.5037037037037037,
+      "grad_norm": 1.6601605415344238,
+      "learning_rate": 1.6648883619837978e-05,
+      "loss": 0.3737,
+      "step": 850
+    },
+    {
+      "epoch": 0.5042962962962962,
+      "grad_norm": 2.329946756362915,
+      "learning_rate": 1.6644931831653826e-05,
+      "loss": 0.4085,
+      "step": 851
+    },
+    {
+      "epoch": 0.5048888888888889,
+      "grad_norm": 2.6266963481903076,
+      "learning_rate": 1.664098004346967e-05,
+      "loss": 0.3658,
+      "step": 852
+    },
+    {
+      "epoch": 0.5054814814814815,
+      "grad_norm": 3.0301928520202637,
+      "learning_rate": 1.663702825528552e-05,
+      "loss": 0.462,
+      "step": 853
+    },
+    {
+      "epoch": 0.5060740740740741,
+      "grad_norm": 1.5690515041351318,
+      "learning_rate": 1.6633076467101366e-05,
+      "loss": 0.3748,
+      "step": 854
+    },
+    {
+      "epoch": 0.5066666666666667,
+      "grad_norm": 1.3143067359924316,
+      "learning_rate": 1.662912467891721e-05,
+      "loss": 0.3167,
+      "step": 855
+    },
+    {
+      "epoch": 0.5072592592592593,
+      "grad_norm": 2.4784634113311768,
+      "learning_rate": 1.662517289073306e-05,
+      "loss": 0.4519,
+      "step": 856
+    },
+    {
+      "epoch": 0.5078518518518519,
+      "grad_norm": 3.215951919555664,
+      "learning_rate": 1.6621221102548907e-05,
+      "loss": 0.5383,
+      "step": 857
+    },
+    {
+      "epoch": 0.5084444444444445,
+      "grad_norm": 2.394343137741089,
+      "learning_rate": 1.661726931436475e-05,
+      "loss": 0.3247,
+      "step": 858
+    },
+    {
+      "epoch": 0.5090370370370371,
+      "grad_norm": 3.1337265968322754,
+      "learning_rate": 1.66133175261806e-05,
+      "loss": 0.6525,
+      "step": 859
+    },
+    {
+      "epoch": 0.5096296296296297,
+      "grad_norm": 1.558388352394104,
+      "learning_rate": 1.6609365737996447e-05,
+      "loss": 0.4199,
+      "step": 860
+    },
+    {
+      "epoch": 0.5102222222222222,
+      "grad_norm": 2.0167205333709717,
+      "learning_rate": 1.6605413949812292e-05,
+      "loss": 0.4222,
+      "step": 861
+    },
+    {
+      "epoch": 0.5108148148148148,
+      "grad_norm": 2.8771591186523438,
+      "learning_rate": 1.660146216162814e-05,
+      "loss": 0.4012,
+      "step": 862
+    },
+    {
+      "epoch": 0.5114074074074074,
+      "grad_norm": 2.9749057292938232,
+      "learning_rate": 1.6597510373443984e-05,
+      "loss": 0.4056,
+      "step": 863
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 1.6909011602401733,
+      "learning_rate": 1.6593558585259832e-05,
+      "loss": 0.4038,
+      "step": 864
+    },
+    {
+      "epoch": 0.5125925925925926,
+      "grad_norm": 2.4711086750030518,
+      "learning_rate": 1.658960679707568e-05,
+      "loss": 0.4246,
+      "step": 865
+    },
+    {
+      "epoch": 0.5131851851851852,
+      "grad_norm": 1.8801990747451782,
+      "learning_rate": 1.6585655008891525e-05,
+      "loss": 0.3629,
+      "step": 866
+    },
+    {
+      "epoch": 0.5137777777777778,
+      "grad_norm": 1.7055833339691162,
+      "learning_rate": 1.6581703220707373e-05,
+      "loss": 0.3761,
+      "step": 867
+    },
+    {
+      "epoch": 0.5143703703703704,
+      "grad_norm": 1.6695557832717896,
+      "learning_rate": 1.657775143252322e-05,
+      "loss": 0.4145,
+      "step": 868
+    },
+    {
+      "epoch": 0.514962962962963,
+      "grad_norm": 1.594416856765747,
+      "learning_rate": 1.6573799644339065e-05,
+      "loss": 0.3535,
+      "step": 869
+    },
+    {
+      "epoch": 0.5155555555555555,
+      "grad_norm": 1.6270569562911987,
+      "learning_rate": 1.6569847856154913e-05,
+      "loss": 0.3849,
+      "step": 870
+    },
+    {
+      "epoch": 0.5161481481481481,
+      "grad_norm": 3.125314235687256,
+      "learning_rate": 1.6565896067970758e-05,
+      "loss": 0.4339,
+      "step": 871
+    },
+    {
+      "epoch": 0.5167407407407407,
+      "grad_norm": 1.5845502614974976,
+      "learning_rate": 1.6561944279786602e-05,
+      "loss": 0.4034,
+      "step": 872
+    },
+    {
+      "epoch": 0.5173333333333333,
+      "grad_norm": 1.953001856803894,
+      "learning_rate": 1.655799249160245e-05,
+      "loss": 0.5003,
+      "step": 873
+    },
+    {
+      "epoch": 0.5179259259259259,
+      "grad_norm": 1.7179248332977295,
+      "learning_rate": 1.6554040703418298e-05,
+      "loss": 0.3934,
+      "step": 874
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 1.9158226251602173,
+      "learning_rate": 1.6550088915234143e-05,
+      "loss": 0.4354,
+      "step": 875
+    },
+    {
+      "epoch": 0.5191111111111111,
+      "grad_norm": 1.9179909229278564,
+      "learning_rate": 1.654613712704999e-05,
+      "loss": 0.3619,
+      "step": 876
+    },
+    {
+      "epoch": 0.5197037037037037,
+      "grad_norm": 2.4242405891418457,
+      "learning_rate": 1.654218533886584e-05,
+      "loss": 0.5157,
+      "step": 877
+    },
+    {
+      "epoch": 0.5202962962962963,
+      "grad_norm": 2.4145619869232178,
+      "learning_rate": 1.6538233550681683e-05,
+      "loss": 0.4571,
+      "step": 878
+    },
+    {
+      "epoch": 0.5208888888888888,
+      "grad_norm": 2.2124950885772705,
+      "learning_rate": 1.653428176249753e-05,
+      "loss": 0.5962,
+      "step": 879
+    },
+    {
+      "epoch": 0.5214814814814814,
+      "grad_norm": 2.4698147773742676,
+      "learning_rate": 1.6530329974313376e-05,
+      "loss": 0.3663,
+      "step": 880
+    },
+    {
+      "epoch": 0.522074074074074,
+      "grad_norm": 1.627457857131958,
+      "learning_rate": 1.6526378186129223e-05,
+      "loss": 0.3627,
+      "step": 881
+    },
+    {
+      "epoch": 0.5226666666666666,
+      "grad_norm": 2.557483196258545,
+      "learning_rate": 1.652242639794507e-05,
+      "loss": 0.3564,
+      "step": 882
+    },
+    {
+      "epoch": 0.5232592592592593,
+      "grad_norm": 1.4525573253631592,
+      "learning_rate": 1.6518474609760916e-05,
+      "loss": 0.3559,
+      "step": 883
+    },
+    {
+      "epoch": 0.5238518518518519,
+      "grad_norm": 1.7047544717788696,
+      "learning_rate": 1.6514522821576764e-05,
+      "loss": 0.4622,
+      "step": 884
+    },
+    {
+      "epoch": 0.5244444444444445,
+      "grad_norm": 2.669679641723633,
+      "learning_rate": 1.6510571033392612e-05,
+      "loss": 0.4067,
+      "step": 885
+    },
+    {
+      "epoch": 0.5250370370370371,
+      "grad_norm": 2.836902379989624,
+      "learning_rate": 1.6506619245208456e-05,
+      "loss": 0.4547,
+      "step": 886
+    },
+    {
+      "epoch": 0.5256296296296297,
+      "grad_norm": 1.8108097314834595,
+      "learning_rate": 1.6502667457024304e-05,
+      "loss": 0.4591,
+      "step": 887
+    },
+    {
+      "epoch": 0.5262222222222223,
+      "grad_norm": 2.50296688079834,
+      "learning_rate": 1.6498715668840152e-05,
+      "loss": 0.5086,
+      "step": 888
+    },
+    {
+      "epoch": 0.5268148148148148,
+      "grad_norm": 2.2456374168395996,
+      "learning_rate": 1.6494763880655997e-05,
+      "loss": 0.3928,
+      "step": 889
+    },
+    {
+      "epoch": 0.5274074074074074,
+      "grad_norm": 1.901884913444519,
+      "learning_rate": 1.6490812092471845e-05,
+      "loss": 0.3317,
+      "step": 890
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 2.8121235370635986,
+      "learning_rate": 1.648686030428769e-05,
+      "loss": 0.3444,
+      "step": 891
+    },
+    {
+      "epoch": 0.5285925925925926,
+      "grad_norm": 2.198828935623169,
+      "learning_rate": 1.6482908516103537e-05,
+      "loss": 0.3497,
+      "step": 892
+    },
+    {
+      "epoch": 0.5291851851851852,
+      "grad_norm": 1.7671470642089844,
+      "learning_rate": 1.6478956727919385e-05,
+      "loss": 0.4632,
+      "step": 893
+    },
+    {
+      "epoch": 0.5297777777777778,
+      "grad_norm": 1.970933198928833,
+      "learning_rate": 1.647500493973523e-05,
+      "loss": 0.4108,
+      "step": 894
+    },
+    {
+      "epoch": 0.5303703703703704,
+      "grad_norm": 2.201775550842285,
+      "learning_rate": 1.6471053151551078e-05,
+      "loss": 0.3559,
+      "step": 895
+    },
+    {
+      "epoch": 0.530962962962963,
+      "grad_norm": 3.8010334968566895,
+      "learning_rate": 1.6467101363366926e-05,
+      "loss": 0.4963,
+      "step": 896
+    },
+    {
+      "epoch": 0.5315555555555556,
+      "grad_norm": 2.0509703159332275,
+      "learning_rate": 1.646314957518277e-05,
+      "loss": 0.4048,
+      "step": 897
+    },
+    {
+      "epoch": 0.5321481481481481,
+      "grad_norm": 1.609839677810669,
+      "learning_rate": 1.6459197786998618e-05,
+      "loss": 0.3826,
+      "step": 898
+    },
+    {
+      "epoch": 0.5327407407407407,
+      "grad_norm": 1.713024616241455,
+      "learning_rate": 1.6455245998814463e-05,
+      "loss": 0.4775,
+      "step": 899
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.9199845790863037,
+      "learning_rate": 1.645129421063031e-05,
+      "loss": 0.3964,
+      "step": 900
+    },
+    {
+      "epoch": 0.5339259259259259,
+      "grad_norm": 3.862412929534912,
+      "learning_rate": 1.644734242244616e-05,
+      "loss": 0.4794,
+      "step": 901
+    },
+    {
+      "epoch": 0.5345185185185185,
+      "grad_norm": 2.282963752746582,
+      "learning_rate": 1.6443390634262003e-05,
+      "loss": 0.409,
+      "step": 902
+    },
+    {
+      "epoch": 0.5351111111111111,
+      "grad_norm": 1.6303036212921143,
+      "learning_rate": 1.643943884607785e-05,
+      "loss": 0.4307,
+      "step": 903
+    },
+    {
+      "epoch": 0.5357037037037037,
+      "grad_norm": 3.242835760116577,
+      "learning_rate": 1.64354870578937e-05,
+      "loss": 0.4538,
+      "step": 904
+    },
+    {
+      "epoch": 0.5362962962962963,
+      "grad_norm": 1.8112117052078247,
+      "learning_rate": 1.6431535269709543e-05,
+      "loss": 0.3326,
+      "step": 905
+    },
+    {
+      "epoch": 0.5368888888888889,
+      "grad_norm": 2.314208745956421,
+      "learning_rate": 1.642758348152539e-05,
+      "loss": 0.4307,
+      "step": 906
+    },
+    {
+      "epoch": 0.5374814814814814,
+      "grad_norm": 2.4123013019561768,
+      "learning_rate": 1.6423631693341236e-05,
+      "loss": 0.5171,
+      "step": 907
+    },
+    {
+      "epoch": 0.538074074074074,
+      "grad_norm": 2.3287627696990967,
+      "learning_rate": 1.6419679905157084e-05,
+      "loss": 0.4521,
+      "step": 908
+    },
+    {
+      "epoch": 0.5386666666666666,
+      "grad_norm": 3.9871442317962646,
+      "learning_rate": 1.6415728116972932e-05,
+      "loss": 0.4435,
+      "step": 909
+    },
+    {
+      "epoch": 0.5392592592592592,
+      "grad_norm": 2.49007511138916,
+      "learning_rate": 1.6411776328788776e-05,
+      "loss": 0.4209,
+      "step": 910
+    },
+    {
+      "epoch": 0.5398518518518518,
+      "grad_norm": 2.3307416439056396,
+      "learning_rate": 1.6407824540604624e-05,
+      "loss": 0.5451,
+      "step": 911
+    },
+    {
+      "epoch": 0.5404444444444444,
+      "grad_norm": 2.6301157474517822,
+      "learning_rate": 1.6403872752420472e-05,
+      "loss": 0.3814,
+      "step": 912
+    },
+    {
+      "epoch": 0.541037037037037,
+      "grad_norm": 2.855236530303955,
+      "learning_rate": 1.6399920964236317e-05,
+      "loss": 0.4146,
+      "step": 913
+    },
+    {
+      "epoch": 0.5416296296296297,
+      "grad_norm": 2.4257519245147705,
+      "learning_rate": 1.6395969176052165e-05,
+      "loss": 0.5281,
+      "step": 914
+    },
+    {
+      "epoch": 0.5422222222222223,
+      "grad_norm": 2.085019111633301,
+      "learning_rate": 1.6392017387868013e-05,
+      "loss": 0.4442,
+      "step": 915
+    },
+    {
+      "epoch": 0.5428148148148149,
+      "grad_norm": 2.906174898147583,
+      "learning_rate": 1.6388065599683857e-05,
+      "loss": 0.472,
+      "step": 916
+    },
+    {
+      "epoch": 0.5434074074074075,
+      "grad_norm": 1.7140858173370361,
+      "learning_rate": 1.6384113811499705e-05,
+      "loss": 0.3709,
+      "step": 917
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 1.4112027883529663,
+      "learning_rate": 1.638016202331555e-05,
+      "loss": 0.308,
+      "step": 918
+    },
+    {
+      "epoch": 0.5445925925925926,
+      "grad_norm": 1.8806043863296509,
+      "learning_rate": 1.6376210235131398e-05,
+      "loss": 0.3339,
+      "step": 919
+    },
+    {
+      "epoch": 0.5451851851851852,
+      "grad_norm": 1.506742000579834,
+      "learning_rate": 1.6372258446947246e-05,
+      "loss": 0.3837,
+      "step": 920
+    },
+    {
+      "epoch": 0.5457777777777778,
+      "grad_norm": 2.1902365684509277,
+      "learning_rate": 1.636830665876309e-05,
+      "loss": 0.4343,
+      "step": 921
+    },
+    {
+      "epoch": 0.5463703703703704,
+      "grad_norm": 5.791208267211914,
+      "learning_rate": 1.6364354870578938e-05,
+      "loss": 0.6911,
+      "step": 922
+    },
+    {
+      "epoch": 0.546962962962963,
+      "grad_norm": 3.284156322479248,
+      "learning_rate": 1.6360403082394786e-05,
+      "loss": 0.433,
+      "step": 923
+    },
+    {
+      "epoch": 0.5475555555555556,
+      "grad_norm": 1.9776136875152588,
+      "learning_rate": 1.635645129421063e-05,
+      "loss": 0.4249,
+      "step": 924
+    },
+    {
+      "epoch": 0.5481481481481482,
+      "grad_norm": 1.4219175577163696,
+      "learning_rate": 1.635249950602648e-05,
+      "loss": 0.2947,
+      "step": 925
+    },
+    {
+      "epoch": 0.5487407407407408,
+      "grad_norm": 1.7355678081512451,
+      "learning_rate": 1.6348547717842323e-05,
+      "loss": 0.3175,
+      "step": 926
+    },
+    {
+      "epoch": 0.5493333333333333,
+      "grad_norm": 1.91839599609375,
+      "learning_rate": 1.634459592965817e-05,
+      "loss": 0.4483,
+      "step": 927
+    },
+    {
+      "epoch": 0.5499259259259259,
+      "grad_norm": 2.0752573013305664,
+      "learning_rate": 1.634064414147402e-05,
+      "loss": 0.4487,
+      "step": 928
+    },
+    {
+      "epoch": 0.5505185185185185,
+      "grad_norm": 1.6736180782318115,
+      "learning_rate": 1.6336692353289863e-05,
+      "loss": 0.3603,
+      "step": 929
+    },
+    {
+      "epoch": 0.5511111111111111,
+      "grad_norm": 1.946247935295105,
+      "learning_rate": 1.633274056510571e-05,
+      "loss": 0.3827,
+      "step": 930
+    },
+    {
+      "epoch": 0.5517037037037037,
+      "grad_norm": 2.393035888671875,
+      "learning_rate": 1.632878877692156e-05,
+      "loss": 0.4452,
+      "step": 931
+    },
+    {
+      "epoch": 0.5522962962962963,
+      "grad_norm": 2.097799777984619,
+      "learning_rate": 1.6324836988737404e-05,
+      "loss": 0.4291,
+      "step": 932
+    },
+    {
+      "epoch": 0.5528888888888889,
+      "grad_norm": 1.7746126651763916,
+      "learning_rate": 1.6320885200553252e-05,
+      "loss": 0.3657,
+      "step": 933
+    },
+    {
+      "epoch": 0.5534814814814815,
+      "grad_norm": 2.3026392459869385,
+      "learning_rate": 1.63169334123691e-05,
+      "loss": 0.4992,
+      "step": 934
+    },
+    {
+      "epoch": 0.554074074074074,
+      "grad_norm": 1.6159485578536987,
+      "learning_rate": 1.6312981624184944e-05,
+      "loss": 0.3753,
+      "step": 935
+    },
+    {
+      "epoch": 0.5546666666666666,
+      "grad_norm": 2.1636786460876465,
+      "learning_rate": 1.6309029836000792e-05,
+      "loss": 0.3792,
+      "step": 936
+    },
+    {
+      "epoch": 0.5552592592592592,
+      "grad_norm": 2.23803973197937,
+      "learning_rate": 1.6305078047816637e-05,
+      "loss": 0.4137,
+      "step": 937
+    },
+    {
+      "epoch": 0.5558518518518518,
+      "grad_norm": 2.9558911323547363,
+      "learning_rate": 1.6301126259632485e-05,
+      "loss": 0.4985,
+      "step": 938
+    },
+    {
+      "epoch": 0.5564444444444444,
+      "grad_norm": 2.588681936264038,
+      "learning_rate": 1.6297174471448333e-05,
+      "loss": 0.4434,
+      "step": 939
+    },
+    {
+      "epoch": 0.557037037037037,
+      "grad_norm": 1.772037386894226,
+      "learning_rate": 1.6293222683264177e-05,
+      "loss": 0.4065,
+      "step": 940
+    },
+    {
+      "epoch": 0.5576296296296296,
+      "grad_norm": 1.7415170669555664,
+      "learning_rate": 1.6289270895080025e-05,
+      "loss": 0.3816,
+      "step": 941
+    },
+    {
+      "epoch": 0.5582222222222222,
+      "grad_norm": 6.751112461090088,
+      "learning_rate": 1.6285319106895873e-05,
+      "loss": 0.8038,
+      "step": 942
+    },
+    {
+      "epoch": 0.5588148148148148,
+      "grad_norm": 2.1469688415527344,
+      "learning_rate": 1.6281367318711718e-05,
+      "loss": 0.447,
+      "step": 943
+    },
+    {
+      "epoch": 0.5594074074074074,
+      "grad_norm": 2.667083978652954,
+      "learning_rate": 1.6277415530527566e-05,
+      "loss": 0.5034,
+      "step": 944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.1367621421813965,
+      "learning_rate": 1.627346374234341e-05,
+      "loss": 0.6283,
+      "step": 945
+    },
+    {
+      "epoch": 0.5605925925925926,
+      "grad_norm": 2.262465000152588,
+      "learning_rate": 1.6269511954159258e-05,
+      "loss": 0.4717,
+      "step": 946
+    },
+    {
+      "epoch": 0.5611851851851852,
+      "grad_norm": 2.656388521194458,
+      "learning_rate": 1.6265560165975106e-05,
+      "loss": 0.4098,
+      "step": 947
+    },
+    {
+      "epoch": 0.5617777777777778,
+      "grad_norm": 2.2412726879119873,
+      "learning_rate": 1.626160837779095e-05,
+      "loss": 0.399,
+      "step": 948
+    },
+    {
+      "epoch": 0.5623703703703704,
+      "grad_norm": 2.5334713459014893,
+      "learning_rate": 1.62576565896068e-05,
+      "loss": 0.3237,
+      "step": 949
+    },
+    {
+      "epoch": 0.562962962962963,
+      "grad_norm": 2.040283441543579,
+      "learning_rate": 1.6253704801422646e-05,
+      "loss": 0.4163,
+      "step": 950
+    },
+    {
+      "epoch": 0.5635555555555556,
+      "grad_norm": 3.829448699951172,
+      "learning_rate": 1.624975301323849e-05,
+      "loss": 0.452,
+      "step": 951
+    },
+    {
+      "epoch": 0.5641481481481482,
+      "grad_norm": 2.3458940982818604,
+      "learning_rate": 1.624580122505434e-05,
+      "loss": 0.394,
+      "step": 952
+    },
+    {
+      "epoch": 0.5647407407407408,
+      "grad_norm": 3.5965986251831055,
+      "learning_rate": 1.6241849436870183e-05,
+      "loss": 0.434,
+      "step": 953
+    },
+    {
+      "epoch": 0.5653333333333334,
+      "grad_norm": 3.055851459503174,
+      "learning_rate": 1.623789764868603e-05,
+      "loss": 0.419,
+      "step": 954
+    },
+    {
+      "epoch": 0.5659259259259259,
+      "grad_norm": 2.7926788330078125,
+      "learning_rate": 1.623394586050188e-05,
+      "loss": 0.4347,
+      "step": 955
+    },
+    {
+      "epoch": 0.5665185185185185,
+      "grad_norm": 2.593423843383789,
+      "learning_rate": 1.6229994072317724e-05,
+      "loss": 0.5499,
+      "step": 956
+    },
+    {
+      "epoch": 0.5671111111111111,
+      "grad_norm": 1.442123293876648,
+      "learning_rate": 1.6226042284133572e-05,
+      "loss": 0.3736,
+      "step": 957
+    },
+    {
+      "epoch": 0.5677037037037037,
+      "grad_norm": 2.270995855331421,
+      "learning_rate": 1.622209049594942e-05,
+      "loss": 0.412,
+      "step": 958
+    },
+    {
+      "epoch": 0.5682962962962963,
+      "grad_norm": 3.025102376937866,
+      "learning_rate": 1.6218138707765264e-05,
+      "loss": 0.4722,
+      "step": 959
+    },
+    {
+      "epoch": 0.5688888888888889,
+      "grad_norm": 2.4052791595458984,
+      "learning_rate": 1.6214186919581112e-05,
+      "loss": 0.3246,
+      "step": 960
+    },
+    {
+      "epoch": 0.5694814814814815,
+      "grad_norm": 1.9652084112167358,
+      "learning_rate": 1.621023513139696e-05,
+      "loss": 0.4106,
+      "step": 961
+    },
+    {
+      "epoch": 0.5700740740740741,
+      "grad_norm": 2.001620292663574,
+      "learning_rate": 1.6206283343212805e-05,
+      "loss": 0.422,
+      "step": 962
+    },
+    {
+      "epoch": 0.5706666666666667,
+      "grad_norm": 1.7509922981262207,
+      "learning_rate": 1.6202331555028653e-05,
+      "loss": 0.425,
+      "step": 963
+    },
+    {
+      "epoch": 0.5712592592592592,
+      "grad_norm": 2.3737003803253174,
+      "learning_rate": 1.6198379766844497e-05,
+      "loss": 0.5023,
+      "step": 964
+    },
+    {
+      "epoch": 0.5718518518518518,
+      "grad_norm": 5.123398303985596,
+      "learning_rate": 1.6194427978660345e-05,
+      "loss": 0.4614,
+      "step": 965
+    },
+    {
+      "epoch": 0.5724444444444444,
+      "grad_norm": 1.8972346782684326,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 0.397,
+      "step": 966
+    },
+    {
+      "epoch": 0.573037037037037,
+      "grad_norm": 2.567357301712036,
+      "learning_rate": 1.6186524402292038e-05,
+      "loss": 0.5287,
+      "step": 967
+    },
+    {
+      "epoch": 0.5736296296296296,
+      "grad_norm": 2.7633681297302246,
+      "learning_rate": 1.6182572614107886e-05,
+      "loss": 0.4786,
+      "step": 968
+    },
+    {
+      "epoch": 0.5742222222222222,
+      "grad_norm": 2.022059917449951,
+      "learning_rate": 1.6178620825923734e-05,
+      "loss": 0.3515,
+      "step": 969
+    },
+    {
+      "epoch": 0.5748148148148148,
+      "grad_norm": 3.052872896194458,
+      "learning_rate": 1.6174669037739578e-05,
+      "loss": 0.4474,
+      "step": 970
+    },
+    {
+      "epoch": 0.5754074074074074,
+      "grad_norm": 3.8861618041992188,
+      "learning_rate": 1.6170717249555426e-05,
+      "loss": 0.3671,
+      "step": 971
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 2.9867069721221924,
+      "learning_rate": 1.616676546137127e-05,
+      "loss": 0.4872,
+      "step": 972
+    },
+    {
+      "epoch": 0.5765925925925925,
+      "grad_norm": 1.821712851524353,
+      "learning_rate": 1.616281367318712e-05,
+      "loss": 0.4087,
+      "step": 973
+    },
+    {
+      "epoch": 0.5771851851851851,
+      "grad_norm": 2.0716867446899414,
+      "learning_rate": 1.6158861885002966e-05,
+      "loss": 0.5099,
+      "step": 974
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 1.8543881177902222,
+      "learning_rate": 1.615491009681881e-05,
+      "loss": 0.4631,
+      "step": 975
+    },
+    {
+      "epoch": 0.5783703703703704,
+      "grad_norm": 1.9422781467437744,
+      "learning_rate": 1.615095830863466e-05,
+      "loss": 0.3649,
+      "step": 976
+    },
+    {
+      "epoch": 0.578962962962963,
+      "grad_norm": 1.2961667776107788,
+      "learning_rate": 1.6147006520450507e-05,
+      "loss": 0.3469,
+      "step": 977
+    },
+    {
+      "epoch": 0.5795555555555556,
+      "grad_norm": 1.8595051765441895,
+      "learning_rate": 1.614305473226635e-05,
+      "loss": 0.4326,
+      "step": 978
+    },
+    {
+      "epoch": 0.5801481481481482,
+      "grad_norm": 1.7076307535171509,
+      "learning_rate": 1.61391029440822e-05,
+      "loss": 0.4016,
+      "step": 979
+    },
+    {
+      "epoch": 0.5807407407407408,
+      "grad_norm": 2.022063970565796,
+      "learning_rate": 1.6135151155898047e-05,
+      "loss": 0.4617,
+      "step": 980
+    },
+    {
+      "epoch": 0.5813333333333334,
+      "grad_norm": 1.8688740730285645,
+      "learning_rate": 1.6131199367713892e-05,
+      "loss": 0.338,
+      "step": 981
+    },
+    {
+      "epoch": 0.581925925925926,
+      "grad_norm": 2.093066930770874,
+      "learning_rate": 1.612724757952974e-05,
+      "loss": 0.5282,
+      "step": 982
+    },
+    {
+      "epoch": 0.5825185185185185,
+      "grad_norm": 2.088667631149292,
+      "learning_rate": 1.6123295791345584e-05,
+      "loss": 0.5015,
+      "step": 983
+    },
+    {
+      "epoch": 0.5831111111111111,
+      "grad_norm": 1.793246865272522,
+      "learning_rate": 1.6119344003161432e-05,
+      "loss": 0.5023,
+      "step": 984
+    },
+    {
+      "epoch": 0.5837037037037037,
+      "grad_norm": 2.483407497406006,
+      "learning_rate": 1.611539221497728e-05,
+      "loss": 0.4678,
+      "step": 985
+    },
+    {
+      "epoch": 0.5842962962962963,
+      "grad_norm": 2.4411873817443848,
+      "learning_rate": 1.6111440426793125e-05,
+      "loss": 0.488,
+      "step": 986
+    },
+    {
+      "epoch": 0.5848888888888889,
+      "grad_norm": 2.807950496673584,
+      "learning_rate": 1.6107488638608973e-05,
+      "loss": 0.446,
+      "step": 987
+    },
+    {
+      "epoch": 0.5854814814814815,
+      "grad_norm": 1.8468725681304932,
+      "learning_rate": 1.610353685042482e-05,
+      "loss": 0.4478,
+      "step": 988
+    },
+    {
+      "epoch": 0.5860740740740741,
+      "grad_norm": 2.450983762741089,
+      "learning_rate": 1.6099585062240665e-05,
+      "loss": 0.419,
+      "step": 989
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 3.04799485206604,
+      "learning_rate": 1.6095633274056513e-05,
+      "loss": 0.5223,
+      "step": 990
+    },
+    {
+      "epoch": 0.5872592592592593,
+      "grad_norm": 3.5726261138916016,
+      "learning_rate": 1.6091681485872358e-05,
+      "loss": 0.3514,
+      "step": 991
+    },
+    {
+      "epoch": 0.5878518518518518,
+      "grad_norm": 3.0815589427948,
+      "learning_rate": 1.6087729697688206e-05,
+      "loss": 0.3171,
+      "step": 992
+    },
+    {
+      "epoch": 0.5884444444444444,
+      "grad_norm": 4.303046703338623,
+      "learning_rate": 1.6083777909504054e-05,
+      "loss": 0.5754,
+      "step": 993
+    },
+    {
+      "epoch": 0.589037037037037,
+      "grad_norm": 1.782226324081421,
+      "learning_rate": 1.6079826121319898e-05,
+      "loss": 0.401,
+      "step": 994
+    },
+    {
+      "epoch": 0.5896296296296296,
+      "grad_norm": 1.6652021408081055,
+      "learning_rate": 1.6075874333135746e-05,
+      "loss": 0.3685,
+      "step": 995
+    },
+    {
+      "epoch": 0.5902222222222222,
+      "grad_norm": 1.855122447013855,
+      "learning_rate": 1.6071922544951594e-05,
+      "loss": 0.352,
+      "step": 996
+    },
+    {
+      "epoch": 0.5908148148148148,
+      "grad_norm": 3.730327844619751,
+      "learning_rate": 1.606797075676744e-05,
+      "loss": 0.4934,
+      "step": 997
+    },
+    {
+      "epoch": 0.5914074074074074,
+      "grad_norm": 1.6755775213241577,
+      "learning_rate": 1.6064018968583286e-05,
+      "loss": 0.3735,
+      "step": 998
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 1.9290437698364258,
+      "learning_rate": 1.606006718039913e-05,
+      "loss": 0.3746,
+      "step": 999
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 2.3501226902008057,
+      "learning_rate": 1.605611539221498e-05,
+      "loss": 0.4525,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5931851851851851,
+      "grad_norm": 1.651233196258545,
+      "learning_rate": 1.6052163604030827e-05,
+      "loss": 0.3366,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5937777777777777,
+      "grad_norm": 2.8399407863616943,
+      "learning_rate": 1.604821181584667e-05,
+      "loss": 0.3947,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5943703703703703,
+      "grad_norm": 2.355510711669922,
+      "learning_rate": 1.604426002766252e-05,
+      "loss": 0.4229,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5949629629629629,
+      "grad_norm": 1.5570646524429321,
+      "learning_rate": 1.6040308239478367e-05,
+      "loss": 0.3656,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5955555555555555,
+      "grad_norm": 2.2392690181732178,
+      "learning_rate": 1.6036356451294212e-05,
+      "loss": 0.4136,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5961481481481481,
+      "grad_norm": 2.2981715202331543,
+      "learning_rate": 1.603240466311006e-05,
+      "loss": 0.4422,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5967407407407408,
+      "grad_norm": 3.1419951915740967,
+      "learning_rate": 1.6028452874925908e-05,
+      "loss": 0.4384,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5973333333333334,
+      "grad_norm": 2.0332517623901367,
+      "learning_rate": 1.6024501086741752e-05,
+      "loss": 0.3746,
+      "step": 1008
+    },
+    {
+      "epoch": 0.597925925925926,
+      "grad_norm": 1.61332106590271,
+      "learning_rate": 1.60205492985576e-05,
+      "loss": 0.3218,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5985185185185186,
+      "grad_norm": 1.4024579524993896,
+      "learning_rate": 1.6016597510373445e-05,
+      "loss": 0.3916,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5991111111111111,
+      "grad_norm": 2.709582805633545,
+      "learning_rate": 1.6012645722189293e-05,
+      "loss": 0.4309,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5997037037037037,
+      "grad_norm": 2.0822994709014893,
+      "learning_rate": 1.600869393400514e-05,
+      "loss": 0.4146,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6002962962962963,
+      "grad_norm": 2.1024348735809326,
+      "learning_rate": 1.6004742145820985e-05,
+      "loss": 0.3281,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6008888888888889,
+      "grad_norm": 2.5806567668914795,
+      "learning_rate": 1.6000790357636833e-05,
+      "loss": 0.5118,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6014814814814815,
+      "grad_norm": 2.113532543182373,
+      "learning_rate": 1.599683856945268e-05,
+      "loss": 0.4468,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6020740740740741,
+      "grad_norm": 1.4582295417785645,
+      "learning_rate": 1.5992886781268526e-05,
+      "loss": 0.334,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6026666666666667,
+      "grad_norm": 2.5854337215423584,
+      "learning_rate": 1.5988934993084373e-05,
+      "loss": 0.5272,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6032592592592593,
+      "grad_norm": 1.4680780172348022,
+      "learning_rate": 1.5984983204900218e-05,
+      "loss": 0.3235,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6038518518518519,
+      "grad_norm": 2.5218048095703125,
+      "learning_rate": 1.5981031416716066e-05,
+      "loss": 0.561,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6044444444444445,
+      "grad_norm": 2.1641554832458496,
+      "learning_rate": 1.5977079628531914e-05,
+      "loss": 0.481,
+      "step": 1020
+    },
+    {
+      "epoch": 0.605037037037037,
+      "grad_norm": 2.1650147438049316,
+      "learning_rate": 1.597312784034776e-05,
+      "loss": 0.3045,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6056296296296296,
+      "grad_norm": 2.313154458999634,
+      "learning_rate": 1.5969176052163606e-05,
+      "loss": 0.4377,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6062222222222222,
+      "grad_norm": 1.9279052019119263,
+      "learning_rate": 1.5965224263979454e-05,
+      "loss": 0.4462,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6068148148148148,
+      "grad_norm": 1.6668338775634766,
+      "learning_rate": 1.59612724757953e-05,
+      "loss": 0.3775,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6074074074074074,
+      "grad_norm": 2.8855206966400146,
+      "learning_rate": 1.5957320687611147e-05,
+      "loss": 0.4933,
+      "step": 1025
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 1.911860466003418,
+      "learning_rate": 1.595336889942699e-05,
+      "loss": 0.3806,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6085925925925926,
+      "grad_norm": 2.454017400741577,
+      "learning_rate": 1.594941711124284e-05,
+      "loss": 0.3455,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6091851851851852,
+      "grad_norm": 1.5567944049835205,
+      "learning_rate": 1.5945465323058687e-05,
+      "loss": 0.3395,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6097777777777778,
+      "grad_norm": 2.044701337814331,
+      "learning_rate": 1.5941513534874532e-05,
+      "loss": 0.4097,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6103703703703703,
+      "grad_norm": 2.0239996910095215,
+      "learning_rate": 1.593756174669038e-05,
+      "loss": 0.3508,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6109629629629629,
+      "grad_norm": 2.4736971855163574,
+      "learning_rate": 1.5933609958506224e-05,
+      "loss": 0.6342,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6115555555555555,
+      "grad_norm": 2.2801601886749268,
+      "learning_rate": 1.5929658170322072e-05,
+      "loss": 0.3539,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6121481481481481,
+      "grad_norm": 2.3134331703186035,
+      "learning_rate": 1.5925706382137917e-05,
+      "loss": 0.3776,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6127407407407407,
+      "grad_norm": 2.3597254753112793,
+      "learning_rate": 1.5921754593953765e-05,
+      "loss": 0.4637,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6133333333333333,
+      "grad_norm": 2.1071596145629883,
+      "learning_rate": 1.5917802805769613e-05,
+      "loss": 0.3753,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6139259259259259,
+      "grad_norm": 2.146023750305176,
+      "learning_rate": 1.5913851017585457e-05,
+      "loss": 0.3586,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6145185185185185,
+      "grad_norm": 2.8998677730560303,
+      "learning_rate": 1.5909899229401305e-05,
+      "loss": 0.4699,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6151111111111112,
+      "grad_norm": 2.451887607574463,
+      "learning_rate": 1.590594744121715e-05,
+      "loss": 0.4468,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6157037037037038,
+      "grad_norm": 3.0691041946411133,
+      "learning_rate": 1.5901995653032998e-05,
+      "loss": 0.3606,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6162962962962963,
+      "grad_norm": 1.8998016119003296,
+      "learning_rate": 1.5898043864848846e-05,
+      "loss": 0.3775,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6168888888888889,
+      "grad_norm": 2.193601369857788,
+      "learning_rate": 1.589409207666469e-05,
+      "loss": 0.4854,
+      "step": 1041
+    },
+    {
+      "epoch": 0.6174814814814815,
+      "grad_norm": 2.230663299560547,
+      "learning_rate": 1.5890140288480538e-05,
+      "loss": 0.421,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6180740740740741,
+      "grad_norm": 2.075199842453003,
+      "learning_rate": 1.5886188500296386e-05,
+      "loss": 0.5271,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6186666666666667,
+      "grad_norm": 3.3343896865844727,
+      "learning_rate": 1.588223671211223e-05,
+      "loss": 0.4958,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6192592592592593,
+      "grad_norm": 2.418895721435547,
+      "learning_rate": 1.587828492392808e-05,
+      "loss": 0.3671,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6198518518518519,
+      "grad_norm": 2.1218631267547607,
+      "learning_rate": 1.5874333135743923e-05,
+      "loss": 0.4651,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6204444444444445,
+      "grad_norm": 3.1008028984069824,
+      "learning_rate": 1.587038134755977e-05,
+      "loss": 0.4365,
+      "step": 1047
+    },
+    {
+      "epoch": 0.621037037037037,
+      "grad_norm": 2.5666632652282715,
+      "learning_rate": 1.586642955937562e-05,
+      "loss": 0.5899,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6216296296296296,
+      "grad_norm": 1.7907891273498535,
+      "learning_rate": 1.5862477771191463e-05,
+      "loss": 0.473,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.5533769130706787,
+      "learning_rate": 1.585852598300731e-05,
+      "loss": 0.3811,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6228148148148148,
+      "grad_norm": 1.6892316341400146,
+      "learning_rate": 1.585457419482316e-05,
+      "loss": 0.4222,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6234074074074074,
+      "grad_norm": 1.5321599245071411,
+      "learning_rate": 1.5850622406639004e-05,
+      "loss": 0.3612,
+      "step": 1052
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 1.897282600402832,
+      "learning_rate": 1.5846670618454852e-05,
+      "loss": 0.4697,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6245925925925926,
+      "grad_norm": 2.3186960220336914,
+      "learning_rate": 1.5842718830270696e-05,
+      "loss": 0.4638,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6251851851851852,
+      "grad_norm": 4.939927577972412,
+      "learning_rate": 1.5838767042086544e-05,
+      "loss": 0.5738,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6257777777777778,
+      "grad_norm": 1.9265767335891724,
+      "learning_rate": 1.5834815253902392e-05,
+      "loss": 0.4011,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6263703703703704,
+      "grad_norm": 1.9809333086013794,
+      "learning_rate": 1.5830863465718237e-05,
+      "loss": 0.3672,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6269629629629629,
+      "grad_norm": 1.8061078786849976,
+      "learning_rate": 1.5826911677534085e-05,
+      "loss": 0.3315,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6275555555555555,
+      "grad_norm": 2.196333646774292,
+      "learning_rate": 1.5822959889349933e-05,
+      "loss": 0.3773,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6281481481481481,
+      "grad_norm": 5.529025077819824,
+      "learning_rate": 1.5819008101165777e-05,
+      "loss": 0.5211,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6287407407407407,
+      "grad_norm": 2.291978597640991,
+      "learning_rate": 1.5815056312981625e-05,
+      "loss": 0.3353,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6293333333333333,
+      "grad_norm": 2.951303720474243,
+      "learning_rate": 1.5811104524797473e-05,
+      "loss": 0.4693,
+      "step": 1062
+    },
+    {
+      "epoch": 0.6299259259259259,
+      "grad_norm": 1.8504918813705444,
+      "learning_rate": 1.5807152736613318e-05,
+      "loss": 0.3286,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6305185185185185,
+      "grad_norm": 1.9202356338500977,
+      "learning_rate": 1.5803200948429166e-05,
+      "loss": 0.4182,
+      "step": 1064
+    },
+    {
+      "epoch": 0.6311111111111111,
+      "grad_norm": 2.3560216426849365,
+      "learning_rate": 1.579924916024501e-05,
+      "loss": 0.3695,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6317037037037037,
+      "grad_norm": 2.4557852745056152,
+      "learning_rate": 1.5795297372060858e-05,
+      "loss": 0.426,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6322962962962962,
+      "grad_norm": 2.5103859901428223,
+      "learning_rate": 1.5791345583876706e-05,
+      "loss": 0.3428,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6328888888888888,
+      "grad_norm": 2.765662431716919,
+      "learning_rate": 1.578739379569255e-05,
+      "loss": 0.5669,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6334814814814815,
+      "grad_norm": 8.941581726074219,
+      "learning_rate": 1.57834420075084e-05,
+      "loss": 0.99,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6340740740740741,
+      "grad_norm": 1.8447152376174927,
+      "learning_rate": 1.5779490219324246e-05,
+      "loss": 0.4568,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6346666666666667,
+      "grad_norm": 2.554399251937866,
+      "learning_rate": 1.577553843114009e-05,
+      "loss": 0.5223,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6352592592592593,
+      "grad_norm": 2.1146976947784424,
+      "learning_rate": 1.577158664295594e-05,
+      "loss": 0.4348,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6358518518518519,
+      "grad_norm": 1.583416223526001,
+      "learning_rate": 1.5767634854771783e-05,
+      "loss": 0.4776,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6364444444444445,
+      "grad_norm": 1.8910413980484009,
+      "learning_rate": 1.576368306658763e-05,
+      "loss": 0.3639,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6370370370370371,
+      "grad_norm": 1.782350778579712,
+      "learning_rate": 1.575973127840348e-05,
+      "loss": 0.4033,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6376296296296297,
+      "grad_norm": 3.338207960128784,
+      "learning_rate": 1.5755779490219324e-05,
+      "loss": 0.5343,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6382222222222222,
+      "grad_norm": 2.0684547424316406,
+      "learning_rate": 1.5751827702035172e-05,
+      "loss": 0.4015,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6388148148148148,
+      "grad_norm": 1.6251686811447144,
+      "learning_rate": 1.574787591385102e-05,
+      "loss": 0.3875,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6394074074074074,
+      "grad_norm": 2.1981751918792725,
+      "learning_rate": 1.5743924125666864e-05,
+      "loss": 0.5746,
+      "step": 1079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9810956716537476,
+      "learning_rate": 1.5739972337482712e-05,
+      "loss": 0.3263,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6405925925925926,
+      "grad_norm": 2.525712251663208,
+      "learning_rate": 1.573602054929856e-05,
+      "loss": 0.473,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6411851851851852,
+      "grad_norm": 2.604447603225708,
+      "learning_rate": 1.5732068761114405e-05,
+      "loss": 0.5555,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6417777777777778,
+      "grad_norm": 5.761559963226318,
+      "learning_rate": 1.5728116972930253e-05,
+      "loss": 0.5919,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6423703703703704,
+      "grad_norm": 2.0423684120178223,
+      "learning_rate": 1.5724165184746097e-05,
+      "loss": 0.4264,
+      "step": 1084
+    },
+    {
+      "epoch": 0.642962962962963,
+      "grad_norm": 3.1846377849578857,
+      "learning_rate": 1.5720213396561945e-05,
+      "loss": 0.4113,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6435555555555555,
+      "grad_norm": 3.3368146419525146,
+      "learning_rate": 1.5716261608377793e-05,
+      "loss": 0.3825,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6441481481481481,
+      "grad_norm": 4.279855251312256,
+      "learning_rate": 1.5712309820193638e-05,
+      "loss": 0.4422,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6447407407407407,
+      "grad_norm": 1.7295254468917847,
+      "learning_rate": 1.5708358032009486e-05,
+      "loss": 0.3045,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6453333333333333,
+      "grad_norm": 1.973081350326538,
+      "learning_rate": 1.5704406243825333e-05,
+      "loss": 0.3619,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6459259259259259,
+      "grad_norm": 2.6521196365356445,
+      "learning_rate": 1.5700454455641178e-05,
+      "loss": 0.4312,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6465185185185185,
+      "grad_norm": 1.6793878078460693,
+      "learning_rate": 1.5696502667457026e-05,
+      "loss": 0.3061,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6471111111111111,
+      "grad_norm": 1.7885286808013916,
+      "learning_rate": 1.569255087927287e-05,
+      "loss": 0.4103,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6477037037037037,
+      "grad_norm": 2.7680506706237793,
+      "learning_rate": 1.568859909108872e-05,
+      "loss": 0.5292,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6482962962962963,
+      "grad_norm": 1.672656536102295,
+      "learning_rate": 1.5684647302904566e-05,
+      "loss": 0.4083,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6488888888888888,
+      "grad_norm": 3.4401204586029053,
+      "learning_rate": 1.568069551472041e-05,
+      "loss": 0.3002,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6494814814814814,
+      "grad_norm": 3.231678009033203,
+      "learning_rate": 1.567674372653626e-05,
+      "loss": 0.3649,
+      "step": 1096
+    },
+    {
+      "epoch": 0.650074074074074,
+      "grad_norm": 1.86122727394104,
+      "learning_rate": 1.5672791938352107e-05,
+      "loss": 0.3617,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6506666666666666,
+      "grad_norm": 1.785387635231018,
+      "learning_rate": 1.566884015016795e-05,
+      "loss": 0.3935,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6512592592592592,
+      "grad_norm": 1.7261079549789429,
+      "learning_rate": 1.56648883619838e-05,
+      "loss": 0.4015,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6518518518518519,
+      "grad_norm": 3.2074809074401855,
+      "learning_rate": 1.5660936573799644e-05,
+      "loss": 0.3982,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6524444444444445,
+      "grad_norm": 1.6235488653182983,
+      "learning_rate": 1.5656984785615492e-05,
+      "loss": 0.4122,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6530370370370371,
+      "grad_norm": 2.7631003856658936,
+      "learning_rate": 1.565303299743134e-05,
+      "loss": 0.3793,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6536296296296297,
+      "grad_norm": 4.715127468109131,
+      "learning_rate": 1.5649081209247184e-05,
+      "loss": 0.5169,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6542222222222223,
+      "grad_norm": 4.173917293548584,
+      "learning_rate": 1.5645129421063032e-05,
+      "loss": 0.6277,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6548148148148148,
+      "grad_norm": 2.030864953994751,
+      "learning_rate": 1.564117763287888e-05,
+      "loss": 0.3334,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6554074074074074,
+      "grad_norm": 3.305692672729492,
+      "learning_rate": 1.5637225844694725e-05,
+      "loss": 0.4595,
+      "step": 1106
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 1.7764016389846802,
+      "learning_rate": 1.5633274056510573e-05,
+      "loss": 0.377,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6565925925925926,
+      "grad_norm": 2.6208419799804688,
+      "learning_rate": 1.562932226832642e-05,
+      "loss": 0.5144,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6571851851851852,
+      "grad_norm": 2.180157423019409,
+      "learning_rate": 1.5625370480142265e-05,
+      "loss": 0.4128,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6577777777777778,
+      "grad_norm": 2.562849998474121,
+      "learning_rate": 1.5621418691958113e-05,
+      "loss": 0.3939,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6583703703703704,
+      "grad_norm": 1.825573205947876,
+      "learning_rate": 1.5617466903773958e-05,
+      "loss": 0.369,
+      "step": 1111
+    },
+    {
+      "epoch": 0.658962962962963,
+      "grad_norm": 1.623281717300415,
+      "learning_rate": 1.5613515115589806e-05,
+      "loss": 0.3812,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6595555555555556,
+      "grad_norm": 1.70048189163208,
+      "learning_rate": 1.5609563327405653e-05,
+      "loss": 0.4769,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6601481481481482,
+      "grad_norm": 1.7525659799575806,
+      "learning_rate": 1.5605611539221498e-05,
+      "loss": 0.3426,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6607407407407407,
+      "grad_norm": 2.0540261268615723,
+      "learning_rate": 1.5601659751037346e-05,
+      "loss": 0.4705,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6613333333333333,
+      "grad_norm": 1.4061357975006104,
+      "learning_rate": 1.5597707962853194e-05,
+      "loss": 0.2656,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6619259259259259,
+      "grad_norm": 2.2261526584625244,
+      "learning_rate": 1.559375617466904e-05,
+      "loss": 0.4043,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6625185185185185,
+      "grad_norm": 1.9714399576187134,
+      "learning_rate": 1.5589804386484886e-05,
+      "loss": 0.4944,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6631111111111111,
+      "grad_norm": 1.6842048168182373,
+      "learning_rate": 1.558585259830073e-05,
+      "loss": 0.5045,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6637037037037037,
+      "grad_norm": 2.9780845642089844,
+      "learning_rate": 1.558190081011658e-05,
+      "loss": 0.3085,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6642962962962963,
+      "grad_norm": 1.7498419284820557,
+      "learning_rate": 1.5577949021932427e-05,
+      "loss": 0.3993,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6648888888888889,
+      "grad_norm": 1.9348454475402832,
+      "learning_rate": 1.557399723374827e-05,
+      "loss": 0.3619,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6654814814814815,
+      "grad_norm": 1.8367990255355835,
+      "learning_rate": 1.557004544556412e-05,
+      "loss": 0.3894,
+      "step": 1123
+    },
+    {
+      "epoch": 0.666074074074074,
+      "grad_norm": 2.3883469104766846,
+      "learning_rate": 1.5566093657379967e-05,
+      "loss": 0.4644,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1.843077301979065,
+      "learning_rate": 1.5562141869195812e-05,
+      "loss": 0.3142,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6672592592592592,
+      "grad_norm": 3.1021127700805664,
+      "learning_rate": 1.555819008101166e-05,
+      "loss": 0.478,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6678518518518518,
+      "grad_norm": 1.94724702835083,
+      "learning_rate": 1.5554238292827508e-05,
+      "loss": 0.3905,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6684444444444444,
+      "grad_norm": 3.4576826095581055,
+      "learning_rate": 1.5550286504643352e-05,
+      "loss": 0.5502,
+      "step": 1128
+    },
+    {
+      "epoch": 0.669037037037037,
+      "grad_norm": 3.1009204387664795,
+      "learning_rate": 1.55463347164592e-05,
+      "loss": 0.4108,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6696296296296296,
+      "grad_norm": 4.121009826660156,
+      "learning_rate": 1.5542382928275045e-05,
+      "loss": 0.3726,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6702222222222223,
+      "grad_norm": 2.669243335723877,
+      "learning_rate": 1.5538431140090893e-05,
+      "loss": 0.3865,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6708148148148149,
+      "grad_norm": 3.084313154220581,
+      "learning_rate": 1.553447935190674e-05,
+      "loss": 0.5991,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6714074074074075,
+      "grad_norm": 2.900257110595703,
+      "learning_rate": 1.5530527563722585e-05,
+      "loss": 0.5456,
+      "step": 1133
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 1.8862411975860596,
+      "learning_rate": 1.5526575775538433e-05,
+      "loss": 0.4482,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6725925925925926,
+      "grad_norm": 2.12424635887146,
+      "learning_rate": 1.552262398735428e-05,
+      "loss": 0.4305,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6731851851851852,
+      "grad_norm": 2.0344738960266113,
+      "learning_rate": 1.5518672199170126e-05,
+      "loss": 0.3208,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6737777777777778,
+      "grad_norm": 2.806122064590454,
+      "learning_rate": 1.5514720410985973e-05,
+      "loss": 0.3757,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6743703703703704,
+      "grad_norm": 2.102069139480591,
+      "learning_rate": 1.5510768622801818e-05,
+      "loss": 0.4225,
+      "step": 1138
+    },
+    {
+      "epoch": 0.674962962962963,
+      "grad_norm": 1.7395190000534058,
+      "learning_rate": 1.5506816834617666e-05,
+      "loss": 0.3464,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6755555555555556,
+      "grad_norm": 1.587647557258606,
+      "learning_rate": 1.5502865046433514e-05,
+      "loss": 0.3272,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6761481481481482,
+      "grad_norm": 1.3612326383590698,
+      "learning_rate": 1.549891325824936e-05,
+      "loss": 0.3688,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6767407407407408,
+      "grad_norm": 2.000469923019409,
+      "learning_rate": 1.5494961470065206e-05,
+      "loss": 0.382,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6773333333333333,
+      "grad_norm": 1.6368337869644165,
+      "learning_rate": 1.5491009681881054e-05,
+      "loss": 0.4465,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6779259259259259,
+      "grad_norm": 2.1554293632507324,
+      "learning_rate": 1.54870578936969e-05,
+      "loss": 0.4532,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6785185185185185,
+      "grad_norm": 4.802222728729248,
+      "learning_rate": 1.5483106105512747e-05,
+      "loss": 0.5016,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6791111111111111,
+      "grad_norm": 1.5911790132522583,
+      "learning_rate": 1.547915431732859e-05,
+      "loss": 0.4104,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6797037037037037,
+      "grad_norm": 2.179105520248413,
+      "learning_rate": 1.547520252914444e-05,
+      "loss": 0.529,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6802962962962963,
+      "grad_norm": 1.9075478315353394,
+      "learning_rate": 1.5471250740960287e-05,
+      "loss": 0.3623,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6808888888888889,
+      "grad_norm": 1.5067347288131714,
+      "learning_rate": 1.5467298952776132e-05,
+      "loss": 0.3236,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6814814814814815,
+      "grad_norm": 1.9329344034194946,
+      "learning_rate": 1.546334716459198e-05,
+      "loss": 0.451,
+      "step": 1150
+    },
+    {
+      "epoch": 0.682074074074074,
+      "grad_norm": 1.7222490310668945,
+      "learning_rate": 1.5459395376407828e-05,
+      "loss": 0.4441,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6826666666666666,
+      "grad_norm": 1.7494182586669922,
+      "learning_rate": 1.5455443588223672e-05,
+      "loss": 0.4424,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6832592592592592,
+      "grad_norm": 1.848321795463562,
+      "learning_rate": 1.545149180003952e-05,
+      "loss": 0.399,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6838518518518518,
+      "grad_norm": 2.2331390380859375,
+      "learning_rate": 1.5447540011855368e-05,
+      "loss": 0.4442,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6844444444444444,
+      "grad_norm": 1.7956246137619019,
+      "learning_rate": 1.5443588223671213e-05,
+      "loss": 0.3284,
+      "step": 1155
+    },
+    {
+      "epoch": 0.685037037037037,
+      "grad_norm": 2.702223777770996,
+      "learning_rate": 1.543963643548706e-05,
+      "loss": 0.6332,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6856296296296296,
+      "grad_norm": 1.6313655376434326,
+      "learning_rate": 1.5435684647302905e-05,
+      "loss": 0.3442,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6862222222222222,
+      "grad_norm": 1.6391613483428955,
+      "learning_rate": 1.5431732859118753e-05,
+      "loss": 0.3281,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6868148148148148,
+      "grad_norm": 7.212156772613525,
+      "learning_rate": 1.54277810709346e-05,
+      "loss": 0.6491,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6874074074074074,
+      "grad_norm": 2.6900265216827393,
+      "learning_rate": 1.5423829282750446e-05,
+      "loss": 0.337,
+      "step": 1160
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 2.1173434257507324,
+      "learning_rate": 1.5419877494566293e-05,
+      "loss": 0.3594,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6885925925925926,
+      "grad_norm": 2.7620763778686523,
+      "learning_rate": 1.541592570638214e-05,
+      "loss": 0.4183,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6891851851851852,
+      "grad_norm": 1.7106410264968872,
+      "learning_rate": 1.5411973918197986e-05,
+      "loss": 0.444,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6897777777777778,
+      "grad_norm": 1.852292776107788,
+      "learning_rate": 1.5408022130013834e-05,
+      "loss": 0.3086,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6903703703703704,
+      "grad_norm": 2.412928819656372,
+      "learning_rate": 1.540407034182968e-05,
+      "loss": 0.3636,
+      "step": 1165
+    },
+    {
+      "epoch": 0.690962962962963,
+      "grad_norm": 1.9604030847549438,
+      "learning_rate": 1.5400118553645526e-05,
+      "loss": 0.389,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6915555555555556,
+      "grad_norm": 2.125000238418579,
+      "learning_rate": 1.5396166765461374e-05,
+      "loss": 0.3651,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6921481481481482,
+      "grad_norm": 2.0500802993774414,
+      "learning_rate": 1.539221497727722e-05,
+      "loss": 0.4391,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6927407407407408,
+      "grad_norm": 2.957080364227295,
+      "learning_rate": 1.5388263189093067e-05,
+      "loss": 0.5129,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "grad_norm": 1.9725420475006104,
+      "learning_rate": 1.5384311400908915e-05,
+      "loss": 0.363,
+      "step": 1170
+    },
+    {
+      "epoch": 0.693925925925926,
+      "grad_norm": 1.9165403842926025,
+      "learning_rate": 1.538035961272476e-05,
+      "loss": 0.3369,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6945185185185185,
+      "grad_norm": 2.2486982345581055,
+      "learning_rate": 1.5376407824540607e-05,
+      "loss": 0.3932,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6951111111111111,
+      "grad_norm": 2.63761568069458,
+      "learning_rate": 1.5372456036356452e-05,
+      "loss": 0.4001,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6957037037037037,
+      "grad_norm": 2.6167306900024414,
+      "learning_rate": 1.53685042481723e-05,
+      "loss": 0.4832,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6962962962962963,
+      "grad_norm": 1.4735889434814453,
+      "learning_rate": 1.5364552459988148e-05,
+      "loss": 0.294,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6968888888888889,
+      "grad_norm": 1.9430512189865112,
+      "learning_rate": 1.5360600671803992e-05,
+      "loss": 0.4086,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6974814814814815,
+      "grad_norm": 1.7597235441207886,
+      "learning_rate": 1.535664888361984e-05,
+      "loss": 0.3829,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6980740740740741,
+      "grad_norm": 1.6026808023452759,
+      "learning_rate": 1.5352697095435688e-05,
+      "loss": 0.3232,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6986666666666667,
+      "grad_norm": 2.097141742706299,
+      "learning_rate": 1.5348745307251533e-05,
+      "loss": 0.3069,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6992592592592592,
+      "grad_norm": 2.8541057109832764,
+      "learning_rate": 1.534479351906738e-05,
+      "loss": 0.4537,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6998518518518518,
+      "grad_norm": 1.5406564474105835,
+      "learning_rate": 1.534084173088323e-05,
+      "loss": 0.3477,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7004444444444444,
+      "grad_norm": 1.8281196355819702,
+      "learning_rate": 1.5336889942699073e-05,
+      "loss": 0.3793,
+      "step": 1182
+    },
+    {
+      "epoch": 0.701037037037037,
+      "grad_norm": 2.2892909049987793,
+      "learning_rate": 1.533293815451492e-05,
+      "loss": 0.427,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7016296296296296,
+      "grad_norm": 3.0731282234191895,
+      "learning_rate": 1.5328986366330766e-05,
+      "loss": 0.5302,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7022222222222222,
+      "grad_norm": 1.8328542709350586,
+      "learning_rate": 1.5325034578146613e-05,
+      "loss": 0.2545,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7028148148148148,
+      "grad_norm": 1.4272211790084839,
+      "learning_rate": 1.532108278996246e-05,
+      "loss": 0.2645,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7034074074074074,
+      "grad_norm": 2.753033399581909,
+      "learning_rate": 1.5317131001778306e-05,
+      "loss": 0.358,
+      "step": 1187
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 2.563620090484619,
+      "learning_rate": 1.5313179213594154e-05,
+      "loss": 0.3793,
+      "step": 1188
+    },
+    {
+      "epoch": 0.7045925925925925,
+      "grad_norm": 2.600966215133667,
+      "learning_rate": 1.530922742541e-05,
+      "loss": 0.4868,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7051851851851851,
+      "grad_norm": 3.00221848487854,
+      "learning_rate": 1.5305275637225846e-05,
+      "loss": 0.3817,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7057777777777777,
+      "grad_norm": 2.659513473510742,
+      "learning_rate": 1.530132384904169e-05,
+      "loss": 0.3492,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7063703703703703,
+      "grad_norm": 3.039992570877075,
+      "learning_rate": 1.529737206085754e-05,
+      "loss": 0.3734,
+      "step": 1192
+    },
+    {
+      "epoch": 0.706962962962963,
+      "grad_norm": 2.263066530227661,
+      "learning_rate": 1.5293420272673383e-05,
+      "loss": 0.3978,
+      "step": 1193
+    },
+    {
+      "epoch": 0.7075555555555556,
+      "grad_norm": 2.2073514461517334,
+      "learning_rate": 1.528946848448923e-05,
+      "loss": 0.3058,
+      "step": 1194
+    },
+    {
+      "epoch": 0.7081481481481482,
+      "grad_norm": 2.1258022785186768,
+      "learning_rate": 1.528551669630508e-05,
+      "loss": 0.4179,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7087407407407408,
+      "grad_norm": 2.255652666091919,
+      "learning_rate": 1.5281564908120924e-05,
+      "loss": 0.3927,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7093333333333334,
+      "grad_norm": 1.7002241611480713,
+      "learning_rate": 1.5277613119936772e-05,
+      "loss": 0.4469,
+      "step": 1197
+    },
+    {
+      "epoch": 0.709925925925926,
+      "grad_norm": 2.5146493911743164,
+      "learning_rate": 1.527366133175262e-05,
+      "loss": 0.4806,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7105185185185185,
+      "grad_norm": 1.8832653760910034,
+      "learning_rate": 1.5269709543568464e-05,
+      "loss": 0.358,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.9547940492630005,
+      "learning_rate": 1.5265757755384312e-05,
+      "loss": 0.3899,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7117037037037037,
+      "grad_norm": 1.7744343280792236,
+      "learning_rate": 1.5261805967200157e-05,
+      "loss": 0.3717,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7122962962962963,
+      "grad_norm": 2.128002882003784,
+      "learning_rate": 1.5257854179016005e-05,
+      "loss": 0.3607,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7128888888888889,
+      "grad_norm": 2.0537898540496826,
+      "learning_rate": 1.5253902390831853e-05,
+      "loss": 0.3935,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7134814814814815,
+      "grad_norm": 1.5756819248199463,
+      "learning_rate": 1.5249950602647699e-05,
+      "loss": 0.3601,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7140740740740741,
+      "grad_norm": 3.0168850421905518,
+      "learning_rate": 1.5245998814463545e-05,
+      "loss": 0.5012,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7146666666666667,
+      "grad_norm": 1.4763816595077515,
+      "learning_rate": 1.5242047026279391e-05,
+      "loss": 0.256,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7152592592592593,
+      "grad_norm": 3.06215238571167,
+      "learning_rate": 1.523809523809524e-05,
+      "loss": 0.3661,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7158518518518519,
+      "grad_norm": 2.0885403156280518,
+      "learning_rate": 1.5234143449911086e-05,
+      "loss": 0.3546,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7164444444444444,
+      "grad_norm": 2.8130972385406494,
+      "learning_rate": 1.5230191661726932e-05,
+      "loss": 0.4484,
+      "step": 1209
+    },
+    {
+      "epoch": 0.717037037037037,
+      "grad_norm": 2.6240174770355225,
+      "learning_rate": 1.5226239873542778e-05,
+      "loss": 0.4531,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7176296296296296,
+      "grad_norm": 1.8575855493545532,
+      "learning_rate": 1.5222288085358626e-05,
+      "loss": 0.3524,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7182222222222222,
+      "grad_norm": 1.8862888813018799,
+      "learning_rate": 1.5218336297174472e-05,
+      "loss": 0.3506,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7188148148148148,
+      "grad_norm": 1.91115403175354,
+      "learning_rate": 1.5214384508990318e-05,
+      "loss": 0.4045,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7194074074074074,
+      "grad_norm": 2.225682020187378,
+      "learning_rate": 1.5210432720806165e-05,
+      "loss": 0.4208,
+      "step": 1214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4980266094207764,
+      "learning_rate": 1.5206480932622013e-05,
+      "loss": 0.3674,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7205925925925926,
+      "grad_norm": 2.0818586349487305,
+      "learning_rate": 1.5202529144437859e-05,
+      "loss": 0.3327,
+      "step": 1216
+    },
+    {
+      "epoch": 0.7211851851851852,
+      "grad_norm": 2.1761889457702637,
+      "learning_rate": 1.5198577356253705e-05,
+      "loss": 0.5044,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7217777777777777,
+      "grad_norm": 2.0202202796936035,
+      "learning_rate": 1.5194625568069551e-05,
+      "loss": 0.4189,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7223703703703703,
+      "grad_norm": 2.988250494003296,
+      "learning_rate": 1.51906737798854e-05,
+      "loss": 0.4234,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7229629629629629,
+      "grad_norm": 2.1585991382598877,
+      "learning_rate": 1.5186721991701246e-05,
+      "loss": 0.3938,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7235555555555555,
+      "grad_norm": 2.328098773956299,
+      "learning_rate": 1.5182770203517092e-05,
+      "loss": 0.3953,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7241481481481481,
+      "grad_norm": 2.783054828643799,
+      "learning_rate": 1.5178818415332938e-05,
+      "loss": 0.4087,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7247407407407407,
+      "grad_norm": 1.9576525688171387,
+      "learning_rate": 1.5174866627148786e-05,
+      "loss": 0.4163,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7253333333333334,
+      "grad_norm": 1.5030555725097656,
+      "learning_rate": 1.5170914838964632e-05,
+      "loss": 0.2838,
+      "step": 1224
+    },
+    {
+      "epoch": 0.725925925925926,
+      "grad_norm": 2.4365127086639404,
+      "learning_rate": 1.5166963050780478e-05,
+      "loss": 0.4011,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7265185185185186,
+      "grad_norm": 1.7001701593399048,
+      "learning_rate": 1.5163011262596325e-05,
+      "loss": 0.4016,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7271111111111112,
+      "grad_norm": 2.0497398376464844,
+      "learning_rate": 1.5159059474412173e-05,
+      "loss": 0.2331,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7277037037037037,
+      "grad_norm": 3.244722604751587,
+      "learning_rate": 1.5155107686228019e-05,
+      "loss": 0.6324,
+      "step": 1228
+    },
+    {
+      "epoch": 0.7282962962962963,
+      "grad_norm": 1.4487639665603638,
+      "learning_rate": 1.5151155898043865e-05,
+      "loss": 0.2991,
+      "step": 1229
+    },
+    {
+      "epoch": 0.7288888888888889,
+      "grad_norm": 1.9585906267166138,
+      "learning_rate": 1.5147204109859713e-05,
+      "loss": 0.365,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7294814814814815,
+      "grad_norm": 2.259113073348999,
+      "learning_rate": 1.514325232167556e-05,
+      "loss": 0.4207,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7300740740740741,
+      "grad_norm": 1.834892749786377,
+      "learning_rate": 1.5139300533491405e-05,
+      "loss": 0.3605,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7306666666666667,
+      "grad_norm": 1.9725133180618286,
+      "learning_rate": 1.5135348745307252e-05,
+      "loss": 0.3525,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7312592592592593,
+      "grad_norm": 2.8778984546661377,
+      "learning_rate": 1.51313969571231e-05,
+      "loss": 0.4882,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7318518518518519,
+      "grad_norm": 1.6605966091156006,
+      "learning_rate": 1.5127445168938946e-05,
+      "loss": 0.252,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7324444444444445,
+      "grad_norm": 2.2206215858459473,
+      "learning_rate": 1.5123493380754792e-05,
+      "loss": 0.4048,
+      "step": 1236
+    },
+    {
+      "epoch": 0.733037037037037,
+      "grad_norm": 2.1556742191314697,
+      "learning_rate": 1.5119541592570638e-05,
+      "loss": 0.4589,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7336296296296296,
+      "grad_norm": 2.425645112991333,
+      "learning_rate": 1.5115589804386486e-05,
+      "loss": 0.3648,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7342222222222222,
+      "grad_norm": 2.053961992263794,
+      "learning_rate": 1.5111638016202333e-05,
+      "loss": 0.4509,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7348148148148148,
+      "grad_norm": 1.8301212787628174,
+      "learning_rate": 1.5107686228018179e-05,
+      "loss": 0.4851,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7354074074074074,
+      "grad_norm": 1.7090506553649902,
+      "learning_rate": 1.5103734439834025e-05,
+      "loss": 0.2933,
+      "step": 1241
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 1.6780797243118286,
+      "learning_rate": 1.5099782651649873e-05,
+      "loss": 0.3511,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7365925925925926,
+      "grad_norm": 1.93254816532135,
+      "learning_rate": 1.509583086346572e-05,
+      "loss": 0.439,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7371851851851852,
+      "grad_norm": 2.001206159591675,
+      "learning_rate": 1.5091879075281565e-05,
+      "loss": 0.3402,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7377777777777778,
+      "grad_norm": 2.058063268661499,
+      "learning_rate": 1.5087927287097412e-05,
+      "loss": 0.4402,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7383703703703703,
+      "grad_norm": 1.7721507549285889,
+      "learning_rate": 1.508397549891326e-05,
+      "loss": 0.3237,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7389629629629629,
+      "grad_norm": 2.8502421379089355,
+      "learning_rate": 1.5080023710729106e-05,
+      "loss": 0.3666,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7395555555555555,
+      "grad_norm": 2.123032569885254,
+      "learning_rate": 1.5076071922544952e-05,
+      "loss": 0.326,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7401481481481481,
+      "grad_norm": 3.407352924346924,
+      "learning_rate": 1.5072120134360798e-05,
+      "loss": 0.5821,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 2.022691011428833,
+      "learning_rate": 1.5068168346176646e-05,
+      "loss": 0.3207,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7413333333333333,
+      "grad_norm": 1.5910940170288086,
+      "learning_rate": 1.5064216557992493e-05,
+      "loss": 0.3848,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7419259259259259,
+      "grad_norm": 2.4330666065216064,
+      "learning_rate": 1.5060264769808339e-05,
+      "loss": 0.4963,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7425185185185185,
+      "grad_norm": 2.063366651535034,
+      "learning_rate": 1.5056312981624187e-05,
+      "loss": 0.3991,
+      "step": 1253
+    },
+    {
+      "epoch": 0.7431111111111111,
+      "grad_norm": 2.6412012577056885,
+      "learning_rate": 1.5052361193440033e-05,
+      "loss": 0.4931,
+      "step": 1254
+    },
+    {
+      "epoch": 0.7437037037037038,
+      "grad_norm": 1.8488892316818237,
+      "learning_rate": 1.504840940525588e-05,
+      "loss": 0.4491,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7442962962962963,
+      "grad_norm": 2.62813663482666,
+      "learning_rate": 1.5044457617071725e-05,
+      "loss": 0.4069,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7448888888888889,
+      "grad_norm": 1.5832606554031372,
+      "learning_rate": 1.5040505828887573e-05,
+      "loss": 0.3457,
+      "step": 1257
+    },
+    {
+      "epoch": 0.7454814814814815,
+      "grad_norm": 1.821006417274475,
+      "learning_rate": 1.503655404070342e-05,
+      "loss": 0.4409,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7460740740740741,
+      "grad_norm": 1.4453284740447998,
+      "learning_rate": 1.5032602252519266e-05,
+      "loss": 0.3527,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "grad_norm": 1.8654431104660034,
+      "learning_rate": 1.5028650464335112e-05,
+      "loss": 0.3686,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7472592592592593,
+      "grad_norm": 2.162553071975708,
+      "learning_rate": 1.502469867615096e-05,
+      "loss": 0.4383,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7478518518518519,
+      "grad_norm": 2.2929067611694336,
+      "learning_rate": 1.5020746887966806e-05,
+      "loss": 0.3218,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7484444444444445,
+      "grad_norm": 1.5168490409851074,
+      "learning_rate": 1.5016795099782653e-05,
+      "loss": 0.3277,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7490370370370371,
+      "grad_norm": 1.6369075775146484,
+      "learning_rate": 1.5012843311598499e-05,
+      "loss": 0.3409,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7496296296296296,
+      "grad_norm": 3.744297742843628,
+      "learning_rate": 1.5008891523414347e-05,
+      "loss": 0.4812,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7502222222222222,
+      "grad_norm": 1.7838548421859741,
+      "learning_rate": 1.5004939735230193e-05,
+      "loss": 0.293,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7508148148148148,
+      "grad_norm": 2.125248432159424,
+      "learning_rate": 1.500098794704604e-05,
+      "loss": 0.3483,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7514074074074074,
+      "grad_norm": 1.7023439407348633,
+      "learning_rate": 1.4997036158861885e-05,
+      "loss": 0.3697,
+      "step": 1268
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 2.5245320796966553,
+      "learning_rate": 1.4993084370677733e-05,
+      "loss": 0.3861,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7525925925925926,
+      "grad_norm": 2.16561222076416,
+      "learning_rate": 1.498913258249358e-05,
+      "loss": 0.3516,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7531851851851852,
+      "grad_norm": 1.9413903951644897,
+      "learning_rate": 1.4985180794309426e-05,
+      "loss": 0.4334,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7537777777777778,
+      "grad_norm": 1.5109609365463257,
+      "learning_rate": 1.4981229006125272e-05,
+      "loss": 0.2691,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7543703703703704,
+      "grad_norm": 1.7790225744247437,
+      "learning_rate": 1.497727721794112e-05,
+      "loss": 0.3927,
+      "step": 1273
+    },
+    {
+      "epoch": 0.754962962962963,
+      "grad_norm": 1.5074994564056396,
+      "learning_rate": 1.4973325429756966e-05,
+      "loss": 0.3442,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7555555555555555,
+      "grad_norm": 2.5574607849121094,
+      "learning_rate": 1.4969373641572813e-05,
+      "loss": 0.4301,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7561481481481481,
+      "grad_norm": 1.790836215019226,
+      "learning_rate": 1.496542185338866e-05,
+      "loss": 0.3261,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7567407407407407,
+      "grad_norm": 1.5770270824432373,
+      "learning_rate": 1.4961470065204507e-05,
+      "loss": 0.3597,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7573333333333333,
+      "grad_norm": 1.5954099893569946,
+      "learning_rate": 1.4957518277020353e-05,
+      "loss": 0.372,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7579259259259259,
+      "grad_norm": 1.4104055166244507,
+      "learning_rate": 1.49535664888362e-05,
+      "loss": 0.3295,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7585185185185185,
+      "grad_norm": 3.6164028644561768,
+      "learning_rate": 1.4949614700652047e-05,
+      "loss": 0.6318,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7591111111111111,
+      "grad_norm": 2.4321279525756836,
+      "learning_rate": 1.4945662912467893e-05,
+      "loss": 0.3153,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7597037037037037,
+      "grad_norm": 1.3936355113983154,
+      "learning_rate": 1.494171112428374e-05,
+      "loss": 0.224,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7602962962962962,
+      "grad_norm": 1.746045470237732,
+      "learning_rate": 1.4937759336099586e-05,
+      "loss": 0.4415,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7608888888888888,
+      "grad_norm": 2.1861984729766846,
+      "learning_rate": 1.4933807547915434e-05,
+      "loss": 0.3582,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7614814814814815,
+      "grad_norm": 2.858353614807129,
+      "learning_rate": 1.492985575973128e-05,
+      "loss": 0.4887,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7620740740740741,
+      "grad_norm": 2.1178908348083496,
+      "learning_rate": 1.4925903971547126e-05,
+      "loss": 0.4191,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7626666666666667,
+      "grad_norm": 2.228424310684204,
+      "learning_rate": 1.4921952183362973e-05,
+      "loss": 0.3612,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7632592592592593,
+      "grad_norm": 2.433380365371704,
+      "learning_rate": 1.491800039517882e-05,
+      "loss": 0.393,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7638518518518519,
+      "grad_norm": 4.5052714347839355,
+      "learning_rate": 1.4914048606994667e-05,
+      "loss": 0.6001,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7644444444444445,
+      "grad_norm": 1.881618618965149,
+      "learning_rate": 1.4910096818810513e-05,
+      "loss": 0.2948,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7650370370370371,
+      "grad_norm": 2.2272636890411377,
+      "learning_rate": 1.490614503062636e-05,
+      "loss": 0.3091,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7656296296296297,
+      "grad_norm": 3.122849702835083,
+      "learning_rate": 1.4902193242442207e-05,
+      "loss": 0.4691,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7662222222222222,
+      "grad_norm": 2.3162851333618164,
+      "learning_rate": 1.4898241454258053e-05,
+      "loss": 0.4009,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7668148148148148,
+      "grad_norm": 2.8754966259002686,
+      "learning_rate": 1.48942896660739e-05,
+      "loss": 0.3832,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7674074074074074,
+      "grad_norm": 2.317288637161255,
+      "learning_rate": 1.4890337877889746e-05,
+      "loss": 0.4269,
+      "step": 1295
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 1.9380499124526978,
+      "learning_rate": 1.4886386089705594e-05,
+      "loss": 0.3774,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7685925925925926,
+      "grad_norm": 2.10446834564209,
+      "learning_rate": 1.488243430152144e-05,
+      "loss": 0.3977,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7691851851851852,
+      "grad_norm": 1.8105900287628174,
+      "learning_rate": 1.4878482513337286e-05,
+      "loss": 0.3387,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7697777777777778,
+      "grad_norm": 2.3005518913269043,
+      "learning_rate": 1.4874530725153134e-05,
+      "loss": 0.3979,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7703703703703704,
+      "grad_norm": 2.4637203216552734,
+      "learning_rate": 1.487057893696898e-05,
+      "loss": 0.3487,
+      "step": 1300
+    },
+    {
+      "epoch": 0.770962962962963,
+      "grad_norm": 4.370997905731201,
+      "learning_rate": 1.4866627148784827e-05,
+      "loss": 0.4481,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7715555555555556,
+      "grad_norm": 1.9340109825134277,
+      "learning_rate": 1.4862675360600673e-05,
+      "loss": 0.3132,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7721481481481481,
+      "grad_norm": 2.3620381355285645,
+      "learning_rate": 1.4858723572416521e-05,
+      "loss": 0.3855,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7727407407407407,
+      "grad_norm": 1.6071555614471436,
+      "learning_rate": 1.4854771784232367e-05,
+      "loss": 0.3082,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7733333333333333,
+      "grad_norm": 1.9710509777069092,
+      "learning_rate": 1.4850819996048213e-05,
+      "loss": 0.337,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7739259259259259,
+      "grad_norm": 3.7946200370788574,
+      "learning_rate": 1.484686820786406e-05,
+      "loss": 0.4023,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7745185185185185,
+      "grad_norm": 2.690908432006836,
+      "learning_rate": 1.4842916419679908e-05,
+      "loss": 0.329,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7751111111111111,
+      "grad_norm": 1.6520037651062012,
+      "learning_rate": 1.4838964631495754e-05,
+      "loss": 0.3213,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7757037037037037,
+      "grad_norm": 1.7754448652267456,
+      "learning_rate": 1.48350128433116e-05,
+      "loss": 0.3951,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7762962962962963,
+      "grad_norm": 2.9733638763427734,
+      "learning_rate": 1.4831061055127446e-05,
+      "loss": 0.4122,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7768888888888889,
+      "grad_norm": 1.8579957485198975,
+      "learning_rate": 1.4827109266943294e-05,
+      "loss": 0.4439,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7774814814814814,
+      "grad_norm": 4.180583477020264,
+      "learning_rate": 1.482315747875914e-05,
+      "loss": 0.4382,
+      "step": 1312
+    },
+    {
+      "epoch": 0.778074074074074,
+      "grad_norm": 2.4216668605804443,
+      "learning_rate": 1.4819205690574987e-05,
+      "loss": 0.3418,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7786666666666666,
+      "grad_norm": 1.4886480569839478,
+      "learning_rate": 1.4815253902390833e-05,
+      "loss": 0.2617,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7792592592592592,
+      "grad_norm": 2.0646674633026123,
+      "learning_rate": 1.4811302114206681e-05,
+      "loss": 0.4966,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7798518518518519,
+      "grad_norm": 2.243691921234131,
+      "learning_rate": 1.4807350326022527e-05,
+      "loss": 0.3816,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7804444444444445,
+      "grad_norm": 3.9576032161712646,
+      "learning_rate": 1.4803398537838373e-05,
+      "loss": 0.566,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7810370370370371,
+      "grad_norm": 2.383575201034546,
+      "learning_rate": 1.479944674965422e-05,
+      "loss": 0.5063,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7816296296296297,
+      "grad_norm": 1.6924495697021484,
+      "learning_rate": 1.4795494961470068e-05,
+      "loss": 0.3186,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7822222222222223,
+      "grad_norm": 2.0734047889709473,
+      "learning_rate": 1.4791543173285914e-05,
+      "loss": 0.3421,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7828148148148149,
+      "grad_norm": 2.4890682697296143,
+      "learning_rate": 1.478759138510176e-05,
+      "loss": 0.3749,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7834074074074074,
+      "grad_norm": 1.7678948640823364,
+      "learning_rate": 1.4783639596917608e-05,
+      "loss": 0.3963,
+      "step": 1322
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 2.1059906482696533,
+      "learning_rate": 1.4779687808733454e-05,
+      "loss": 0.423,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7845925925925926,
+      "grad_norm": 1.7259266376495361,
+      "learning_rate": 1.47757360205493e-05,
+      "loss": 0.2536,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7851851851851852,
+      "grad_norm": 3.4462227821350098,
+      "learning_rate": 1.4771784232365147e-05,
+      "loss": 0.4867,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7857777777777778,
+      "grad_norm": 2.091338634490967,
+      "learning_rate": 1.4767832444180995e-05,
+      "loss": 0.3421,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7863703703703704,
+      "grad_norm": 1.5237092971801758,
+      "learning_rate": 1.4763880655996841e-05,
+      "loss": 0.2875,
+      "step": 1327
+    },
+    {
+      "epoch": 0.786962962962963,
+      "grad_norm": 2.1271753311157227,
+      "learning_rate": 1.4759928867812687e-05,
+      "loss": 0.381,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7875555555555556,
+      "grad_norm": 2.1506075859069824,
+      "learning_rate": 1.4755977079628533e-05,
+      "loss": 0.4986,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7881481481481482,
+      "grad_norm": 2.0281431674957275,
+      "learning_rate": 1.4752025291444381e-05,
+      "loss": 0.3286,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7887407407407407,
+      "grad_norm": 1.9858900308609009,
+      "learning_rate": 1.4748073503260228e-05,
+      "loss": 0.3328,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7893333333333333,
+      "grad_norm": 1.3083425760269165,
+      "learning_rate": 1.4744121715076074e-05,
+      "loss": 0.3177,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7899259259259259,
+      "grad_norm": 2.0950069427490234,
+      "learning_rate": 1.474016992689192e-05,
+      "loss": 0.373,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7905185185185185,
+      "grad_norm": 3.00944185256958,
+      "learning_rate": 1.4736218138707768e-05,
+      "loss": 0.358,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7911111111111111,
+      "grad_norm": 2.4436309337615967,
+      "learning_rate": 1.4732266350523614e-05,
+      "loss": 0.3577,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7917037037037037,
+      "grad_norm": 2.0806429386138916,
+      "learning_rate": 1.472831456233946e-05,
+      "loss": 0.3604,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7922962962962963,
+      "grad_norm": 2.0733890533447266,
+      "learning_rate": 1.4724362774155307e-05,
+      "loss": 0.3115,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7928888888888889,
+      "grad_norm": 1.5340299606323242,
+      "learning_rate": 1.4720410985971155e-05,
+      "loss": 0.3372,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7934814814814815,
+      "grad_norm": 1.9966020584106445,
+      "learning_rate": 1.4716459197787001e-05,
+      "loss": 0.3437,
+      "step": 1339
+    },
+    {
+      "epoch": 0.794074074074074,
+      "grad_norm": 3.080676794052124,
+      "learning_rate": 1.4712507409602847e-05,
+      "loss": 0.4307,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7946666666666666,
+      "grad_norm": 3.173445463180542,
+      "learning_rate": 1.4708555621418693e-05,
+      "loss": 0.5766,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7952592592592592,
+      "grad_norm": 2.173215627670288,
+      "learning_rate": 1.4704603833234541e-05,
+      "loss": 0.4004,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7958518518518518,
+      "grad_norm": 1.8060274124145508,
+      "learning_rate": 1.4700652045050388e-05,
+      "loss": 0.3333,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7964444444444444,
+      "grad_norm": 2.1241540908813477,
+      "learning_rate": 1.4696700256866234e-05,
+      "loss": 0.4473,
+      "step": 1344
+    },
+    {
+      "epoch": 0.797037037037037,
+      "grad_norm": 1.5435324907302856,
+      "learning_rate": 1.4692748468682082e-05,
+      "loss": 0.3445,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7976296296296296,
+      "grad_norm": 1.5190536975860596,
+      "learning_rate": 1.4688796680497928e-05,
+      "loss": 0.325,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7982222222222223,
+      "grad_norm": 1.949932336807251,
+      "learning_rate": 1.4684844892313773e-05,
+      "loss": 0.4584,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7988148148148149,
+      "grad_norm": 1.4186478853225708,
+      "learning_rate": 1.4680893104129619e-05,
+      "loss": 0.2751,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7994074074074075,
+      "grad_norm": 1.691709041595459,
+      "learning_rate": 1.4676941315945465e-05,
+      "loss": 0.4021,
+      "step": 1349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.624911904335022,
+      "learning_rate": 1.4672989527761313e-05,
+      "loss": 0.371,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8005925925925926,
+      "grad_norm": 2.847836494445801,
+      "learning_rate": 1.466903773957716e-05,
+      "loss": 0.5035,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8011851851851852,
+      "grad_norm": 2.0204355716705322,
+      "learning_rate": 1.4665085951393005e-05,
+      "loss": 0.3528,
+      "step": 1352
+    },
+    {
+      "epoch": 0.8017777777777778,
+      "grad_norm": 2.081942081451416,
+      "learning_rate": 1.4661134163208852e-05,
+      "loss": 0.4,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8023703703703704,
+      "grad_norm": 2.5324795246124268,
+      "learning_rate": 1.46571823750247e-05,
+      "loss": 0.2457,
+      "step": 1354
+    },
+    {
+      "epoch": 0.802962962962963,
+      "grad_norm": 3.0072240829467773,
+      "learning_rate": 1.4653230586840546e-05,
+      "loss": 0.5633,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8035555555555556,
+      "grad_norm": 2.2367117404937744,
+      "learning_rate": 1.4649278798656392e-05,
+      "loss": 0.4123,
+      "step": 1356
+    },
+    {
+      "epoch": 0.8041481481481482,
+      "grad_norm": 1.6455742120742798,
+      "learning_rate": 1.4645327010472238e-05,
+      "loss": 0.3446,
+      "step": 1357
+    },
+    {
+      "epoch": 0.8047407407407408,
+      "grad_norm": 2.4111199378967285,
+      "learning_rate": 1.4641375222288086e-05,
+      "loss": 0.4357,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8053333333333333,
+      "grad_norm": 1.9356091022491455,
+      "learning_rate": 1.4637423434103933e-05,
+      "loss": 0.3906,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8059259259259259,
+      "grad_norm": 1.6176751852035522,
+      "learning_rate": 1.4633471645919779e-05,
+      "loss": 0.2996,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8065185185185185,
+      "grad_norm": 1.9958466291427612,
+      "learning_rate": 1.4629519857735625e-05,
+      "loss": 0.2988,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8071111111111111,
+      "grad_norm": 2.080132007598877,
+      "learning_rate": 1.4625568069551473e-05,
+      "loss": 0.3764,
+      "step": 1362
+    },
+    {
+      "epoch": 0.8077037037037037,
+      "grad_norm": 3.5478086471557617,
+      "learning_rate": 1.462161628136732e-05,
+      "loss": 0.5304,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8082962962962963,
+      "grad_norm": 2.3769078254699707,
+      "learning_rate": 1.4617664493183165e-05,
+      "loss": 0.3382,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8088888888888889,
+      "grad_norm": 2.092668294906616,
+      "learning_rate": 1.4613712704999012e-05,
+      "loss": 0.4588,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8094814814814815,
+      "grad_norm": 1.5667452812194824,
+      "learning_rate": 1.460976091681486e-05,
+      "loss": 0.3049,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8100740740740741,
+      "grad_norm": 1.8400635719299316,
+      "learning_rate": 1.4605809128630706e-05,
+      "loss": 0.3671,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8106666666666666,
+      "grad_norm": 1.982218623161316,
+      "learning_rate": 1.4601857340446552e-05,
+      "loss": 0.4191,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8112592592592592,
+      "grad_norm": 2.1156904697418213,
+      "learning_rate": 1.4597905552262398e-05,
+      "loss": 0.4175,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8118518518518518,
+      "grad_norm": 1.6793193817138672,
+      "learning_rate": 1.4593953764078246e-05,
+      "loss": 0.3839,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8124444444444444,
+      "grad_norm": 1.4875460863113403,
+      "learning_rate": 1.4590001975894093e-05,
+      "loss": 0.3022,
+      "step": 1371
+    },
+    {
+      "epoch": 0.813037037037037,
+      "grad_norm": 1.823776364326477,
+      "learning_rate": 1.4586050187709939e-05,
+      "loss": 0.4292,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8136296296296296,
+      "grad_norm": 1.692374587059021,
+      "learning_rate": 1.4582098399525785e-05,
+      "loss": 0.3712,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8142222222222222,
+      "grad_norm": 2.1633989810943604,
+      "learning_rate": 1.4578146611341633e-05,
+      "loss": 0.3551,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 2.1315886974334717,
+      "learning_rate": 1.457419482315748e-05,
+      "loss": 0.338,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8154074074074074,
+      "grad_norm": 2.629422426223755,
+      "learning_rate": 1.4570243034973325e-05,
+      "loss": 0.4366,
+      "step": 1376
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 2.3739476203918457,
+      "learning_rate": 1.4566291246789173e-05,
+      "loss": 0.3908,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8165925925925926,
+      "grad_norm": 1.7683513164520264,
+      "learning_rate": 1.456233945860502e-05,
+      "loss": 0.3022,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8171851851851852,
+      "grad_norm": 2.912783145904541,
+      "learning_rate": 1.4558387670420866e-05,
+      "loss": 0.4246,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8177777777777778,
+      "grad_norm": 2.947299003601074,
+      "learning_rate": 1.4554435882236712e-05,
+      "loss": 0.3866,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8183703703703704,
+      "grad_norm": 2.9963865280151367,
+      "learning_rate": 1.455048409405256e-05,
+      "loss": 0.3901,
+      "step": 1381
+    },
+    {
+      "epoch": 0.818962962962963,
+      "grad_norm": 2.4726345539093018,
+      "learning_rate": 1.4546532305868406e-05,
+      "loss": 0.4136,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8195555555555556,
+      "grad_norm": 2.2182836532592773,
+      "learning_rate": 1.4542580517684253e-05,
+      "loss": 0.358,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8201481481481482,
+      "grad_norm": 1.721076488494873,
+      "learning_rate": 1.4538628729500099e-05,
+      "loss": 0.3646,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8207407407407408,
+      "grad_norm": 1.6971954107284546,
+      "learning_rate": 1.4534676941315947e-05,
+      "loss": 0.3906,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8213333333333334,
+      "grad_norm": 1.4247231483459473,
+      "learning_rate": 1.4530725153131793e-05,
+      "loss": 0.3725,
+      "step": 1386
+    },
+    {
+      "epoch": 0.821925925925926,
+      "grad_norm": 1.8083750009536743,
+      "learning_rate": 1.452677336494764e-05,
+      "loss": 0.3196,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8225185185185185,
+      "grad_norm": 1.8028483390808105,
+      "learning_rate": 1.4522821576763485e-05,
+      "loss": 0.2671,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8231111111111111,
+      "grad_norm": 1.6177033185958862,
+      "learning_rate": 1.4518869788579333e-05,
+      "loss": 0.3803,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8237037037037037,
+      "grad_norm": 2.3013787269592285,
+      "learning_rate": 1.451491800039518e-05,
+      "loss": 0.5159,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8242962962962963,
+      "grad_norm": 2.6012673377990723,
+      "learning_rate": 1.4510966212211026e-05,
+      "loss": 0.4217,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8248888888888889,
+      "grad_norm": 2.8011391162872314,
+      "learning_rate": 1.4507014424026872e-05,
+      "loss": 0.4379,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8254814814814815,
+      "grad_norm": 2.377157211303711,
+      "learning_rate": 1.450306263584272e-05,
+      "loss": 0.3852,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8260740740740741,
+      "grad_norm": 3.0142533779144287,
+      "learning_rate": 1.4499110847658566e-05,
+      "loss": 0.5167,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8266666666666667,
+      "grad_norm": 1.9622479677200317,
+      "learning_rate": 1.4495159059474413e-05,
+      "loss": 0.3535,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8272592592592592,
+      "grad_norm": 1.7239298820495605,
+      "learning_rate": 1.4491207271290259e-05,
+      "loss": 0.3121,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8278518518518518,
+      "grad_norm": 1.6423649787902832,
+      "learning_rate": 1.4487255483106107e-05,
+      "loss": 0.3601,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8284444444444444,
+      "grad_norm": 1.9214917421340942,
+      "learning_rate": 1.4483303694921953e-05,
+      "loss": 0.4007,
+      "step": 1398
+    },
+    {
+      "epoch": 0.829037037037037,
+      "grad_norm": 1.364236831665039,
+      "learning_rate": 1.44793519067378e-05,
+      "loss": 0.2907,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8296296296296296,
+      "grad_norm": 3.852635383605957,
+      "learning_rate": 1.4475400118553647e-05,
+      "loss": 0.4499,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8302222222222222,
+      "grad_norm": 2.2103161811828613,
+      "learning_rate": 1.4471448330369493e-05,
+      "loss": 0.4151,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8308148148148148,
+      "grad_norm": 1.897921085357666,
+      "learning_rate": 1.446749654218534e-05,
+      "loss": 0.4607,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8314074074074074,
+      "grad_norm": 3.086670398712158,
+      "learning_rate": 1.4463544754001186e-05,
+      "loss": 0.4821,
+      "step": 1403
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 1.827023983001709,
+      "learning_rate": 1.4459592965817034e-05,
+      "loss": 0.2613,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8325925925925926,
+      "grad_norm": 2.178452491760254,
+      "learning_rate": 1.445564117763288e-05,
+      "loss": 0.3234,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8331851851851851,
+      "grad_norm": 2.0375664234161377,
+      "learning_rate": 1.4451689389448726e-05,
+      "loss": 0.4478,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8337777777777777,
+      "grad_norm": 2.264516830444336,
+      "learning_rate": 1.4447737601264573e-05,
+      "loss": 0.382,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8343703703703703,
+      "grad_norm": 2.119260311126709,
+      "learning_rate": 1.444378581308042e-05,
+      "loss": 0.2977,
+      "step": 1408
+    },
+    {
+      "epoch": 0.834962962962963,
+      "grad_norm": 2.167792558670044,
+      "learning_rate": 1.4439834024896267e-05,
+      "loss": 0.3541,
+      "step": 1409
+    },
+    {
+      "epoch": 0.8355555555555556,
+      "grad_norm": 1.8270937204360962,
+      "learning_rate": 1.4435882236712113e-05,
+      "loss": 0.4304,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8361481481481482,
+      "grad_norm": 1.7227752208709717,
+      "learning_rate": 1.443193044852796e-05,
+      "loss": 0.3458,
+      "step": 1411
+    },
+    {
+      "epoch": 0.8367407407407408,
+      "grad_norm": 1.4413201808929443,
+      "learning_rate": 1.4427978660343807e-05,
+      "loss": 0.3268,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8373333333333334,
+      "grad_norm": 1.5207862854003906,
+      "learning_rate": 1.4424026872159653e-05,
+      "loss": 0.2922,
+      "step": 1413
+    },
+    {
+      "epoch": 0.837925925925926,
+      "grad_norm": 1.7909324169158936,
+      "learning_rate": 1.44200750839755e-05,
+      "loss": 0.3591,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8385185185185186,
+      "grad_norm": 2.0871431827545166,
+      "learning_rate": 1.4416123295791346e-05,
+      "loss": 0.3744,
+      "step": 1415
+    },
+    {
+      "epoch": 0.8391111111111111,
+      "grad_norm": 4.103446006774902,
+      "learning_rate": 1.4412171507607194e-05,
+      "loss": 0.4774,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8397037037037037,
+      "grad_norm": 1.7280763387680054,
+      "learning_rate": 1.440821971942304e-05,
+      "loss": 0.2836,
+      "step": 1417
+    },
+    {
+      "epoch": 0.8402962962962963,
+      "grad_norm": 1.5372905731201172,
+      "learning_rate": 1.4404267931238886e-05,
+      "loss": 0.3042,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8408888888888889,
+      "grad_norm": 2.559356689453125,
+      "learning_rate": 1.4400316143054733e-05,
+      "loss": 0.4966,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8414814814814815,
+      "grad_norm": 1.9310855865478516,
+      "learning_rate": 1.439636435487058e-05,
+      "loss": 0.421,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8420740740740741,
+      "grad_norm": 1.9700080156326294,
+      "learning_rate": 1.4392412566686427e-05,
+      "loss": 0.3444,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8426666666666667,
+      "grad_norm": 1.5184448957443237,
+      "learning_rate": 1.4388460778502273e-05,
+      "loss": 0.322,
+      "step": 1422
+    },
+    {
+      "epoch": 0.8432592592592593,
+      "grad_norm": 1.696308970451355,
+      "learning_rate": 1.4384508990318121e-05,
+      "loss": 0.257,
+      "step": 1423
+    },
+    {
+      "epoch": 0.8438518518518519,
+      "grad_norm": 2.2121431827545166,
+      "learning_rate": 1.4380557202133967e-05,
+      "loss": 0.346,
+      "step": 1424
+    },
+    {
+      "epoch": 0.8444444444444444,
+      "grad_norm": 1.7427834272384644,
+      "learning_rate": 1.4376605413949813e-05,
+      "loss": 0.2737,
+      "step": 1425
+    },
+    {
+      "epoch": 0.845037037037037,
+      "grad_norm": 4.085221290588379,
+      "learning_rate": 1.437265362576566e-05,
+      "loss": 0.4625,
+      "step": 1426
+    },
+    {
+      "epoch": 0.8456296296296296,
+      "grad_norm": 2.2473208904266357,
+      "learning_rate": 1.4368701837581508e-05,
+      "loss": 0.4979,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8462222222222222,
+      "grad_norm": 4.116520404815674,
+      "learning_rate": 1.4364750049397354e-05,
+      "loss": 0.4253,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8468148148148148,
+      "grad_norm": 2.1700146198272705,
+      "learning_rate": 1.43607982612132e-05,
+      "loss": 0.3042,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8474074074074074,
+      "grad_norm": 1.680627465248108,
+      "learning_rate": 1.4356846473029046e-05,
+      "loss": 0.3298,
+      "step": 1430
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 1.9758901596069336,
+      "learning_rate": 1.4352894684844894e-05,
+      "loss": 0.364,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8485925925925926,
+      "grad_norm": 1.9136720895767212,
+      "learning_rate": 1.434894289666074e-05,
+      "loss": 0.4339,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8491851851851852,
+      "grad_norm": 1.5847541093826294,
+      "learning_rate": 1.4344991108476587e-05,
+      "loss": 0.3536,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8497777777777777,
+      "grad_norm": 2.077934741973877,
+      "learning_rate": 1.4341039320292433e-05,
+      "loss": 0.506,
+      "step": 1434
+    },
+    {
+      "epoch": 0.8503703703703703,
+      "grad_norm": 2.307828187942505,
+      "learning_rate": 1.4337087532108281e-05,
+      "loss": 0.3834,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8509629629629629,
+      "grad_norm": 1.9874824285507202,
+      "learning_rate": 1.4333135743924127e-05,
+      "loss": 0.3897,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8515555555555555,
+      "grad_norm": 2.7278780937194824,
+      "learning_rate": 1.4329183955739973e-05,
+      "loss": 0.3168,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8521481481481481,
+      "grad_norm": 2.6640405654907227,
+      "learning_rate": 1.432523216755582e-05,
+      "loss": 0.4962,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8527407407407407,
+      "grad_norm": 4.093876838684082,
+      "learning_rate": 1.4321280379371668e-05,
+      "loss": 0.5168,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "grad_norm": 2.5805482864379883,
+      "learning_rate": 1.4317328591187514e-05,
+      "loss": 0.4144,
+      "step": 1440
+    },
+    {
+      "epoch": 0.853925925925926,
+      "grad_norm": 1.6923569440841675,
+      "learning_rate": 1.431337680300336e-05,
+      "loss": 0.3421,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8545185185185186,
+      "grad_norm": 1.777841567993164,
+      "learning_rate": 1.4309425014819206e-05,
+      "loss": 0.3344,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8551111111111112,
+      "grad_norm": 1.870968222618103,
+      "learning_rate": 1.4305473226635054e-05,
+      "loss": 0.424,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8557037037037037,
+      "grad_norm": 1.8172470331192017,
+      "learning_rate": 1.43015214384509e-05,
+      "loss": 0.2993,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8562962962962963,
+      "grad_norm": 1.4466150999069214,
+      "learning_rate": 1.4297569650266747e-05,
+      "loss": 0.2465,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8568888888888889,
+      "grad_norm": 2.4257454872131348,
+      "learning_rate": 1.4293617862082595e-05,
+      "loss": 0.4596,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8574814814814815,
+      "grad_norm": 1.7246711254119873,
+      "learning_rate": 1.4289666073898441e-05,
+      "loss": 0.3556,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8580740740740741,
+      "grad_norm": 2.919105052947998,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.2992,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8586666666666667,
+      "grad_norm": 2.547780752182007,
+      "learning_rate": 1.4281762497530133e-05,
+      "loss": 0.3888,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8592592592592593,
+      "grad_norm": 2.0909159183502197,
+      "learning_rate": 1.4277810709345981e-05,
+      "loss": 0.407,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8598518518518519,
+      "grad_norm": 2.2178869247436523,
+      "learning_rate": 1.4273858921161828e-05,
+      "loss": 0.4444,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8604444444444445,
+      "grad_norm": 2.482827663421631,
+      "learning_rate": 1.4269907132977674e-05,
+      "loss": 0.4161,
+      "step": 1452
+    },
+    {
+      "epoch": 0.861037037037037,
+      "grad_norm": 1.7337634563446045,
+      "learning_rate": 1.426595534479352e-05,
+      "loss": 0.4986,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8616296296296296,
+      "grad_norm": 6.733527660369873,
+      "learning_rate": 1.4262003556609368e-05,
+      "loss": 0.4695,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8622222222222222,
+      "grad_norm": 2.813152551651001,
+      "learning_rate": 1.4258051768425214e-05,
+      "loss": 0.3401,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8628148148148148,
+      "grad_norm": 3.0269768238067627,
+      "learning_rate": 1.425409998024106e-05,
+      "loss": 0.5348,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8634074074074074,
+      "grad_norm": 2.199885845184326,
+      "learning_rate": 1.4250148192056907e-05,
+      "loss": 0.4102,
+      "step": 1457
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 1.8111685514450073,
+      "learning_rate": 1.4246196403872755e-05,
+      "loss": 0.277,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8645925925925926,
+      "grad_norm": 3.0107901096343994,
+      "learning_rate": 1.4242244615688601e-05,
+      "loss": 0.3812,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8651851851851852,
+      "grad_norm": 1.7943427562713623,
+      "learning_rate": 1.4238292827504447e-05,
+      "loss": 0.3629,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8657777777777778,
+      "grad_norm": 2.667762279510498,
+      "learning_rate": 1.4234341039320293e-05,
+      "loss": 0.4208,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8663703703703703,
+      "grad_norm": 1.4568886756896973,
+      "learning_rate": 1.4230389251136141e-05,
+      "loss": 0.2667,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8669629629629629,
+      "grad_norm": 1.87785005569458,
+      "learning_rate": 1.4226437462951988e-05,
+      "loss": 0.3237,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8675555555555555,
+      "grad_norm": 2.4205000400543213,
+      "learning_rate": 1.4222485674767834e-05,
+      "loss": 0.4028,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8681481481481481,
+      "grad_norm": 2.0882771015167236,
+      "learning_rate": 1.421853388658368e-05,
+      "loss": 0.3836,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8687407407407407,
+      "grad_norm": 1.940891981124878,
+      "learning_rate": 1.4214582098399528e-05,
+      "loss": 0.3116,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8693333333333333,
+      "grad_norm": 2.0996110439300537,
+      "learning_rate": 1.4210630310215374e-05,
+      "loss": 0.2457,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8699259259259259,
+      "grad_norm": 2.129406452178955,
+      "learning_rate": 1.420667852203122e-05,
+      "loss": 0.4343,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8705185185185185,
+      "grad_norm": 2.1356394290924072,
+      "learning_rate": 1.4202726733847068e-05,
+      "loss": 0.4051,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8711111111111111,
+      "grad_norm": 2.3795502185821533,
+      "learning_rate": 1.4198774945662915e-05,
+      "loss": 0.4421,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8717037037037038,
+      "grad_norm": 2.483039379119873,
+      "learning_rate": 1.4194823157478761e-05,
+      "loss": 0.3546,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8722962962962963,
+      "grad_norm": 2.407362937927246,
+      "learning_rate": 1.4190871369294607e-05,
+      "loss": 0.4527,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8728888888888889,
+      "grad_norm": 1.9372997283935547,
+      "learning_rate": 1.4186919581110455e-05,
+      "loss": 0.4081,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8734814814814815,
+      "grad_norm": 1.8833132982254028,
+      "learning_rate": 1.4182967792926301e-05,
+      "loss": 0.3714,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8740740740740741,
+      "grad_norm": 2.57008957862854,
+      "learning_rate": 1.4179016004742148e-05,
+      "loss": 0.3364,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8746666666666667,
+      "grad_norm": 3.96630859375,
+      "learning_rate": 1.4175064216557994e-05,
+      "loss": 0.6189,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8752592592592593,
+      "grad_norm": 1.9076176881790161,
+      "learning_rate": 1.4171112428373842e-05,
+      "loss": 0.3909,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8758518518518519,
+      "grad_norm": 2.583515167236328,
+      "learning_rate": 1.4167160640189688e-05,
+      "loss": 0.3478,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8764444444444445,
+      "grad_norm": 1.9139636754989624,
+      "learning_rate": 1.4163208852005534e-05,
+      "loss": 0.4505,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8770370370370371,
+      "grad_norm": 2.2974934577941895,
+      "learning_rate": 1.415925706382138e-05,
+      "loss": 0.3764,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8776296296296296,
+      "grad_norm": 1.8406239748001099,
+      "learning_rate": 1.4155305275637228e-05,
+      "loss": 0.459,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8782222222222222,
+      "grad_norm": 2.0807511806488037,
+      "learning_rate": 1.4151353487453075e-05,
+      "loss": 0.3442,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8788148148148148,
+      "grad_norm": 1.7351526021957397,
+      "learning_rate": 1.4147401699268921e-05,
+      "loss": 0.3741,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8794074074074074,
+      "grad_norm": 1.6013861894607544,
+      "learning_rate": 1.4143449911084767e-05,
+      "loss": 0.3027,
+      "step": 1484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9782830476760864,
+      "learning_rate": 1.4139498122900615e-05,
+      "loss": 0.3248,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8805925925925926,
+      "grad_norm": 1.6817258596420288,
+      "learning_rate": 1.4135546334716461e-05,
+      "loss": 0.3453,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8811851851851852,
+      "grad_norm": 1.990211844444275,
+      "learning_rate": 1.4131594546532308e-05,
+      "loss": 0.3474,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8817777777777778,
+      "grad_norm": 2.407019853591919,
+      "learning_rate": 1.4127642758348154e-05,
+      "loss": 0.4238,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8823703703703704,
+      "grad_norm": 1.7395319938659668,
+      "learning_rate": 1.4123690970164002e-05,
+      "loss": 0.3234,
+      "step": 1489
+    },
+    {
+      "epoch": 0.882962962962963,
+      "grad_norm": 1.4735463857650757,
+      "learning_rate": 1.4119739181979848e-05,
+      "loss": 0.2772,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8835555555555555,
+      "grad_norm": 1.9431641101837158,
+      "learning_rate": 1.4115787393795694e-05,
+      "loss": 0.3901,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8841481481481481,
+      "grad_norm": 1.6273599863052368,
+      "learning_rate": 1.4111835605611542e-05,
+      "loss": 0.3996,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8847407407407407,
+      "grad_norm": 2.558722734451294,
+      "learning_rate": 1.4107883817427388e-05,
+      "loss": 0.3452,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8853333333333333,
+      "grad_norm": 2.1233022212982178,
+      "learning_rate": 1.4103932029243235e-05,
+      "loss": 0.3409,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8859259259259259,
+      "grad_norm": 2.1418938636779785,
+      "learning_rate": 1.4099980241059081e-05,
+      "loss": 0.3232,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8865185185185185,
+      "grad_norm": 1.7888809442520142,
+      "learning_rate": 1.4096028452874929e-05,
+      "loss": 0.3882,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8871111111111111,
+      "grad_norm": 2.4423131942749023,
+      "learning_rate": 1.4092076664690775e-05,
+      "loss": 0.488,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8877037037037037,
+      "grad_norm": 2.378862142562866,
+      "learning_rate": 1.4088124876506621e-05,
+      "loss": 0.4901,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8882962962962963,
+      "grad_norm": 2.3293371200561523,
+      "learning_rate": 1.4084173088322468e-05,
+      "loss": 0.3301,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 2.1011767387390137,
+      "learning_rate": 1.4080221300138315e-05,
+      "loss": 0.3928,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8894814814814814,
+      "grad_norm": 2.0898020267486572,
+      "learning_rate": 1.4076269511954162e-05,
+      "loss": 0.3931,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8900740740740741,
+      "grad_norm": 2.2307989597320557,
+      "learning_rate": 1.4072317723770008e-05,
+      "loss": 0.4077,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8906666666666667,
+      "grad_norm": 1.8513140678405762,
+      "learning_rate": 1.4068365935585854e-05,
+      "loss": 0.3238,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8912592592592593,
+      "grad_norm": 1.7100841999053955,
+      "learning_rate": 1.4064414147401702e-05,
+      "loss": 0.3337,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8918518518518519,
+      "grad_norm": 1.688895583152771,
+      "learning_rate": 1.4060462359217547e-05,
+      "loss": 0.3522,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8924444444444445,
+      "grad_norm": 1.6983891725540161,
+      "learning_rate": 1.4056510571033393e-05,
+      "loss": 0.3881,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8930370370370371,
+      "grad_norm": 1.4721825122833252,
+      "learning_rate": 1.405255878284924e-05,
+      "loss": 0.2668,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8936296296296297,
+      "grad_norm": 1.635797381401062,
+      "learning_rate": 1.4048606994665085e-05,
+      "loss": 0.2686,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8942222222222223,
+      "grad_norm": 2.396505355834961,
+      "learning_rate": 1.4044655206480933e-05,
+      "loss": 0.5346,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8948148148148148,
+      "grad_norm": 3.2132060527801514,
+      "learning_rate": 1.404070341829678e-05,
+      "loss": 0.4198,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8954074074074074,
+      "grad_norm": 2.307114839553833,
+      "learning_rate": 1.4036751630112626e-05,
+      "loss": 0.4343,
+      "step": 1511
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 3.4945361614227295,
+      "learning_rate": 1.4032799841928472e-05,
+      "loss": 0.4681,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8965925925925926,
+      "grad_norm": 1.5815505981445312,
+      "learning_rate": 1.402884805374432e-05,
+      "loss": 0.2895,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8971851851851852,
+      "grad_norm": 2.2456374168395996,
+      "learning_rate": 1.4024896265560166e-05,
+      "loss": 0.2966,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8977777777777778,
+      "grad_norm": 2.144205331802368,
+      "learning_rate": 1.4020944477376013e-05,
+      "loss": 0.348,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8983703703703704,
+      "grad_norm": 1.6796717643737793,
+      "learning_rate": 1.4016992689191859e-05,
+      "loss": 0.3721,
+      "step": 1516
+    },
+    {
+      "epoch": 0.898962962962963,
+      "grad_norm": 1.5109621286392212,
+      "learning_rate": 1.4013040901007707e-05,
+      "loss": 0.3225,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8995555555555556,
+      "grad_norm": 1.5097479820251465,
+      "learning_rate": 1.4009089112823553e-05,
+      "loss": 0.2259,
+      "step": 1518
+    },
+    {
+      "epoch": 0.9001481481481481,
+      "grad_norm": 2.2350926399230957,
+      "learning_rate": 1.40051373246394e-05,
+      "loss": 0.3597,
+      "step": 1519
+    },
+    {
+      "epoch": 0.9007407407407407,
+      "grad_norm": 1.9831076860427856,
+      "learning_rate": 1.4001185536455245e-05,
+      "loss": 0.414,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9013333333333333,
+      "grad_norm": 1.8318032026290894,
+      "learning_rate": 1.3997233748271093e-05,
+      "loss": 0.3992,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9019259259259259,
+      "grad_norm": 2.2491469383239746,
+      "learning_rate": 1.399328196008694e-05,
+      "loss": 0.413,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9025185185185185,
+      "grad_norm": 2.2753543853759766,
+      "learning_rate": 1.3989330171902786e-05,
+      "loss": 0.426,
+      "step": 1523
+    },
+    {
+      "epoch": 0.9031111111111111,
+      "grad_norm": 2.0645904541015625,
+      "learning_rate": 1.3985378383718634e-05,
+      "loss": 0.426,
+      "step": 1524
+    },
+    {
+      "epoch": 0.9037037037037037,
+      "grad_norm": 2.6436667442321777,
+      "learning_rate": 1.398142659553448e-05,
+      "loss": 0.3455,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9042962962962963,
+      "grad_norm": 2.0623185634613037,
+      "learning_rate": 1.3977474807350326e-05,
+      "loss": 0.4806,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9048888888888889,
+      "grad_norm": 2.2691359519958496,
+      "learning_rate": 1.3973523019166172e-05,
+      "loss": 0.3326,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9054814814814814,
+      "grad_norm": 1.5437867641448975,
+      "learning_rate": 1.396957123098202e-05,
+      "loss": 0.295,
+      "step": 1528
+    },
+    {
+      "epoch": 0.906074074074074,
+      "grad_norm": 2.423393726348877,
+      "learning_rate": 1.3965619442797867e-05,
+      "loss": 0.329,
+      "step": 1529
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "grad_norm": 2.2363760471343994,
+      "learning_rate": 1.3961667654613713e-05,
+      "loss": 0.3655,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9072592592592592,
+      "grad_norm": 2.0261332988739014,
+      "learning_rate": 1.395771586642956e-05,
+      "loss": 0.3923,
+      "step": 1531
+    },
+    {
+      "epoch": 0.9078518518518518,
+      "grad_norm": 2.1730196475982666,
+      "learning_rate": 1.3953764078245407e-05,
+      "loss": 0.3977,
+      "step": 1532
+    },
+    {
+      "epoch": 0.9084444444444445,
+      "grad_norm": 3.6098618507385254,
+      "learning_rate": 1.3949812290061253e-05,
+      "loss": 0.3607,
+      "step": 1533
+    },
+    {
+      "epoch": 0.9090370370370371,
+      "grad_norm": 2.2252938747406006,
+      "learning_rate": 1.39458605018771e-05,
+      "loss": 0.4116,
+      "step": 1534
+    },
+    {
+      "epoch": 0.9096296296296297,
+      "grad_norm": 1.7330560684204102,
+      "learning_rate": 1.3941908713692946e-05,
+      "loss": 0.3774,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9102222222222223,
+      "grad_norm": 1.715484380722046,
+      "learning_rate": 1.3937956925508794e-05,
+      "loss": 0.358,
+      "step": 1536
+    },
+    {
+      "epoch": 0.9108148148148149,
+      "grad_norm": 1.744504690170288,
+      "learning_rate": 1.393400513732464e-05,
+      "loss": 0.3128,
+      "step": 1537
+    },
+    {
+      "epoch": 0.9114074074074074,
+      "grad_norm": 2.700943946838379,
+      "learning_rate": 1.3930053349140486e-05,
+      "loss": 0.3753,
+      "step": 1538
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 3.2259740829467773,
+      "learning_rate": 1.3926101560956332e-05,
+      "loss": 0.4834,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9125925925925926,
+      "grad_norm": 1.7506686449050903,
+      "learning_rate": 1.392214977277218e-05,
+      "loss": 0.2753,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9131851851851852,
+      "grad_norm": 2.628603219985962,
+      "learning_rate": 1.3918197984588027e-05,
+      "loss": 0.4053,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9137777777777778,
+      "grad_norm": 2.7034926414489746,
+      "learning_rate": 1.3914246196403873e-05,
+      "loss": 0.5,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9143703703703704,
+      "grad_norm": 3.0678889751434326,
+      "learning_rate": 1.391029440821972e-05,
+      "loss": 0.3098,
+      "step": 1543
+    },
+    {
+      "epoch": 0.914962962962963,
+      "grad_norm": 1.6574033498764038,
+      "learning_rate": 1.3906342620035567e-05,
+      "loss": 0.3146,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9155555555555556,
+      "grad_norm": 2.494941473007202,
+      "learning_rate": 1.3902390831851413e-05,
+      "loss": 0.3589,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9161481481481482,
+      "grad_norm": 1.7242512702941895,
+      "learning_rate": 1.389843904366726e-05,
+      "loss": 0.4158,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9167407407407407,
+      "grad_norm": 3.4666688442230225,
+      "learning_rate": 1.3894487255483108e-05,
+      "loss": 0.5317,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9173333333333333,
+      "grad_norm": 2.06310772895813,
+      "learning_rate": 1.3890535467298954e-05,
+      "loss": 0.3188,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9179259259259259,
+      "grad_norm": 2.827470064163208,
+      "learning_rate": 1.38865836791148e-05,
+      "loss": 0.3503,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9185185185185185,
+      "grad_norm": 1.7820212841033936,
+      "learning_rate": 1.3882631890930646e-05,
+      "loss": 0.4023,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9191111111111111,
+      "grad_norm": 2.8055365085601807,
+      "learning_rate": 1.3878680102746494e-05,
+      "loss": 0.3665,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9197037037037037,
+      "grad_norm": 1.608599305152893,
+      "learning_rate": 1.387472831456234e-05,
+      "loss": 0.3179,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9202962962962963,
+      "grad_norm": 2.2686362266540527,
+      "learning_rate": 1.3870776526378187e-05,
+      "loss": 0.3906,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9208888888888889,
+      "grad_norm": 1.458475112915039,
+      "learning_rate": 1.3866824738194033e-05,
+      "loss": 0.2892,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9214814814814815,
+      "grad_norm": 2.7000083923339844,
+      "learning_rate": 1.3862872950009881e-05,
+      "loss": 0.408,
+      "step": 1555
+    },
+    {
+      "epoch": 0.922074074074074,
+      "grad_norm": 2.5656180381774902,
+      "learning_rate": 1.3858921161825727e-05,
+      "loss": 0.383,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9226666666666666,
+      "grad_norm": 2.9311397075653076,
+      "learning_rate": 1.3854969373641573e-05,
+      "loss": 0.4086,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9232592592592592,
+      "grad_norm": 2.723477840423584,
+      "learning_rate": 1.385101758545742e-05,
+      "loss": 0.3856,
+      "step": 1558
+    },
+    {
+      "epoch": 0.9238518518518518,
+      "grad_norm": 2.310074806213379,
+      "learning_rate": 1.3847065797273268e-05,
+      "loss": 0.3559,
+      "step": 1559
+    },
+    {
+      "epoch": 0.9244444444444444,
+      "grad_norm": 3.121511936187744,
+      "learning_rate": 1.3843114009089114e-05,
+      "loss": 0.4224,
+      "step": 1560
+    },
+    {
+      "epoch": 0.925037037037037,
+      "grad_norm": 2.807938575744629,
+      "learning_rate": 1.383916222090496e-05,
+      "loss": 0.2827,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9256296296296296,
+      "grad_norm": 3.8324713706970215,
+      "learning_rate": 1.3835210432720806e-05,
+      "loss": 0.5119,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9262222222222222,
+      "grad_norm": 4.149698734283447,
+      "learning_rate": 1.3831258644536654e-05,
+      "loss": 0.6039,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9268148148148149,
+      "grad_norm": 1.707141399383545,
+      "learning_rate": 1.38273068563525e-05,
+      "loss": 0.2977,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9274074074074075,
+      "grad_norm": 3.2056448459625244,
+      "learning_rate": 1.3823355068168347e-05,
+      "loss": 0.4949,
+      "step": 1565
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 1.7254548072814941,
+      "learning_rate": 1.3819403279984193e-05,
+      "loss": 0.3962,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9285925925925926,
+      "grad_norm": 1.8970190286636353,
+      "learning_rate": 1.381545149180004e-05,
+      "loss": 0.3145,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9291851851851852,
+      "grad_norm": 1.7479379177093506,
+      "learning_rate": 1.3811499703615887e-05,
+      "loss": 0.3529,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9297777777777778,
+      "grad_norm": 1.9552879333496094,
+      "learning_rate": 1.3807547915431733e-05,
+      "loss": 0.3464,
+      "step": 1569
+    },
+    {
+      "epoch": 0.9303703703703704,
+      "grad_norm": 2.0383613109588623,
+      "learning_rate": 1.3803596127247581e-05,
+      "loss": 0.5264,
+      "step": 1570
+    },
+    {
+      "epoch": 0.930962962962963,
+      "grad_norm": 3.800870895385742,
+      "learning_rate": 1.3799644339063428e-05,
+      "loss": 0.3281,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9315555555555556,
+      "grad_norm": 2.423962354660034,
+      "learning_rate": 1.3795692550879274e-05,
+      "loss": 0.3559,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9321481481481482,
+      "grad_norm": 1.3963474035263062,
+      "learning_rate": 1.379174076269512e-05,
+      "loss": 0.3075,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9327407407407408,
+      "grad_norm": 1.6716946363449097,
+      "learning_rate": 1.3787788974510968e-05,
+      "loss": 0.3153,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 2.164868116378784,
+      "learning_rate": 1.3783837186326814e-05,
+      "loss": 0.3729,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9339259259259259,
+      "grad_norm": 1.8524284362792969,
+      "learning_rate": 1.377988539814266e-05,
+      "loss": 0.3593,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9345185185185185,
+      "grad_norm": 2.6210479736328125,
+      "learning_rate": 1.3775933609958507e-05,
+      "loss": 0.4063,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9351111111111111,
+      "grad_norm": 2.7707602977752686,
+      "learning_rate": 1.3771981821774355e-05,
+      "loss": 0.3494,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9357037037037037,
+      "grad_norm": 2.4279770851135254,
+      "learning_rate": 1.37680300335902e-05,
+      "loss": 0.3614,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9362962962962963,
+      "grad_norm": 1.785167932510376,
+      "learning_rate": 1.3764078245406047e-05,
+      "loss": 0.3589,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9368888888888889,
+      "grad_norm": 2.049751043319702,
+      "learning_rate": 1.3760126457221893e-05,
+      "loss": 0.4387,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9374814814814815,
+      "grad_norm": 2.352332353591919,
+      "learning_rate": 1.3756174669037741e-05,
+      "loss": 0.466,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9380740740740741,
+      "grad_norm": 1.5891728401184082,
+      "learning_rate": 1.3752222880853588e-05,
+      "loss": 0.3483,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9386666666666666,
+      "grad_norm": 1.582592248916626,
+      "learning_rate": 1.3748271092669434e-05,
+      "loss": 0.3414,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9392592592592592,
+      "grad_norm": 2.1939351558685303,
+      "learning_rate": 1.374431930448528e-05,
+      "loss": 0.2849,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9398518518518518,
+      "grad_norm": 2.047973394393921,
+      "learning_rate": 1.3740367516301128e-05,
+      "loss": 0.3461,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9404444444444444,
+      "grad_norm": 1.6822198629379272,
+      "learning_rate": 1.3736415728116974e-05,
+      "loss": 0.3439,
+      "step": 1587
+    },
+    {
+      "epoch": 0.941037037037037,
+      "grad_norm": 2.081071376800537,
+      "learning_rate": 1.373246393993282e-05,
+      "loss": 0.3398,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9416296296296296,
+      "grad_norm": 2.057239055633545,
+      "learning_rate": 1.3728512151748667e-05,
+      "loss": 0.3477,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9422222222222222,
+      "grad_norm": 1.8352714776992798,
+      "learning_rate": 1.3724560363564515e-05,
+      "loss": 0.3755,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9428148148148148,
+      "grad_norm": 2.0469765663146973,
+      "learning_rate": 1.372060857538036e-05,
+      "loss": 0.3661,
+      "step": 1591
+    },
+    {
+      "epoch": 0.9434074074074074,
+      "grad_norm": 1.6702176332473755,
+      "learning_rate": 1.3716656787196207e-05,
+      "loss": 0.3176,
+      "step": 1592
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 1.963241696357727,
+      "learning_rate": 1.3712704999012055e-05,
+      "loss": 0.3304,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9445925925925925,
+      "grad_norm": 2.0911741256713867,
+      "learning_rate": 1.3708753210827901e-05,
+      "loss": 0.2803,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9451851851851852,
+      "grad_norm": 1.9866443872451782,
+      "learning_rate": 1.3704801422643748e-05,
+      "loss": 0.3664,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9457777777777778,
+      "grad_norm": 3.062960147857666,
+      "learning_rate": 1.3700849634459594e-05,
+      "loss": 0.4198,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9463703703703704,
+      "grad_norm": 3.103468418121338,
+      "learning_rate": 1.3696897846275442e-05,
+      "loss": 0.4329,
+      "step": 1597
+    },
+    {
+      "epoch": 0.946962962962963,
+      "grad_norm": 2.7486770153045654,
+      "learning_rate": 1.3692946058091288e-05,
+      "loss": 0.4218,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9475555555555556,
+      "grad_norm": 2.4107203483581543,
+      "learning_rate": 1.3688994269907134e-05,
+      "loss": 0.277,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9481481481481482,
+      "grad_norm": 1.5365214347839355,
+      "learning_rate": 1.368504248172298e-05,
+      "loss": 0.3195,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9487407407407408,
+      "grad_norm": 2.01750111579895,
+      "learning_rate": 1.3681090693538828e-05,
+      "loss": 0.3319,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9493333333333334,
+      "grad_norm": 1.509074091911316,
+      "learning_rate": 1.3677138905354675e-05,
+      "loss": 0.3191,
+      "step": 1602
+    },
+    {
+      "epoch": 0.949925925925926,
+      "grad_norm": 2.720151424407959,
+      "learning_rate": 1.367318711717052e-05,
+      "loss": 0.3594,
+      "step": 1603
+    },
+    {
+      "epoch": 0.9505185185185185,
+      "grad_norm": 1.6067456007003784,
+      "learning_rate": 1.3669235328986367e-05,
+      "loss": 0.2744,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9511111111111111,
+      "grad_norm": 2.0137879848480225,
+      "learning_rate": 1.3665283540802215e-05,
+      "loss": 0.3267,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9517037037037037,
+      "grad_norm": 1.5869941711425781,
+      "learning_rate": 1.3661331752618061e-05,
+      "loss": 0.3169,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9522962962962963,
+      "grad_norm": 1.958805799484253,
+      "learning_rate": 1.3657379964433908e-05,
+      "loss": 0.4021,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9528888888888889,
+      "grad_norm": 1.6996593475341797,
+      "learning_rate": 1.3653428176249754e-05,
+      "loss": 0.3627,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9534814814814815,
+      "grad_norm": 2.5950467586517334,
+      "learning_rate": 1.3649476388065602e-05,
+      "loss": 0.41,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9540740740740741,
+      "grad_norm": 2.288168430328369,
+      "learning_rate": 1.3645524599881448e-05,
+      "loss": 0.4085,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9546666666666667,
+      "grad_norm": 1.9445436000823975,
+      "learning_rate": 1.3641572811697294e-05,
+      "loss": 0.3854,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9552592592592593,
+      "grad_norm": 2.369734048843384,
+      "learning_rate": 1.363762102351314e-05,
+      "loss": 0.4685,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9558518518518518,
+      "grad_norm": 2.1718034744262695,
+      "learning_rate": 1.3633669235328988e-05,
+      "loss": 0.4968,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9564444444444444,
+      "grad_norm": 3.3376498222351074,
+      "learning_rate": 1.3629717447144835e-05,
+      "loss": 0.4876,
+      "step": 1614
+    },
+    {
+      "epoch": 0.957037037037037,
+      "grad_norm": 1.6046192646026611,
+      "learning_rate": 1.362576565896068e-05,
+      "loss": 0.33,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9576296296296296,
+      "grad_norm": 1.6745736598968506,
+      "learning_rate": 1.3621813870776529e-05,
+      "loss": 0.3355,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9582222222222222,
+      "grad_norm": 1.7163114547729492,
+      "learning_rate": 1.3617862082592375e-05,
+      "loss": 0.2774,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9588148148148148,
+      "grad_norm": 1.7283052206039429,
+      "learning_rate": 1.3613910294408221e-05,
+      "loss": 0.4033,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9594074074074074,
+      "grad_norm": 2.1650280952453613,
+      "learning_rate": 1.3609958506224067e-05,
+      "loss": 0.4589,
+      "step": 1619
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9469177722930908,
+      "learning_rate": 1.3606006718039915e-05,
+      "loss": 0.3371,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9605925925925926,
+      "grad_norm": 2.1861088275909424,
+      "learning_rate": 1.3602054929855762e-05,
+      "loss": 0.393,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9611851851851851,
+      "grad_norm": 3.6768798828125,
+      "learning_rate": 1.3598103141671608e-05,
+      "loss": 0.6151,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9617777777777777,
+      "grad_norm": 5.292967796325684,
+      "learning_rate": 1.3594151353487454e-05,
+      "loss": 0.4944,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9623703703703703,
+      "grad_norm": 2.5039379596710205,
+      "learning_rate": 1.3590199565303302e-05,
+      "loss": 0.4798,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 2.106205463409424,
+      "learning_rate": 1.3586247777119148e-05,
+      "loss": 0.4558,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9635555555555556,
+      "grad_norm": 2.0037834644317627,
+      "learning_rate": 1.3582295988934995e-05,
+      "loss": 0.3389,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9641481481481482,
+      "grad_norm": 2.0743606090545654,
+      "learning_rate": 1.357834420075084e-05,
+      "loss": 0.2899,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9647407407407408,
+      "grad_norm": 2.174769401550293,
+      "learning_rate": 1.3574392412566689e-05,
+      "loss": 0.3146,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9653333333333334,
+      "grad_norm": 1.9437662363052368,
+      "learning_rate": 1.3570440624382535e-05,
+      "loss": 0.3388,
+      "step": 1629
+    },
+    {
+      "epoch": 0.965925925925926,
+      "grad_norm": 1.6727485656738281,
+      "learning_rate": 1.3566488836198381e-05,
+      "loss": 0.3283,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9665185185185186,
+      "grad_norm": 2.011276960372925,
+      "learning_rate": 1.3562537048014227e-05,
+      "loss": 0.318,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9671111111111111,
+      "grad_norm": 1.899776816368103,
+      "learning_rate": 1.3558585259830075e-05,
+      "loss": 0.3718,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9677037037037037,
+      "grad_norm": 1.6111485958099365,
+      "learning_rate": 1.3554633471645922e-05,
+      "loss": 0.3439,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9682962962962963,
+      "grad_norm": 1.9891409873962402,
+      "learning_rate": 1.3550681683461768e-05,
+      "loss": 0.3689,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9688888888888889,
+      "grad_norm": 1.9124521017074585,
+      "learning_rate": 1.3546729895277614e-05,
+      "loss": 0.2838,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9694814814814815,
+      "grad_norm": 2.517293930053711,
+      "learning_rate": 1.3542778107093462e-05,
+      "loss": 0.3008,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9700740740740741,
+      "grad_norm": 1.9142873287200928,
+      "learning_rate": 1.3538826318909308e-05,
+      "loss": 0.3148,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9706666666666667,
+      "grad_norm": 1.940642237663269,
+      "learning_rate": 1.3534874530725155e-05,
+      "loss": 0.3448,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9712592592592593,
+      "grad_norm": 2.1416375637054443,
+      "learning_rate": 1.3530922742541003e-05,
+      "loss": 0.3721,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9718518518518519,
+      "grad_norm": 5.717188835144043,
+      "learning_rate": 1.3526970954356849e-05,
+      "loss": 0.6531,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9724444444444444,
+      "grad_norm": 2.409088373184204,
+      "learning_rate": 1.3523019166172695e-05,
+      "loss": 0.398,
+      "step": 1641
+    },
+    {
+      "epoch": 0.973037037037037,
+      "grad_norm": 3.443333387374878,
+      "learning_rate": 1.3519067377988541e-05,
+      "loss": 0.3365,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9736296296296296,
+      "grad_norm": 2.441636800765991,
+      "learning_rate": 1.351511558980439e-05,
+      "loss": 0.3783,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9742222222222222,
+      "grad_norm": 1.9496097564697266,
+      "learning_rate": 1.3511163801620235e-05,
+      "loss": 0.3494,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9748148148148148,
+      "grad_norm": 1.797947883605957,
+      "learning_rate": 1.3507212013436082e-05,
+      "loss": 0.4149,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9754074074074074,
+      "grad_norm": 3.0423450469970703,
+      "learning_rate": 1.3503260225251928e-05,
+      "loss": 0.5294,
+      "step": 1646
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 2.0449111461639404,
+      "learning_rate": 1.3499308437067776e-05,
+      "loss": 0.426,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9765925925925926,
+      "grad_norm": 2.6027724742889404,
+      "learning_rate": 1.3495356648883622e-05,
+      "loss": 0.345,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9771851851851852,
+      "grad_norm": 2.178757905960083,
+      "learning_rate": 1.3491404860699468e-05,
+      "loss": 0.3699,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 2.059175968170166,
+      "learning_rate": 1.3487453072515315e-05,
+      "loss": 0.4047,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9783703703703703,
+      "grad_norm": 1.6668440103530884,
+      "learning_rate": 1.3483501284331163e-05,
+      "loss": 0.3394,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9789629629629629,
+      "grad_norm": 1.3882774114608765,
+      "learning_rate": 1.3479549496147009e-05,
+      "loss": 0.3077,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9795555555555555,
+      "grad_norm": 1.6716232299804688,
+      "learning_rate": 1.3475597707962855e-05,
+      "loss": 0.3043,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9801481481481481,
+      "grad_norm": 3.099900484085083,
+      "learning_rate": 1.3471645919778701e-05,
+      "loss": 0.259,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9807407407407407,
+      "grad_norm": 1.5654165744781494,
+      "learning_rate": 1.346769413159455e-05,
+      "loss": 0.2362,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9813333333333333,
+      "grad_norm": 2.3863437175750732,
+      "learning_rate": 1.3463742343410395e-05,
+      "loss": 0.3308,
+      "step": 1656
+    },
+    {
+      "epoch": 0.981925925925926,
+      "grad_norm": 3.4624874591827393,
+      "learning_rate": 1.3459790555226242e-05,
+      "loss": 0.4794,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9825185185185186,
+      "grad_norm": 1.6766304969787598,
+      "learning_rate": 1.3455838767042088e-05,
+      "loss": 0.3278,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9831111111111112,
+      "grad_norm": 2.325833797454834,
+      "learning_rate": 1.3451886978857936e-05,
+      "loss": 0.3886,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9837037037037037,
+      "grad_norm": 2.014470338821411,
+      "learning_rate": 1.3447935190673782e-05,
+      "loss": 0.4857,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9842962962962963,
+      "grad_norm": 1.7585229873657227,
+      "learning_rate": 1.3443983402489628e-05,
+      "loss": 0.2721,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9848888888888889,
+      "grad_norm": 1.8703911304473877,
+      "learning_rate": 1.3440031614305476e-05,
+      "loss": 0.3657,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9854814814814815,
+      "grad_norm": 1.7473621368408203,
+      "learning_rate": 1.3436079826121319e-05,
+      "loss": 0.3719,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9860740740740741,
+      "grad_norm": 2.563368558883667,
+      "learning_rate": 1.3432128037937167e-05,
+      "loss": 0.3574,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9866666666666667,
+      "grad_norm": 1.9676768779754639,
+      "learning_rate": 1.3428176249753013e-05,
+      "loss": 0.3906,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9872592592592593,
+      "grad_norm": 2.1842663288116455,
+      "learning_rate": 1.342422446156886e-05,
+      "loss": 0.4447,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9878518518518519,
+      "grad_norm": 2.120403289794922,
+      "learning_rate": 1.3420272673384706e-05,
+      "loss": 0.4901,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9884444444444445,
+      "grad_norm": 2.085205078125,
+      "learning_rate": 1.3416320885200554e-05,
+      "loss": 0.3411,
+      "step": 1668
+    },
+    {
+      "epoch": 0.989037037037037,
+      "grad_norm": 2.151130199432373,
+      "learning_rate": 1.34123690970164e-05,
+      "loss": 0.449,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9896296296296296,
+      "grad_norm": 1.674979329109192,
+      "learning_rate": 1.3408417308832246e-05,
+      "loss": 0.273,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9902222222222222,
+      "grad_norm": 2.493623971939087,
+      "learning_rate": 1.3404465520648094e-05,
+      "loss": 0.3784,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9908148148148148,
+      "grad_norm": 1.7574785947799683,
+      "learning_rate": 1.340051373246394e-05,
+      "loss": 0.2512,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9914074074074074,
+      "grad_norm": 1.7129690647125244,
+      "learning_rate": 1.3396561944279787e-05,
+      "loss": 0.3178,
+      "step": 1673
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 2.0228893756866455,
+      "learning_rate": 1.3392610156095633e-05,
+      "loss": 0.3093,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9925925925925926,
+      "grad_norm": 2.0579919815063477,
+      "learning_rate": 1.338865836791148e-05,
+      "loss": 0.329,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9931851851851852,
+      "grad_norm": 2.565741777420044,
+      "learning_rate": 1.3384706579727327e-05,
+      "loss": 0.5066,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9937777777777778,
+      "grad_norm": 1.8564770221710205,
+      "learning_rate": 1.3380754791543173e-05,
+      "loss": 0.3689,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9943703703703703,
+      "grad_norm": 1.7160310745239258,
+      "learning_rate": 1.337680300335902e-05,
+      "loss": 0.3518,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9949629629629629,
+      "grad_norm": 1.6673554182052612,
+      "learning_rate": 1.3372851215174867e-05,
+      "loss": 0.4296,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9955555555555555,
+      "grad_norm": 1.6364470720291138,
+      "learning_rate": 1.3368899426990714e-05,
+      "loss": 0.3967,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9961481481481481,
+      "grad_norm": 2.3556740283966064,
+      "learning_rate": 1.336494763880656e-05,
+      "loss": 0.2926,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9967407407407407,
+      "grad_norm": 1.668472409248352,
+      "learning_rate": 1.3360995850622406e-05,
+      "loss": 0.2166,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9973333333333333,
+      "grad_norm": 1.7163176536560059,
+      "learning_rate": 1.3357044062438254e-05,
+      "loss": 0.4401,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9979259259259259,
+      "grad_norm": 2.0760464668273926,
+      "learning_rate": 1.33530922742541e-05,
+      "loss": 0.3451,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9985185185185185,
+      "grad_norm": 1.814353346824646,
+      "learning_rate": 1.3349140486069947e-05,
+      "loss": 0.3625,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9991111111111111,
+      "grad_norm": 2.025660514831543,
+      "learning_rate": 1.3345188697885793e-05,
+      "loss": 0.4021,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9997037037037036,
+      "grad_norm": 2.3432743549346924,
+      "learning_rate": 1.334123690970164e-05,
+      "loss": 0.3083,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9997037037037036,
+      "eval_loss": 0.3622141182422638,
+      "eval_runtime": 165.5115,
+      "eval_samples_per_second": 36.251,
+      "eval_steps_per_second": 36.251,
+      "step": 1687
+    },
+    {
+      "epoch": 1.0002962962962962,
+      "grad_norm": 1.5975393056869507,
+      "learning_rate": 1.3337285121517487e-05,
+      "loss": 0.2761,
+      "step": 1688
+    },
+    {
+      "epoch": 1.000888888888889,
+      "grad_norm": 1.730049967765808,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.2664,
+      "step": 1689
+    },
+    {
+      "epoch": 1.0014814814814814,
+      "grad_norm": 4.199815273284912,
+      "learning_rate": 1.332938154514918e-05,
+      "loss": 0.4907,
+      "step": 1690
+    },
+    {
+      "epoch": 1.0020740740740741,
+      "grad_norm": 1.8063157796859741,
+      "learning_rate": 1.3325429756965027e-05,
+      "loss": 0.3227,
+      "step": 1691
+    },
+    {
+      "epoch": 1.0026666666666666,
+      "grad_norm": 3.144071340560913,
+      "learning_rate": 1.3321477968780874e-05,
+      "loss": 0.3777,
+      "step": 1692
+    },
+    {
+      "epoch": 1.0032592592592593,
+      "grad_norm": 1.6420443058013916,
+      "learning_rate": 1.331752618059672e-05,
+      "loss": 0.3637,
+      "step": 1693
+    },
+    {
+      "epoch": 1.0038518518518518,
+      "grad_norm": 1.955870270729065,
+      "learning_rate": 1.3313574392412568e-05,
+      "loss": 0.3555,
+      "step": 1694
+    },
+    {
+      "epoch": 1.0044444444444445,
+      "grad_norm": 2.379291296005249,
+      "learning_rate": 1.3309622604228414e-05,
+      "loss": 0.3364,
+      "step": 1695
+    },
+    {
+      "epoch": 1.005037037037037,
+      "grad_norm": 4.130725860595703,
+      "learning_rate": 1.330567081604426e-05,
+      "loss": 0.3758,
+      "step": 1696
+    },
+    {
+      "epoch": 1.0056296296296297,
+      "grad_norm": 1.3443374633789062,
+      "learning_rate": 1.3301719027860107e-05,
+      "loss": 0.2902,
+      "step": 1697
+    },
+    {
+      "epoch": 1.0062222222222221,
+      "grad_norm": 1.4191011190414429,
+      "learning_rate": 1.3297767239675955e-05,
+      "loss": 0.2898,
+      "step": 1698
+    },
+    {
+      "epoch": 1.0068148148148148,
+      "grad_norm": 2.4037094116210938,
+      "learning_rate": 1.32938154514918e-05,
+      "loss": 0.3966,
+      "step": 1699
+    },
+    {
+      "epoch": 1.0074074074074073,
+      "grad_norm": 1.4890761375427246,
+      "learning_rate": 1.3289863663307647e-05,
+      "loss": 0.3236,
+      "step": 1700
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 3.1404643058776855,
+      "learning_rate": 1.3285911875123493e-05,
+      "loss": 0.4053,
+      "step": 1701
+    },
+    {
+      "epoch": 1.0085925925925925,
+      "grad_norm": 2.697660207748413,
+      "learning_rate": 1.3281960086939341e-05,
+      "loss": 0.2989,
+      "step": 1702
+    },
+    {
+      "epoch": 1.0091851851851852,
+      "grad_norm": 2.6603684425354004,
+      "learning_rate": 1.3278008298755187e-05,
+      "loss": 0.4077,
+      "step": 1703
+    },
+    {
+      "epoch": 1.0097777777777779,
+      "grad_norm": 2.2935678958892822,
+      "learning_rate": 1.3274056510571034e-05,
+      "loss": 0.3971,
+      "step": 1704
+    },
+    {
+      "epoch": 1.0103703703703704,
+      "grad_norm": 1.7843064069747925,
+      "learning_rate": 1.327010472238688e-05,
+      "loss": 0.3788,
+      "step": 1705
+    },
+    {
+      "epoch": 1.010962962962963,
+      "grad_norm": 2.375601053237915,
+      "learning_rate": 1.3266152934202728e-05,
+      "loss": 0.3808,
+      "step": 1706
+    },
+    {
+      "epoch": 1.0115555555555555,
+      "grad_norm": 4.416742324829102,
+      "learning_rate": 1.3262201146018574e-05,
+      "loss": 0.6023,
+      "step": 1707
+    },
+    {
+      "epoch": 1.0121481481481482,
+      "grad_norm": 3.0418670177459717,
+      "learning_rate": 1.325824935783442e-05,
+      "loss": 0.4348,
+      "step": 1708
+    },
+    {
+      "epoch": 1.0127407407407407,
+      "grad_norm": 2.2529866695404053,
+      "learning_rate": 1.3254297569650267e-05,
+      "loss": 0.3559,
+      "step": 1709
+    },
+    {
+      "epoch": 1.0133333333333334,
+      "grad_norm": 2.646756410598755,
+      "learning_rate": 1.3250345781466115e-05,
+      "loss": 0.401,
+      "step": 1710
+    },
+    {
+      "epoch": 1.013925925925926,
+      "grad_norm": 2.8170294761657715,
+      "learning_rate": 1.324639399328196e-05,
+      "loss": 0.3705,
+      "step": 1711
+    },
+    {
+      "epoch": 1.0145185185185186,
+      "grad_norm": 2.1138737201690674,
+      "learning_rate": 1.3242442205097807e-05,
+      "loss": 0.3245,
+      "step": 1712
+    },
+    {
+      "epoch": 1.015111111111111,
+      "grad_norm": 1.812298059463501,
+      "learning_rate": 1.3238490416913653e-05,
+      "loss": 0.2915,
+      "step": 1713
+    },
+    {
+      "epoch": 1.0157037037037038,
+      "grad_norm": 1.8100494146347046,
+      "learning_rate": 1.3234538628729501e-05,
+      "loss": 0.3285,
+      "step": 1714
+    },
+    {
+      "epoch": 1.0162962962962963,
+      "grad_norm": 1.786063551902771,
+      "learning_rate": 1.3230586840545347e-05,
+      "loss": 0.1852,
+      "step": 1715
+    },
+    {
+      "epoch": 1.016888888888889,
+      "grad_norm": 2.0514869689941406,
+      "learning_rate": 1.3226635052361194e-05,
+      "loss": 0.3027,
+      "step": 1716
+    },
+    {
+      "epoch": 1.0174814814814814,
+      "grad_norm": 1.6924946308135986,
+      "learning_rate": 1.3222683264177042e-05,
+      "loss": 0.2429,
+      "step": 1717
+    },
+    {
+      "epoch": 1.0180740740740741,
+      "grad_norm": 2.1519765853881836,
+      "learning_rate": 1.3218731475992888e-05,
+      "loss": 0.3055,
+      "step": 1718
+    },
+    {
+      "epoch": 1.0186666666666666,
+      "grad_norm": 2.16658616065979,
+      "learning_rate": 1.3214779687808734e-05,
+      "loss": 0.278,
+      "step": 1719
+    },
+    {
+      "epoch": 1.0192592592592593,
+      "grad_norm": 2.6385083198547363,
+      "learning_rate": 1.321082789962458e-05,
+      "loss": 0.3678,
+      "step": 1720
+    },
+    {
+      "epoch": 1.0198518518518518,
+      "grad_norm": 2.0564777851104736,
+      "learning_rate": 1.3206876111440428e-05,
+      "loss": 0.4239,
+      "step": 1721
+    },
+    {
+      "epoch": 1.0204444444444445,
+      "grad_norm": 2.344243049621582,
+      "learning_rate": 1.3202924323256275e-05,
+      "loss": 0.2918,
+      "step": 1722
+    },
+    {
+      "epoch": 1.021037037037037,
+      "grad_norm": 2.208996057510376,
+      "learning_rate": 1.319897253507212e-05,
+      "loss": 0.361,
+      "step": 1723
+    },
+    {
+      "epoch": 1.0216296296296297,
+      "grad_norm": 1.7210078239440918,
+      "learning_rate": 1.3195020746887967e-05,
+      "loss": 0.2381,
+      "step": 1724
+    },
+    {
+      "epoch": 1.0222222222222221,
+      "grad_norm": 1.8177205324172974,
+      "learning_rate": 1.3191068958703815e-05,
+      "loss": 0.3579,
+      "step": 1725
+    },
+    {
+      "epoch": 1.0228148148148148,
+      "grad_norm": 1.4479060173034668,
+      "learning_rate": 1.3187117170519661e-05,
+      "loss": 0.184,
+      "step": 1726
+    },
+    {
+      "epoch": 1.0234074074074073,
+      "grad_norm": 1.6788142919540405,
+      "learning_rate": 1.3183165382335507e-05,
+      "loss": 0.3928,
+      "step": 1727
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 3.1818785667419434,
+      "learning_rate": 1.3179213594151354e-05,
+      "loss": 0.4359,
+      "step": 1728
+    },
+    {
+      "epoch": 1.0245925925925925,
+      "grad_norm": 1.7369277477264404,
+      "learning_rate": 1.3175261805967202e-05,
+      "loss": 0.334,
+      "step": 1729
+    },
+    {
+      "epoch": 1.0251851851851852,
+      "grad_norm": 2.15474271774292,
+      "learning_rate": 1.3171310017783048e-05,
+      "loss": 0.3931,
+      "step": 1730
+    },
+    {
+      "epoch": 1.0257777777777777,
+      "grad_norm": 2.4502861499786377,
+      "learning_rate": 1.3167358229598894e-05,
+      "loss": 0.3733,
+      "step": 1731
+    },
+    {
+      "epoch": 1.0263703703703704,
+      "grad_norm": 2.0976309776306152,
+      "learning_rate": 1.316340644141474e-05,
+      "loss": 0.3423,
+      "step": 1732
+    },
+    {
+      "epoch": 1.0269629629629629,
+      "grad_norm": 1.8920936584472656,
+      "learning_rate": 1.3159454653230588e-05,
+      "loss": 0.4494,
+      "step": 1733
+    },
+    {
+      "epoch": 1.0275555555555556,
+      "grad_norm": 2.3755311965942383,
+      "learning_rate": 1.3155502865046435e-05,
+      "loss": 0.325,
+      "step": 1734
+    },
+    {
+      "epoch": 1.0281481481481483,
+      "grad_norm": 2.158097505569458,
+      "learning_rate": 1.315155107686228e-05,
+      "loss": 0.3035,
+      "step": 1735
+    },
+    {
+      "epoch": 1.0287407407407407,
+      "grad_norm": 2.2832205295562744,
+      "learning_rate": 1.3147599288678127e-05,
+      "loss": 0.2955,
+      "step": 1736
+    },
+    {
+      "epoch": 1.0293333333333334,
+      "grad_norm": 3.1825811862945557,
+      "learning_rate": 1.3143647500493975e-05,
+      "loss": 0.4508,
+      "step": 1737
+    },
+    {
+      "epoch": 1.029925925925926,
+      "grad_norm": 2.6581389904022217,
+      "learning_rate": 1.3139695712309821e-05,
+      "loss": 0.3829,
+      "step": 1738
+    },
+    {
+      "epoch": 1.0305185185185186,
+      "grad_norm": 2.5814218521118164,
+      "learning_rate": 1.3135743924125667e-05,
+      "loss": 0.3008,
+      "step": 1739
+    },
+    {
+      "epoch": 1.031111111111111,
+      "grad_norm": 2.1190433502197266,
+      "learning_rate": 1.3131792135941515e-05,
+      "loss": 0.3453,
+      "step": 1740
+    },
+    {
+      "epoch": 1.0317037037037038,
+      "grad_norm": 2.176351308822632,
+      "learning_rate": 1.3127840347757362e-05,
+      "loss": 0.4115,
+      "step": 1741
+    },
+    {
+      "epoch": 1.0322962962962963,
+      "grad_norm": 1.845057725906372,
+      "learning_rate": 1.3123888559573208e-05,
+      "loss": 0.3322,
+      "step": 1742
+    },
+    {
+      "epoch": 1.032888888888889,
+      "grad_norm": 2.2533206939697266,
+      "learning_rate": 1.3119936771389054e-05,
+      "loss": 0.2736,
+      "step": 1743
+    },
+    {
+      "epoch": 1.0334814814814814,
+      "grad_norm": 2.3348441123962402,
+      "learning_rate": 1.3115984983204902e-05,
+      "loss": 0.3586,
+      "step": 1744
+    },
+    {
+      "epoch": 1.0340740740740741,
+      "grad_norm": 2.6695566177368164,
+      "learning_rate": 1.3112033195020748e-05,
+      "loss": 0.305,
+      "step": 1745
+    },
+    {
+      "epoch": 1.0346666666666666,
+      "grad_norm": 2.115570306777954,
+      "learning_rate": 1.3108081406836595e-05,
+      "loss": 0.282,
+      "step": 1746
+    },
+    {
+      "epoch": 1.0352592592592593,
+      "grad_norm": 2.1562795639038086,
+      "learning_rate": 1.310412961865244e-05,
+      "loss": 0.331,
+      "step": 1747
+    },
+    {
+      "epoch": 1.0358518518518518,
+      "grad_norm": 1.8694641590118408,
+      "learning_rate": 1.3100177830468289e-05,
+      "loss": 0.3085,
+      "step": 1748
+    },
+    {
+      "epoch": 1.0364444444444445,
+      "grad_norm": 1.7038936614990234,
+      "learning_rate": 1.3096226042284135e-05,
+      "loss": 0.2519,
+      "step": 1749
+    },
+    {
+      "epoch": 1.037037037037037,
+      "grad_norm": 2.0586745738983154,
+      "learning_rate": 1.3092274254099981e-05,
+      "loss": 0.3959,
+      "step": 1750
+    },
+    {
+      "epoch": 1.0376296296296297,
+      "grad_norm": 2.07562255859375,
+      "learning_rate": 1.3088322465915827e-05,
+      "loss": 0.2714,
+      "step": 1751
+    },
+    {
+      "epoch": 1.0382222222222222,
+      "grad_norm": 2.008974075317383,
+      "learning_rate": 1.3084370677731675e-05,
+      "loss": 0.3296,
+      "step": 1752
+    },
+    {
+      "epoch": 1.0388148148148149,
+      "grad_norm": 2.249042510986328,
+      "learning_rate": 1.3080418889547522e-05,
+      "loss": 0.4078,
+      "step": 1753
+    },
+    {
+      "epoch": 1.0394074074074073,
+      "grad_norm": 1.6912949085235596,
+      "learning_rate": 1.3076467101363368e-05,
+      "loss": 0.2501,
+      "step": 1754
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.3434394598007202,
+      "learning_rate": 1.3072515313179214e-05,
+      "loss": 0.2774,
+      "step": 1755
+    },
+    {
+      "epoch": 1.0405925925925925,
+      "grad_norm": 1.9379388093948364,
+      "learning_rate": 1.3068563524995062e-05,
+      "loss": 0.251,
+      "step": 1756
+    },
+    {
+      "epoch": 1.0411851851851852,
+      "grad_norm": 2.0610194206237793,
+      "learning_rate": 1.3064611736810908e-05,
+      "loss": 0.3081,
+      "step": 1757
+    },
+    {
+      "epoch": 1.0417777777777777,
+      "grad_norm": 2.069087028503418,
+      "learning_rate": 1.3060659948626755e-05,
+      "loss": 0.3508,
+      "step": 1758
+    },
+    {
+      "epoch": 1.0423703703703704,
+      "grad_norm": 2.991574287414551,
+      "learning_rate": 1.30567081604426e-05,
+      "loss": 0.4736,
+      "step": 1759
+    },
+    {
+      "epoch": 1.0429629629629629,
+      "grad_norm": 2.0768723487854004,
+      "learning_rate": 1.3052756372258449e-05,
+      "loss": 0.2606,
+      "step": 1760
+    },
+    {
+      "epoch": 1.0435555555555556,
+      "grad_norm": 1.6583445072174072,
+      "learning_rate": 1.3048804584074295e-05,
+      "loss": 0.3034,
+      "step": 1761
+    },
+    {
+      "epoch": 1.044148148148148,
+      "grad_norm": 2.1030757427215576,
+      "learning_rate": 1.3044852795890141e-05,
+      "loss": 0.2768,
+      "step": 1762
+    },
+    {
+      "epoch": 1.0447407407407407,
+      "grad_norm": 2.251863718032837,
+      "learning_rate": 1.3040901007705989e-05,
+      "loss": 0.4229,
+      "step": 1763
+    },
+    {
+      "epoch": 1.0453333333333332,
+      "grad_norm": 1.6404879093170166,
+      "learning_rate": 1.3036949219521835e-05,
+      "loss": 0.2839,
+      "step": 1764
+    },
+    {
+      "epoch": 1.045925925925926,
+      "grad_norm": 1.9033992290496826,
+      "learning_rate": 1.3032997431337682e-05,
+      "loss": 0.3422,
+      "step": 1765
+    },
+    {
+      "epoch": 1.0465185185185186,
+      "grad_norm": 1.664802074432373,
+      "learning_rate": 1.3029045643153528e-05,
+      "loss": 0.2573,
+      "step": 1766
+    },
+    {
+      "epoch": 1.047111111111111,
+      "grad_norm": 1.868331789970398,
+      "learning_rate": 1.3025093854969376e-05,
+      "loss": 0.2604,
+      "step": 1767
+    },
+    {
+      "epoch": 1.0477037037037038,
+      "grad_norm": 2.4397478103637695,
+      "learning_rate": 1.3021142066785222e-05,
+      "loss": 0.319,
+      "step": 1768
+    },
+    {
+      "epoch": 1.0482962962962963,
+      "grad_norm": 1.9023634195327759,
+      "learning_rate": 1.3017190278601068e-05,
+      "loss": 0.256,
+      "step": 1769
+    },
+    {
+      "epoch": 1.048888888888889,
+      "grad_norm": 3.1318371295928955,
+      "learning_rate": 1.3013238490416915e-05,
+      "loss": 0.341,
+      "step": 1770
+    },
+    {
+      "epoch": 1.0494814814814815,
+      "grad_norm": 1.9885759353637695,
+      "learning_rate": 1.3009286702232762e-05,
+      "loss": 0.3119,
+      "step": 1771
+    },
+    {
+      "epoch": 1.0500740740740742,
+      "grad_norm": 3.52811598777771,
+      "learning_rate": 1.3005334914048609e-05,
+      "loss": 0.3164,
+      "step": 1772
+    },
+    {
+      "epoch": 1.0506666666666666,
+      "grad_norm": 1.7433916330337524,
+      "learning_rate": 1.3001383125864455e-05,
+      "loss": 0.2844,
+      "step": 1773
+    },
+    {
+      "epoch": 1.0512592592592593,
+      "grad_norm": 2.8051693439483643,
+      "learning_rate": 1.2997431337680301e-05,
+      "loss": 0.3922,
+      "step": 1774
+    },
+    {
+      "epoch": 1.0518518518518518,
+      "grad_norm": 2.136664628982544,
+      "learning_rate": 1.2993479549496149e-05,
+      "loss": 0.3888,
+      "step": 1775
+    },
+    {
+      "epoch": 1.0524444444444445,
+      "grad_norm": 2.935901403427124,
+      "learning_rate": 1.2989527761311995e-05,
+      "loss": 0.2879,
+      "step": 1776
+    },
+    {
+      "epoch": 1.053037037037037,
+      "grad_norm": 1.6086357831954956,
+      "learning_rate": 1.2985575973127842e-05,
+      "loss": 0.2333,
+      "step": 1777
+    },
+    {
+      "epoch": 1.0536296296296297,
+      "grad_norm": 2.8088557720184326,
+      "learning_rate": 1.2981624184943688e-05,
+      "loss": 0.3504,
+      "step": 1778
+    },
+    {
+      "epoch": 1.0542222222222222,
+      "grad_norm": 2.725360631942749,
+      "learning_rate": 1.2977672396759536e-05,
+      "loss": 0.4801,
+      "step": 1779
+    },
+    {
+      "epoch": 1.0548148148148149,
+      "grad_norm": 1.6681886911392212,
+      "learning_rate": 1.2973720608575382e-05,
+      "loss": 0.2274,
+      "step": 1780
+    },
+    {
+      "epoch": 1.0554074074074073,
+      "grad_norm": 1.7153695821762085,
+      "learning_rate": 1.2969768820391228e-05,
+      "loss": 0.3739,
+      "step": 1781
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 1.7304407358169556,
+      "learning_rate": 1.2965817032207075e-05,
+      "loss": 0.293,
+      "step": 1782
+    },
+    {
+      "epoch": 1.0565925925925925,
+      "grad_norm": 1.8602337837219238,
+      "learning_rate": 1.2961865244022922e-05,
+      "loss": 0.3441,
+      "step": 1783
+    },
+    {
+      "epoch": 1.0571851851851852,
+      "grad_norm": 3.002453327178955,
+      "learning_rate": 1.2957913455838769e-05,
+      "loss": 0.3524,
+      "step": 1784
+    },
+    {
+      "epoch": 1.0577777777777777,
+      "grad_norm": 1.6927968263626099,
+      "learning_rate": 1.2953961667654615e-05,
+      "loss": 0.28,
+      "step": 1785
+    },
+    {
+      "epoch": 1.0583703703703704,
+      "grad_norm": 1.9887079000473022,
+      "learning_rate": 1.2950009879470463e-05,
+      "loss": 0.3071,
+      "step": 1786
+    },
+    {
+      "epoch": 1.0589629629629629,
+      "grad_norm": 1.8570650815963745,
+      "learning_rate": 1.2946058091286309e-05,
+      "loss": 0.3367,
+      "step": 1787
+    },
+    {
+      "epoch": 1.0595555555555556,
+      "grad_norm": 2.031632900238037,
+      "learning_rate": 1.2942106303102155e-05,
+      "loss": 0.3639,
+      "step": 1788
+    },
+    {
+      "epoch": 1.060148148148148,
+      "grad_norm": 1.5636905431747437,
+      "learning_rate": 1.2938154514918002e-05,
+      "loss": 0.2392,
+      "step": 1789
+    },
+    {
+      "epoch": 1.0607407407407408,
+      "grad_norm": 1.7872003316879272,
+      "learning_rate": 1.293420272673385e-05,
+      "loss": 0.29,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0613333333333332,
+      "grad_norm": 1.6676671504974365,
+      "learning_rate": 1.2930250938549696e-05,
+      "loss": 0.2825,
+      "step": 1791
+    },
+    {
+      "epoch": 1.061925925925926,
+      "grad_norm": 2.2727038860321045,
+      "learning_rate": 1.2926299150365542e-05,
+      "loss": 0.2227,
+      "step": 1792
+    },
+    {
+      "epoch": 1.0625185185185184,
+      "grad_norm": 2.576676607131958,
+      "learning_rate": 1.2922347362181388e-05,
+      "loss": 0.3193,
+      "step": 1793
+    },
+    {
+      "epoch": 1.0631111111111111,
+      "grad_norm": 2.583242177963257,
+      "learning_rate": 1.2918395573997236e-05,
+      "loss": 0.3307,
+      "step": 1794
+    },
+    {
+      "epoch": 1.0637037037037036,
+      "grad_norm": 1.5671871900558472,
+      "learning_rate": 1.2914443785813082e-05,
+      "loss": 0.2653,
+      "step": 1795
+    },
+    {
+      "epoch": 1.0642962962962963,
+      "grad_norm": 1.7060147523880005,
+      "learning_rate": 1.2910491997628929e-05,
+      "loss": 0.3076,
+      "step": 1796
+    },
+    {
+      "epoch": 1.064888888888889,
+      "grad_norm": 1.9187260866165161,
+      "learning_rate": 1.2906540209444775e-05,
+      "loss": 0.3018,
+      "step": 1797
+    },
+    {
+      "epoch": 1.0654814814814815,
+      "grad_norm": 2.8953983783721924,
+      "learning_rate": 1.2902588421260623e-05,
+      "loss": 0.3583,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0660740740740742,
+      "grad_norm": 2.1018364429473877,
+      "learning_rate": 1.2898636633076469e-05,
+      "loss": 0.358,
+      "step": 1799
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 1.9201173782348633,
+      "learning_rate": 1.2894684844892315e-05,
+      "loss": 0.2938,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0672592592592594,
+      "grad_norm": 1.9895683526992798,
+      "learning_rate": 1.2890733056708162e-05,
+      "loss": 0.316,
+      "step": 1801
+    },
+    {
+      "epoch": 1.0678518518518518,
+      "grad_norm": 1.395429253578186,
+      "learning_rate": 1.288678126852401e-05,
+      "loss": 0.2057,
+      "step": 1802
+    },
+    {
+      "epoch": 1.0684444444444445,
+      "grad_norm": 2.5708868503570557,
+      "learning_rate": 1.2882829480339856e-05,
+      "loss": 0.3642,
+      "step": 1803
+    },
+    {
+      "epoch": 1.069037037037037,
+      "grad_norm": 2.1280272006988525,
+      "learning_rate": 1.2878877692155702e-05,
+      "loss": 0.3245,
+      "step": 1804
+    },
+    {
+      "epoch": 1.0696296296296297,
+      "grad_norm": 1.2355867624282837,
+      "learning_rate": 1.2874925903971548e-05,
+      "loss": 0.2721,
+      "step": 1805
+    },
+    {
+      "epoch": 1.0702222222222222,
+      "grad_norm": 3.826097249984741,
+      "learning_rate": 1.2870974115787396e-05,
+      "loss": 0.5513,
+      "step": 1806
+    },
+    {
+      "epoch": 1.0708148148148149,
+      "grad_norm": 2.089851140975952,
+      "learning_rate": 1.2867022327603242e-05,
+      "loss": 0.3429,
+      "step": 1807
+    },
+    {
+      "epoch": 1.0714074074074074,
+      "grad_norm": 1.8819403648376465,
+      "learning_rate": 1.2863070539419089e-05,
+      "loss": 0.2668,
+      "step": 1808
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 2.6977930068969727,
+      "learning_rate": 1.2859118751234937e-05,
+      "loss": 0.282,
+      "step": 1809
+    },
+    {
+      "epoch": 1.0725925925925925,
+      "grad_norm": 2.893220901489258,
+      "learning_rate": 1.2855166963050783e-05,
+      "loss": 0.3457,
+      "step": 1810
+    },
+    {
+      "epoch": 1.0731851851851852,
+      "grad_norm": 2.8467774391174316,
+      "learning_rate": 1.2851215174866629e-05,
+      "loss": 0.4388,
+      "step": 1811
+    },
+    {
+      "epoch": 1.0737777777777777,
+      "grad_norm": 1.8851909637451172,
+      "learning_rate": 1.2847263386682475e-05,
+      "loss": 0.3292,
+      "step": 1812
+    },
+    {
+      "epoch": 1.0743703703703704,
+      "grad_norm": 2.0641794204711914,
+      "learning_rate": 1.2843311598498323e-05,
+      "loss": 0.3711,
+      "step": 1813
+    },
+    {
+      "epoch": 1.074962962962963,
+      "grad_norm": 3.6416850090026855,
+      "learning_rate": 1.283935981031417e-05,
+      "loss": 0.37,
+      "step": 1814
+    },
+    {
+      "epoch": 1.0755555555555556,
+      "grad_norm": 2.7931509017944336,
+      "learning_rate": 1.2835408022130016e-05,
+      "loss": 0.3815,
+      "step": 1815
+    },
+    {
+      "epoch": 1.076148148148148,
+      "grad_norm": 1.6958404779434204,
+      "learning_rate": 1.2831456233945862e-05,
+      "loss": 0.2831,
+      "step": 1816
+    },
+    {
+      "epoch": 1.0767407407407408,
+      "grad_norm": 2.0574419498443604,
+      "learning_rate": 1.282750444576171e-05,
+      "loss": 0.2779,
+      "step": 1817
+    },
+    {
+      "epoch": 1.0773333333333333,
+      "grad_norm": 1.6727770566940308,
+      "learning_rate": 1.2823552657577556e-05,
+      "loss": 0.3348,
+      "step": 1818
+    },
+    {
+      "epoch": 1.077925925925926,
+      "grad_norm": 3.555209159851074,
+      "learning_rate": 1.2819600869393402e-05,
+      "loss": 0.3975,
+      "step": 1819
+    },
+    {
+      "epoch": 1.0785185185185184,
+      "grad_norm": 3.727966070175171,
+      "learning_rate": 1.2815649081209249e-05,
+      "loss": 0.4573,
+      "step": 1820
+    },
+    {
+      "epoch": 1.0791111111111111,
+      "grad_norm": 2.7147629261016846,
+      "learning_rate": 1.2811697293025093e-05,
+      "loss": 0.3055,
+      "step": 1821
+    },
+    {
+      "epoch": 1.0797037037037036,
+      "grad_norm": 1.4640800952911377,
+      "learning_rate": 1.2807745504840941e-05,
+      "loss": 0.2078,
+      "step": 1822
+    },
+    {
+      "epoch": 1.0802962962962963,
+      "grad_norm": 1.8691457509994507,
+      "learning_rate": 1.2803793716656787e-05,
+      "loss": 0.3314,
+      "step": 1823
+    },
+    {
+      "epoch": 1.0808888888888888,
+      "grad_norm": 3.008765935897827,
+      "learning_rate": 1.2799841928472634e-05,
+      "loss": 0.3336,
+      "step": 1824
+    },
+    {
+      "epoch": 1.0814814814814815,
+      "grad_norm": 2.273707151412964,
+      "learning_rate": 1.279589014028848e-05,
+      "loss": 0.3619,
+      "step": 1825
+    },
+    {
+      "epoch": 1.082074074074074,
+      "grad_norm": 2.0732288360595703,
+      "learning_rate": 1.2791938352104328e-05,
+      "loss": 0.4266,
+      "step": 1826
+    },
+    {
+      "epoch": 1.0826666666666667,
+      "grad_norm": 3.33415150642395,
+      "learning_rate": 1.2787986563920174e-05,
+      "loss": 0.3775,
+      "step": 1827
+    },
+    {
+      "epoch": 1.0832592592592594,
+      "grad_norm": 2.4406235218048096,
+      "learning_rate": 1.278403477573602e-05,
+      "loss": 0.3131,
+      "step": 1828
+    },
+    {
+      "epoch": 1.0838518518518518,
+      "grad_norm": 1.94198739528656,
+      "learning_rate": 1.2780082987551867e-05,
+      "loss": 0.354,
+      "step": 1829
+    },
+    {
+      "epoch": 1.0844444444444445,
+      "grad_norm": 2.0611324310302734,
+      "learning_rate": 1.2776131199367715e-05,
+      "loss": 0.2979,
+      "step": 1830
+    },
+    {
+      "epoch": 1.085037037037037,
+      "grad_norm": 2.0678794384002686,
+      "learning_rate": 1.277217941118356e-05,
+      "loss": 0.1981,
+      "step": 1831
+    },
+    {
+      "epoch": 1.0856296296296297,
+      "grad_norm": 1.4900182485580444,
+      "learning_rate": 1.2768227622999407e-05,
+      "loss": 0.2699,
+      "step": 1832
+    },
+    {
+      "epoch": 1.0862222222222222,
+      "grad_norm": 2.17694091796875,
+      "learning_rate": 1.2764275834815253e-05,
+      "loss": 0.3282,
+      "step": 1833
+    },
+    {
+      "epoch": 1.086814814814815,
+      "grad_norm": 2.172736883163452,
+      "learning_rate": 1.2760324046631101e-05,
+      "loss": 0.3897,
+      "step": 1834
+    },
+    {
+      "epoch": 1.0874074074074074,
+      "grad_norm": 3.2572712898254395,
+      "learning_rate": 1.2756372258446947e-05,
+      "loss": 0.3742,
+      "step": 1835
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 2.551409959793091,
+      "learning_rate": 1.2752420470262794e-05,
+      "loss": 0.3449,
+      "step": 1836
+    },
+    {
+      "epoch": 1.0885925925925926,
+      "grad_norm": 1.9356296062469482,
+      "learning_rate": 1.274846868207864e-05,
+      "loss": 0.397,
+      "step": 1837
+    },
+    {
+      "epoch": 1.0891851851851853,
+      "grad_norm": 1.5107135772705078,
+      "learning_rate": 1.2744516893894488e-05,
+      "loss": 0.2586,
+      "step": 1838
+    },
+    {
+      "epoch": 1.0897777777777777,
+      "grad_norm": 1.8557285070419312,
+      "learning_rate": 1.2740565105710334e-05,
+      "loss": 0.3473,
+      "step": 1839
+    },
+    {
+      "epoch": 1.0903703703703704,
+      "grad_norm": 2.211355686187744,
+      "learning_rate": 1.273661331752618e-05,
+      "loss": 0.2524,
+      "step": 1840
+    },
+    {
+      "epoch": 1.090962962962963,
+      "grad_norm": 2.0612549781799316,
+      "learning_rate": 1.2732661529342028e-05,
+      "loss": 0.3904,
+      "step": 1841
+    },
+    {
+      "epoch": 1.0915555555555556,
+      "grad_norm": 2.826061725616455,
+      "learning_rate": 1.2728709741157875e-05,
+      "loss": 0.3483,
+      "step": 1842
+    },
+    {
+      "epoch": 1.092148148148148,
+      "grad_norm": 2.166558027267456,
+      "learning_rate": 1.272475795297372e-05,
+      "loss": 0.3706,
+      "step": 1843
+    },
+    {
+      "epoch": 1.0927407407407408,
+      "grad_norm": 2.165097713470459,
+      "learning_rate": 1.2720806164789567e-05,
+      "loss": 0.3126,
+      "step": 1844
+    },
+    {
+      "epoch": 1.0933333333333333,
+      "grad_norm": 1.4105703830718994,
+      "learning_rate": 1.2716854376605415e-05,
+      "loss": 0.2767,
+      "step": 1845
+    },
+    {
+      "epoch": 1.093925925925926,
+      "grad_norm": 2.1482605934143066,
+      "learning_rate": 1.2712902588421261e-05,
+      "loss": 0.3429,
+      "step": 1846
+    },
+    {
+      "epoch": 1.0945185185185184,
+      "grad_norm": 1.823941946029663,
+      "learning_rate": 1.2708950800237107e-05,
+      "loss": 0.2258,
+      "step": 1847
+    },
+    {
+      "epoch": 1.0951111111111111,
+      "grad_norm": 1.7751283645629883,
+      "learning_rate": 1.2704999012052954e-05,
+      "loss": 0.3696,
+      "step": 1848
+    },
+    {
+      "epoch": 1.0957037037037036,
+      "grad_norm": 1.3459594249725342,
+      "learning_rate": 1.2701047223868802e-05,
+      "loss": 0.2395,
+      "step": 1849
+    },
+    {
+      "epoch": 1.0962962962962963,
+      "grad_norm": 1.761081337928772,
+      "learning_rate": 1.2697095435684648e-05,
+      "loss": 0.2074,
+      "step": 1850
+    },
+    {
+      "epoch": 1.0968888888888888,
+      "grad_norm": 1.9211677312850952,
+      "learning_rate": 1.2693143647500494e-05,
+      "loss": 0.3494,
+      "step": 1851
+    },
+    {
+      "epoch": 1.0974814814814815,
+      "grad_norm": 2.153923749923706,
+      "learning_rate": 1.268919185931634e-05,
+      "loss": 0.2766,
+      "step": 1852
+    },
+    {
+      "epoch": 1.098074074074074,
+      "grad_norm": 2.131762742996216,
+      "learning_rate": 1.2685240071132188e-05,
+      "loss": 0.2722,
+      "step": 1853
+    },
+    {
+      "epoch": 1.0986666666666667,
+      "grad_norm": 2.578961133956909,
+      "learning_rate": 1.2681288282948035e-05,
+      "loss": 0.4398,
+      "step": 1854
+    },
+    {
+      "epoch": 1.0992592592592592,
+      "grad_norm": 3.002182722091675,
+      "learning_rate": 1.267733649476388e-05,
+      "loss": 0.4252,
+      "step": 1855
+    },
+    {
+      "epoch": 1.0998518518518519,
+      "grad_norm": 2.205030918121338,
+      "learning_rate": 1.2673384706579727e-05,
+      "loss": 0.301,
+      "step": 1856
+    },
+    {
+      "epoch": 1.1004444444444443,
+      "grad_norm": 1.795343041419983,
+      "learning_rate": 1.2669432918395575e-05,
+      "loss": 0.2634,
+      "step": 1857
+    },
+    {
+      "epoch": 1.101037037037037,
+      "grad_norm": 1.8401360511779785,
+      "learning_rate": 1.2665481130211421e-05,
+      "loss": 0.2833,
+      "step": 1858
+    },
+    {
+      "epoch": 1.1016296296296297,
+      "grad_norm": 2.339531660079956,
+      "learning_rate": 1.2661529342027267e-05,
+      "loss": 0.4447,
+      "step": 1859
+    },
+    {
+      "epoch": 1.1022222222222222,
+      "grad_norm": 2.7905216217041016,
+      "learning_rate": 1.2657577553843114e-05,
+      "loss": 0.3443,
+      "step": 1860
+    },
+    {
+      "epoch": 1.102814814814815,
+      "grad_norm": 1.8456348180770874,
+      "learning_rate": 1.2653625765658962e-05,
+      "loss": 0.2951,
+      "step": 1861
+    },
+    {
+      "epoch": 1.1034074074074074,
+      "grad_norm": 3.3890531063079834,
+      "learning_rate": 1.2649673977474808e-05,
+      "loss": 0.3728,
+      "step": 1862
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 2.684581756591797,
+      "learning_rate": 1.2645722189290654e-05,
+      "loss": 0.3182,
+      "step": 1863
+    },
+    {
+      "epoch": 1.1045925925925926,
+      "grad_norm": 1.6578736305236816,
+      "learning_rate": 1.2641770401106502e-05,
+      "loss": 0.3212,
+      "step": 1864
+    },
+    {
+      "epoch": 1.1051851851851853,
+      "grad_norm": 1.817847728729248,
+      "learning_rate": 1.2637818612922348e-05,
+      "loss": 0.2831,
+      "step": 1865
+    },
+    {
+      "epoch": 1.1057777777777777,
+      "grad_norm": 1.7664296627044678,
+      "learning_rate": 1.2633866824738195e-05,
+      "loss": 0.3555,
+      "step": 1866
+    },
+    {
+      "epoch": 1.1063703703703704,
+      "grad_norm": 2.1410794258117676,
+      "learning_rate": 1.262991503655404e-05,
+      "loss": 0.3593,
+      "step": 1867
+    },
+    {
+      "epoch": 1.106962962962963,
+      "grad_norm": 1.8158884048461914,
+      "learning_rate": 1.2625963248369889e-05,
+      "loss": 0.2581,
+      "step": 1868
+    },
+    {
+      "epoch": 1.1075555555555556,
+      "grad_norm": 2.0211212635040283,
+      "learning_rate": 1.2622011460185735e-05,
+      "loss": 0.2678,
+      "step": 1869
+    },
+    {
+      "epoch": 1.108148148148148,
+      "grad_norm": 1.4682796001434326,
+      "learning_rate": 1.2618059672001581e-05,
+      "loss": 0.2303,
+      "step": 1870
+    },
+    {
+      "epoch": 1.1087407407407408,
+      "grad_norm": 3.4720418453216553,
+      "learning_rate": 1.2614107883817427e-05,
+      "loss": 0.4231,
+      "step": 1871
+    },
+    {
+      "epoch": 1.1093333333333333,
+      "grad_norm": 1.7730364799499512,
+      "learning_rate": 1.2610156095633275e-05,
+      "loss": 0.2565,
+      "step": 1872
+    },
+    {
+      "epoch": 1.109925925925926,
+      "grad_norm": 2.1205031871795654,
+      "learning_rate": 1.2606204307449122e-05,
+      "loss": 0.3163,
+      "step": 1873
+    },
+    {
+      "epoch": 1.1105185185185185,
+      "grad_norm": 1.7159559726715088,
+      "learning_rate": 1.2602252519264968e-05,
+      "loss": 0.2668,
+      "step": 1874
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 1.9449148178100586,
+      "learning_rate": 1.2598300731080814e-05,
+      "loss": 0.3066,
+      "step": 1875
+    },
+    {
+      "epoch": 1.1117037037037036,
+      "grad_norm": 1.3417363166809082,
+      "learning_rate": 1.2594348942896662e-05,
+      "loss": 0.2765,
+      "step": 1876
+    },
+    {
+      "epoch": 1.1122962962962963,
+      "grad_norm": 2.0653650760650635,
+      "learning_rate": 1.2590397154712508e-05,
+      "loss": 0.4214,
+      "step": 1877
+    },
+    {
+      "epoch": 1.1128888888888888,
+      "grad_norm": 2.295518398284912,
+      "learning_rate": 1.2586445366528355e-05,
+      "loss": 0.3227,
+      "step": 1878
+    },
+    {
+      "epoch": 1.1134814814814815,
+      "grad_norm": 1.5910754203796387,
+      "learning_rate": 1.25824935783442e-05,
+      "loss": 0.2432,
+      "step": 1879
+    },
+    {
+      "epoch": 1.114074074074074,
+      "grad_norm": 1.8974294662475586,
+      "learning_rate": 1.2578541790160049e-05,
+      "loss": 0.3639,
+      "step": 1880
+    },
+    {
+      "epoch": 1.1146666666666667,
+      "grad_norm": 1.6099746227264404,
+      "learning_rate": 1.2574590001975895e-05,
+      "loss": 0.3124,
+      "step": 1881
+    },
+    {
+      "epoch": 1.1152592592592592,
+      "grad_norm": 2.0279176235198975,
+      "learning_rate": 1.2570638213791741e-05,
+      "loss": 0.2587,
+      "step": 1882
+    },
+    {
+      "epoch": 1.1158518518518519,
+      "grad_norm": 2.3197193145751953,
+      "learning_rate": 1.2566686425607587e-05,
+      "loss": 0.274,
+      "step": 1883
+    },
+    {
+      "epoch": 1.1164444444444444,
+      "grad_norm": 2.4602913856506348,
+      "learning_rate": 1.2562734637423435e-05,
+      "loss": 0.3557,
+      "step": 1884
+    },
+    {
+      "epoch": 1.117037037037037,
+      "grad_norm": 2.150665760040283,
+      "learning_rate": 1.2558782849239282e-05,
+      "loss": 0.3799,
+      "step": 1885
+    },
+    {
+      "epoch": 1.1176296296296295,
+      "grad_norm": 2.1627111434936523,
+      "learning_rate": 1.2554831061055128e-05,
+      "loss": 0.3349,
+      "step": 1886
+    },
+    {
+      "epoch": 1.1182222222222222,
+      "grad_norm": 1.6806094646453857,
+      "learning_rate": 1.2550879272870976e-05,
+      "loss": 0.251,
+      "step": 1887
+    },
+    {
+      "epoch": 1.1188148148148147,
+      "grad_norm": 3.740407705307007,
+      "learning_rate": 1.2546927484686822e-05,
+      "loss": 0.3094,
+      "step": 1888
+    },
+    {
+      "epoch": 1.1194074074074074,
+      "grad_norm": 2.235539197921753,
+      "learning_rate": 1.2542975696502668e-05,
+      "loss": 0.369,
+      "step": 1889
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.6282660961151123,
+      "learning_rate": 1.2539023908318515e-05,
+      "loss": 0.2636,
+      "step": 1890
+    },
+    {
+      "epoch": 1.1205925925925926,
+      "grad_norm": 1.5923211574554443,
+      "learning_rate": 1.2535072120134362e-05,
+      "loss": 0.3207,
+      "step": 1891
+    },
+    {
+      "epoch": 1.1211851851851853,
+      "grad_norm": 2.2770426273345947,
+      "learning_rate": 1.2531120331950209e-05,
+      "loss": 0.3628,
+      "step": 1892
+    },
+    {
+      "epoch": 1.1217777777777778,
+      "grad_norm": 2.001372814178467,
+      "learning_rate": 1.2527168543766055e-05,
+      "loss": 0.2976,
+      "step": 1893
+    },
+    {
+      "epoch": 1.1223703703703705,
+      "grad_norm": 2.353450059890747,
+      "learning_rate": 1.2523216755581901e-05,
+      "loss": 0.2997,
+      "step": 1894
+    },
+    {
+      "epoch": 1.122962962962963,
+      "grad_norm": 1.6503175497055054,
+      "learning_rate": 1.2519264967397749e-05,
+      "loss": 0.2829,
+      "step": 1895
+    },
+    {
+      "epoch": 1.1235555555555556,
+      "grad_norm": 1.5174890756607056,
+      "learning_rate": 1.2515313179213595e-05,
+      "loss": 0.2365,
+      "step": 1896
+    },
+    {
+      "epoch": 1.1241481481481481,
+      "grad_norm": 2.0546960830688477,
+      "learning_rate": 1.2511361391029442e-05,
+      "loss": 0.3215,
+      "step": 1897
+    },
+    {
+      "epoch": 1.1247407407407408,
+      "grad_norm": 1.5290073156356812,
+      "learning_rate": 1.2507409602845288e-05,
+      "loss": 0.2151,
+      "step": 1898
+    },
+    {
+      "epoch": 1.1253333333333333,
+      "grad_norm": 1.9244683980941772,
+      "learning_rate": 1.2503457814661136e-05,
+      "loss": 0.2873,
+      "step": 1899
+    },
+    {
+      "epoch": 1.125925925925926,
+      "grad_norm": 2.156874418258667,
+      "learning_rate": 1.2499506026476982e-05,
+      "loss": 0.3545,
+      "step": 1900
+    },
+    {
+      "epoch": 1.1265185185185185,
+      "grad_norm": 2.109956741333008,
+      "learning_rate": 1.2495554238292828e-05,
+      "loss": 0.2774,
+      "step": 1901
+    },
+    {
+      "epoch": 1.1271111111111112,
+      "grad_norm": 1.801192045211792,
+      "learning_rate": 1.2491602450108675e-05,
+      "loss": 0.2651,
+      "step": 1902
+    },
+    {
+      "epoch": 1.1277037037037037,
+      "grad_norm": 2.350994825363159,
+      "learning_rate": 1.2487650661924522e-05,
+      "loss": 0.3193,
+      "step": 1903
+    },
+    {
+      "epoch": 1.1282962962962964,
+      "grad_norm": 1.8946239948272705,
+      "learning_rate": 1.2483698873740369e-05,
+      "loss": 0.2688,
+      "step": 1904
+    },
+    {
+      "epoch": 1.1288888888888888,
+      "grad_norm": 1.4050941467285156,
+      "learning_rate": 1.2479747085556215e-05,
+      "loss": 0.2517,
+      "step": 1905
+    },
+    {
+      "epoch": 1.1294814814814815,
+      "grad_norm": 2.3291611671447754,
+      "learning_rate": 1.2475795297372061e-05,
+      "loss": 0.3481,
+      "step": 1906
+    },
+    {
+      "epoch": 1.130074074074074,
+      "grad_norm": 2.467792272567749,
+      "learning_rate": 1.2471843509187909e-05,
+      "loss": 0.3049,
+      "step": 1907
+    },
+    {
+      "epoch": 1.1306666666666667,
+      "grad_norm": 3.052844762802124,
+      "learning_rate": 1.2467891721003755e-05,
+      "loss": 0.3289,
+      "step": 1908
+    },
+    {
+      "epoch": 1.1312592592592592,
+      "grad_norm": 2.384774923324585,
+      "learning_rate": 1.2463939932819602e-05,
+      "loss": 0.4095,
+      "step": 1909
+    },
+    {
+      "epoch": 1.1318518518518519,
+      "grad_norm": 1.7308002710342407,
+      "learning_rate": 1.245998814463545e-05,
+      "loss": 0.3072,
+      "step": 1910
+    },
+    {
+      "epoch": 1.1324444444444444,
+      "grad_norm": 1.3490586280822754,
+      "learning_rate": 1.2456036356451296e-05,
+      "loss": 0.1907,
+      "step": 1911
+    },
+    {
+      "epoch": 1.133037037037037,
+      "grad_norm": 2.2601358890533447,
+      "learning_rate": 1.2452084568267142e-05,
+      "loss": 0.2924,
+      "step": 1912
+    },
+    {
+      "epoch": 1.1336296296296295,
+      "grad_norm": 2.2636375427246094,
+      "learning_rate": 1.2448132780082988e-05,
+      "loss": 0.2804,
+      "step": 1913
+    },
+    {
+      "epoch": 1.1342222222222222,
+      "grad_norm": 2.0308291912078857,
+      "learning_rate": 1.2444180991898836e-05,
+      "loss": 0.2492,
+      "step": 1914
+    },
+    {
+      "epoch": 1.1348148148148147,
+      "grad_norm": 1.9660959243774414,
+      "learning_rate": 1.2440229203714682e-05,
+      "loss": 0.3269,
+      "step": 1915
+    },
+    {
+      "epoch": 1.1354074074074074,
+      "grad_norm": 1.7326987981796265,
+      "learning_rate": 1.2436277415530529e-05,
+      "loss": 0.3011,
+      "step": 1916
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 7.089484691619873,
+      "learning_rate": 1.2432325627346375e-05,
+      "loss": 0.4904,
+      "step": 1917
+    },
+    {
+      "epoch": 1.1365925925925926,
+      "grad_norm": 1.8060382604599,
+      "learning_rate": 1.2428373839162223e-05,
+      "loss": 0.2454,
+      "step": 1918
+    },
+    {
+      "epoch": 1.137185185185185,
+      "grad_norm": 2.45862078666687,
+      "learning_rate": 1.2424422050978069e-05,
+      "loss": 0.3344,
+      "step": 1919
+    },
+    {
+      "epoch": 1.1377777777777778,
+      "grad_norm": 2.1789793968200684,
+      "learning_rate": 1.2420470262793915e-05,
+      "loss": 0.3386,
+      "step": 1920
+    },
+    {
+      "epoch": 1.1383703703703705,
+      "grad_norm": 1.8194804191589355,
+      "learning_rate": 1.2416518474609762e-05,
+      "loss": 0.2756,
+      "step": 1921
+    },
+    {
+      "epoch": 1.138962962962963,
+      "grad_norm": 1.8616869449615479,
+      "learning_rate": 1.241256668642561e-05,
+      "loss": 0.3011,
+      "step": 1922
+    },
+    {
+      "epoch": 1.1395555555555554,
+      "grad_norm": 3.504300117492676,
+      "learning_rate": 1.2408614898241456e-05,
+      "loss": 0.2957,
+      "step": 1923
+    },
+    {
+      "epoch": 1.1401481481481481,
+      "grad_norm": 2.150808334350586,
+      "learning_rate": 1.2404663110057302e-05,
+      "loss": 0.3324,
+      "step": 1924
+    },
+    {
+      "epoch": 1.1407407407407408,
+      "grad_norm": 1.7600092887878418,
+      "learning_rate": 1.2400711321873148e-05,
+      "loss": 0.2354,
+      "step": 1925
+    },
+    {
+      "epoch": 1.1413333333333333,
+      "grad_norm": 2.109464645385742,
+      "learning_rate": 1.2396759533688996e-05,
+      "loss": 0.3629,
+      "step": 1926
+    },
+    {
+      "epoch": 1.141925925925926,
+      "grad_norm": 2.624311685562134,
+      "learning_rate": 1.2392807745504842e-05,
+      "loss": 0.3857,
+      "step": 1927
+    },
+    {
+      "epoch": 1.1425185185185185,
+      "grad_norm": 6.3917460441589355,
+      "learning_rate": 1.2388855957320689e-05,
+      "loss": 0.8785,
+      "step": 1928
+    },
+    {
+      "epoch": 1.1431111111111112,
+      "grad_norm": 1.9344691038131714,
+      "learning_rate": 1.2384904169136535e-05,
+      "loss": 0.3236,
+      "step": 1929
+    },
+    {
+      "epoch": 1.1437037037037037,
+      "grad_norm": 1.9372371435165405,
+      "learning_rate": 1.2380952380952383e-05,
+      "loss": 0.2952,
+      "step": 1930
+    },
+    {
+      "epoch": 1.1442962962962964,
+      "grad_norm": 2.8723034858703613,
+      "learning_rate": 1.2377000592768229e-05,
+      "loss": 0.2922,
+      "step": 1931
+    },
+    {
+      "epoch": 1.1448888888888888,
+      "grad_norm": 2.2677321434020996,
+      "learning_rate": 1.2373048804584075e-05,
+      "loss": 0.3765,
+      "step": 1932
+    },
+    {
+      "epoch": 1.1454814814814815,
+      "grad_norm": 1.2886396646499634,
+      "learning_rate": 1.2369097016399923e-05,
+      "loss": 0.2397,
+      "step": 1933
+    },
+    {
+      "epoch": 1.146074074074074,
+      "grad_norm": 2.265031099319458,
+      "learning_rate": 1.236514522821577e-05,
+      "loss": 0.2381,
+      "step": 1934
+    },
+    {
+      "epoch": 1.1466666666666667,
+      "grad_norm": 3.343719720840454,
+      "learning_rate": 1.2361193440031616e-05,
+      "loss": 0.4305,
+      "step": 1935
+    },
+    {
+      "epoch": 1.1472592592592592,
+      "grad_norm": 3.084427833557129,
+      "learning_rate": 1.2357241651847462e-05,
+      "loss": 0.3685,
+      "step": 1936
+    },
+    {
+      "epoch": 1.147851851851852,
+      "grad_norm": 2.9208555221557617,
+      "learning_rate": 1.235328986366331e-05,
+      "loss": 0.3977,
+      "step": 1937
+    },
+    {
+      "epoch": 1.1484444444444444,
+      "grad_norm": 1.830222249031067,
+      "learning_rate": 1.2349338075479156e-05,
+      "loss": 0.2821,
+      "step": 1938
+    },
+    {
+      "epoch": 1.149037037037037,
+      "grad_norm": 2.0731523036956787,
+      "learning_rate": 1.2345386287295002e-05,
+      "loss": 0.3247,
+      "step": 1939
+    },
+    {
+      "epoch": 1.1496296296296296,
+      "grad_norm": 1.79550302028656,
+      "learning_rate": 1.2341434499110849e-05,
+      "loss": 0.2635,
+      "step": 1940
+    },
+    {
+      "epoch": 1.1502222222222223,
+      "grad_norm": 2.6288888454437256,
+      "learning_rate": 1.2337482710926697e-05,
+      "loss": 0.3351,
+      "step": 1941
+    },
+    {
+      "epoch": 1.1508148148148147,
+      "grad_norm": 1.5039900541305542,
+      "learning_rate": 1.2333530922742543e-05,
+      "loss": 0.2663,
+      "step": 1942
+    },
+    {
+      "epoch": 1.1514074074074074,
+      "grad_norm": 1.9347316026687622,
+      "learning_rate": 1.2329579134558389e-05,
+      "loss": 0.3506,
+      "step": 1943
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 2.579625368118286,
+      "learning_rate": 1.2325627346374235e-05,
+      "loss": 0.2659,
+      "step": 1944
+    },
+    {
+      "epoch": 1.1525925925925926,
+      "grad_norm": 3.8022983074188232,
+      "learning_rate": 1.2321675558190083e-05,
+      "loss": 0.3394,
+      "step": 1945
+    },
+    {
+      "epoch": 1.153185185185185,
+      "grad_norm": 2.901304006576538,
+      "learning_rate": 1.231772377000593e-05,
+      "loss": 0.4226,
+      "step": 1946
+    },
+    {
+      "epoch": 1.1537777777777778,
+      "grad_norm": 1.6792738437652588,
+      "learning_rate": 1.2313771981821776e-05,
+      "loss": 0.2428,
+      "step": 1947
+    },
+    {
+      "epoch": 1.1543703703703705,
+      "grad_norm": 2.4017229080200195,
+      "learning_rate": 1.2309820193637622e-05,
+      "loss": 0.3313,
+      "step": 1948
+    },
+    {
+      "epoch": 1.154962962962963,
+      "grad_norm": 1.7834665775299072,
+      "learning_rate": 1.230586840545347e-05,
+      "loss": 0.2271,
+      "step": 1949
+    },
+    {
+      "epoch": 1.1555555555555554,
+      "grad_norm": 1.975946307182312,
+      "learning_rate": 1.2301916617269316e-05,
+      "loss": 0.3469,
+      "step": 1950
+    },
+    {
+      "epoch": 1.1561481481481481,
+      "grad_norm": 1.6455470323562622,
+      "learning_rate": 1.2297964829085162e-05,
+      "loss": 0.2561,
+      "step": 1951
+    },
+    {
+      "epoch": 1.1567407407407408,
+      "grad_norm": 2.05991530418396,
+      "learning_rate": 1.2294013040901009e-05,
+      "loss": 0.3018,
+      "step": 1952
+    },
+    {
+      "epoch": 1.1573333333333333,
+      "grad_norm": 2.06345534324646,
+      "learning_rate": 1.2290061252716857e-05,
+      "loss": 0.3615,
+      "step": 1953
+    },
+    {
+      "epoch": 1.1579259259259258,
+      "grad_norm": 1.715617060661316,
+      "learning_rate": 1.2286109464532703e-05,
+      "loss": 0.285,
+      "step": 1954
+    },
+    {
+      "epoch": 1.1585185185185185,
+      "grad_norm": 1.4985648393630981,
+      "learning_rate": 1.2282157676348549e-05,
+      "loss": 0.2589,
+      "step": 1955
+    },
+    {
+      "epoch": 1.1591111111111112,
+      "grad_norm": 2.4440648555755615,
+      "learning_rate": 1.2278205888164397e-05,
+      "loss": 0.2605,
+      "step": 1956
+    },
+    {
+      "epoch": 1.1597037037037037,
+      "grad_norm": 2.3640692234039307,
+      "learning_rate": 1.2274254099980243e-05,
+      "loss": 0.3341,
+      "step": 1957
+    },
+    {
+      "epoch": 1.1602962962962964,
+      "grad_norm": 2.054830551147461,
+      "learning_rate": 1.227030231179609e-05,
+      "loss": 0.3354,
+      "step": 1958
+    },
+    {
+      "epoch": 1.1608888888888889,
+      "grad_norm": 2.0688793659210205,
+      "learning_rate": 1.2266350523611936e-05,
+      "loss": 0.3164,
+      "step": 1959
+    },
+    {
+      "epoch": 1.1614814814814816,
+      "grad_norm": 1.9762091636657715,
+      "learning_rate": 1.2262398735427784e-05,
+      "loss": 0.2556,
+      "step": 1960
+    },
+    {
+      "epoch": 1.162074074074074,
+      "grad_norm": 2.693821430206299,
+      "learning_rate": 1.225844694724363e-05,
+      "loss": 0.3327,
+      "step": 1961
+    },
+    {
+      "epoch": 1.1626666666666667,
+      "grad_norm": 2.242702007293701,
+      "learning_rate": 1.2254495159059476e-05,
+      "loss": 0.3221,
+      "step": 1962
+    },
+    {
+      "epoch": 1.1632592592592592,
+      "grad_norm": 3.221470832824707,
+      "learning_rate": 1.2250543370875322e-05,
+      "loss": 0.3425,
+      "step": 1963
+    },
+    {
+      "epoch": 1.163851851851852,
+      "grad_norm": 3.1904783248901367,
+      "learning_rate": 1.224659158269117e-05,
+      "loss": 0.3514,
+      "step": 1964
+    },
+    {
+      "epoch": 1.1644444444444444,
+      "grad_norm": 2.975409984588623,
+      "learning_rate": 1.2242639794507017e-05,
+      "loss": 0.3816,
+      "step": 1965
+    },
+    {
+      "epoch": 1.165037037037037,
+      "grad_norm": 1.992355227470398,
+      "learning_rate": 1.2238688006322863e-05,
+      "loss": 0.3419,
+      "step": 1966
+    },
+    {
+      "epoch": 1.1656296296296296,
+      "grad_norm": 1.5968272686004639,
+      "learning_rate": 1.2234736218138709e-05,
+      "loss": 0.2343,
+      "step": 1967
+    },
+    {
+      "epoch": 1.1662222222222223,
+      "grad_norm": 2.058190107345581,
+      "learning_rate": 1.2230784429954557e-05,
+      "loss": 0.2879,
+      "step": 1968
+    },
+    {
+      "epoch": 1.1668148148148147,
+      "grad_norm": 2.152205467224121,
+      "learning_rate": 1.2226832641770403e-05,
+      "loss": 0.2625,
+      "step": 1969
+    },
+    {
+      "epoch": 1.1674074074074074,
+      "grad_norm": 2.150320291519165,
+      "learning_rate": 1.222288085358625e-05,
+      "loss": 0.2977,
+      "step": 1970
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 2.7342724800109863,
+      "learning_rate": 1.2218929065402096e-05,
+      "loss": 0.3965,
+      "step": 1971
+    },
+    {
+      "epoch": 1.1685925925925926,
+      "grad_norm": 1.77459716796875,
+      "learning_rate": 1.2214977277217944e-05,
+      "loss": 0.343,
+      "step": 1972
+    },
+    {
+      "epoch": 1.169185185185185,
+      "grad_norm": 1.9470113515853882,
+      "learning_rate": 1.221102548903379e-05,
+      "loss": 0.298,
+      "step": 1973
+    },
+    {
+      "epoch": 1.1697777777777778,
+      "grad_norm": 1.9173747301101685,
+      "learning_rate": 1.2207073700849636e-05,
+      "loss": 0.341,
+      "step": 1974
+    },
+    {
+      "epoch": 1.1703703703703703,
+      "grad_norm": 2.2375001907348633,
+      "learning_rate": 1.2203121912665482e-05,
+      "loss": 0.3753,
+      "step": 1975
+    },
+    {
+      "epoch": 1.170962962962963,
+      "grad_norm": 1.5332649946212769,
+      "learning_rate": 1.219917012448133e-05,
+      "loss": 0.2356,
+      "step": 1976
+    },
+    {
+      "epoch": 1.1715555555555555,
+      "grad_norm": 2.086400270462036,
+      "learning_rate": 1.2195218336297177e-05,
+      "loss": 0.2931,
+      "step": 1977
+    },
+    {
+      "epoch": 1.1721481481481482,
+      "grad_norm": 3.1385691165924072,
+      "learning_rate": 1.2191266548113023e-05,
+      "loss": 0.3615,
+      "step": 1978
+    },
+    {
+      "epoch": 1.1727407407407409,
+      "grad_norm": 1.948525071144104,
+      "learning_rate": 1.2187314759928867e-05,
+      "loss": 0.2673,
+      "step": 1979
+    },
+    {
+      "epoch": 1.1733333333333333,
+      "grad_norm": 2.8600330352783203,
+      "learning_rate": 1.2183362971744714e-05,
+      "loss": 0.289,
+      "step": 1980
+    },
+    {
+      "epoch": 1.1739259259259258,
+      "grad_norm": 1.3946692943572998,
+      "learning_rate": 1.2179411183560562e-05,
+      "loss": 0.2321,
+      "step": 1981
+    },
+    {
+      "epoch": 1.1745185185185185,
+      "grad_norm": 1.9421837329864502,
+      "learning_rate": 1.2175459395376408e-05,
+      "loss": 0.1717,
+      "step": 1982
+    },
+    {
+      "epoch": 1.1751111111111112,
+      "grad_norm": 1.7265747785568237,
+      "learning_rate": 1.2171507607192254e-05,
+      "loss": 0.211,
+      "step": 1983
+    },
+    {
+      "epoch": 1.1757037037037037,
+      "grad_norm": 1.9337495565414429,
+      "learning_rate": 1.21675558190081e-05,
+      "loss": 0.3056,
+      "step": 1984
+    },
+    {
+      "epoch": 1.1762962962962962,
+      "grad_norm": 2.8913819789886475,
+      "learning_rate": 1.2163604030823948e-05,
+      "loss": 0.3664,
+      "step": 1985
+    },
+    {
+      "epoch": 1.1768888888888889,
+      "grad_norm": 1.6633763313293457,
+      "learning_rate": 1.2159652242639794e-05,
+      "loss": 0.2918,
+      "step": 1986
+    },
+    {
+      "epoch": 1.1774814814814816,
+      "grad_norm": 2.0304315090179443,
+      "learning_rate": 1.215570045445564e-05,
+      "loss": 0.2791,
+      "step": 1987
+    },
+    {
+      "epoch": 1.178074074074074,
+      "grad_norm": 1.940256118774414,
+      "learning_rate": 1.2151748666271489e-05,
+      "loss": 0.3332,
+      "step": 1988
+    },
+    {
+      "epoch": 1.1786666666666668,
+      "grad_norm": 1.9191358089447021,
+      "learning_rate": 1.2147796878087335e-05,
+      "loss": 0.2889,
+      "step": 1989
+    },
+    {
+      "epoch": 1.1792592592592592,
+      "grad_norm": 3.3052220344543457,
+      "learning_rate": 1.2143845089903181e-05,
+      "loss": 0.3442,
+      "step": 1990
+    },
+    {
+      "epoch": 1.179851851851852,
+      "grad_norm": 2.1602249145507812,
+      "learning_rate": 1.2139893301719027e-05,
+      "loss": 0.3274,
+      "step": 1991
+    },
+    {
+      "epoch": 1.1804444444444444,
+      "grad_norm": 1.7052943706512451,
+      "learning_rate": 1.2135941513534875e-05,
+      "loss": 0.2177,
+      "step": 1992
+    },
+    {
+      "epoch": 1.181037037037037,
+      "grad_norm": 2.00001859664917,
+      "learning_rate": 1.2131989725350722e-05,
+      "loss": 0.3493,
+      "step": 1993
+    },
+    {
+      "epoch": 1.1816296296296296,
+      "grad_norm": 2.4642608165740967,
+      "learning_rate": 1.2128037937166568e-05,
+      "loss": 0.3362,
+      "step": 1994
+    },
+    {
+      "epoch": 1.1822222222222223,
+      "grad_norm": 2.249380350112915,
+      "learning_rate": 1.2124086148982414e-05,
+      "loss": 0.2556,
+      "step": 1995
+    },
+    {
+      "epoch": 1.1828148148148148,
+      "grad_norm": 2.069532632827759,
+      "learning_rate": 1.2120134360798262e-05,
+      "loss": 0.4175,
+      "step": 1996
+    },
+    {
+      "epoch": 1.1834074074074075,
+      "grad_norm": 1.9915450811386108,
+      "learning_rate": 1.2116182572614108e-05,
+      "loss": 0.3186,
+      "step": 1997
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 1.8363286256790161,
+      "learning_rate": 1.2112230784429954e-05,
+      "loss": 0.3319,
+      "step": 1998
+    },
+    {
+      "epoch": 1.1845925925925926,
+      "grad_norm": 1.9424128532409668,
+      "learning_rate": 1.21082789962458e-05,
+      "loss": 0.3212,
+      "step": 1999
+    },
+    {
+      "epoch": 1.1851851851851851,
+      "grad_norm": 1.622134804725647,
+      "learning_rate": 1.2104327208061649e-05,
+      "loss": 0.2572,
+      "step": 2000
+    },
+    {
+      "epoch": 1.1857777777777778,
+      "grad_norm": 1.8149138689041138,
+      "learning_rate": 1.2100375419877495e-05,
+      "loss": 0.3215,
+      "step": 2001
+    },
+    {
+      "epoch": 1.1863703703703703,
+      "grad_norm": 2.6152725219726562,
+      "learning_rate": 1.2096423631693341e-05,
+      "loss": 0.3652,
+      "step": 2002
+    },
+    {
+      "epoch": 1.186962962962963,
+      "grad_norm": 2.0863900184631348,
+      "learning_rate": 1.2092471843509187e-05,
+      "loss": 0.2759,
+      "step": 2003
+    },
+    {
+      "epoch": 1.1875555555555555,
+      "grad_norm": 1.6766539812088013,
+      "learning_rate": 1.2088520055325035e-05,
+      "loss": 0.318,
+      "step": 2004
+    },
+    {
+      "epoch": 1.1881481481481482,
+      "grad_norm": 2.189783811569214,
+      "learning_rate": 1.2084568267140882e-05,
+      "loss": 0.2357,
+      "step": 2005
+    },
+    {
+      "epoch": 1.1887407407407407,
+      "grad_norm": 1.8441532850265503,
+      "learning_rate": 1.2080616478956728e-05,
+      "loss": 0.2346,
+      "step": 2006
+    },
+    {
+      "epoch": 1.1893333333333334,
+      "grad_norm": 1.679923415184021,
+      "learning_rate": 1.2076664690772574e-05,
+      "loss": 0.3615,
+      "step": 2007
+    },
+    {
+      "epoch": 1.1899259259259258,
+      "grad_norm": 2.114039182662964,
+      "learning_rate": 1.2072712902588422e-05,
+      "loss": 0.4546,
+      "step": 2008
+    },
+    {
+      "epoch": 1.1905185185185185,
+      "grad_norm": 2.164518356323242,
+      "learning_rate": 1.2068761114404268e-05,
+      "loss": 0.3951,
+      "step": 2009
+    },
+    {
+      "epoch": 1.1911111111111112,
+      "grad_norm": 2.032184362411499,
+      "learning_rate": 1.2064809326220114e-05,
+      "loss": 0.3135,
+      "step": 2010
+    },
+    {
+      "epoch": 1.1917037037037037,
+      "grad_norm": 2.439563274383545,
+      "learning_rate": 1.2060857538035962e-05,
+      "loss": 0.3717,
+      "step": 2011
+    },
+    {
+      "epoch": 1.1922962962962962,
+      "grad_norm": 2.2781553268432617,
+      "learning_rate": 1.2056905749851809e-05,
+      "loss": 0.3471,
+      "step": 2012
+    },
+    {
+      "epoch": 1.1928888888888889,
+      "grad_norm": 2.4690799713134766,
+      "learning_rate": 1.2052953961667655e-05,
+      "loss": 0.3494,
+      "step": 2013
+    },
+    {
+      "epoch": 1.1934814814814816,
+      "grad_norm": 2.079693555831909,
+      "learning_rate": 1.2049002173483501e-05,
+      "loss": 0.2966,
+      "step": 2014
+    },
+    {
+      "epoch": 1.194074074074074,
+      "grad_norm": 1.982282280921936,
+      "learning_rate": 1.2045050385299349e-05,
+      "loss": 0.3648,
+      "step": 2015
+    },
+    {
+      "epoch": 1.1946666666666665,
+      "grad_norm": 1.9974536895751953,
+      "learning_rate": 1.2041098597115195e-05,
+      "loss": 0.3084,
+      "step": 2016
+    },
+    {
+      "epoch": 1.1952592592592592,
+      "grad_norm": 1.8868674039840698,
+      "learning_rate": 1.2037146808931042e-05,
+      "loss": 0.2567,
+      "step": 2017
+    },
+    {
+      "epoch": 1.195851851851852,
+      "grad_norm": 2.809453248977661,
+      "learning_rate": 1.2033195020746888e-05,
+      "loss": 0.3908,
+      "step": 2018
+    },
+    {
+      "epoch": 1.1964444444444444,
+      "grad_norm": 1.7884621620178223,
+      "learning_rate": 1.2029243232562736e-05,
+      "loss": 0.3812,
+      "step": 2019
+    },
+    {
+      "epoch": 1.1970370370370371,
+      "grad_norm": 1.9826323986053467,
+      "learning_rate": 1.2025291444378582e-05,
+      "loss": 0.3687,
+      "step": 2020
+    },
+    {
+      "epoch": 1.1976296296296296,
+      "grad_norm": 1.482724905014038,
+      "learning_rate": 1.2021339656194428e-05,
+      "loss": 0.2376,
+      "step": 2021
+    },
+    {
+      "epoch": 1.1982222222222223,
+      "grad_norm": 1.6407848596572876,
+      "learning_rate": 1.2017387868010274e-05,
+      "loss": 0.2408,
+      "step": 2022
+    },
+    {
+      "epoch": 1.1988148148148148,
+      "grad_norm": 2.0094785690307617,
+      "learning_rate": 1.2013436079826122e-05,
+      "loss": 0.3401,
+      "step": 2023
+    },
+    {
+      "epoch": 1.1994074074074075,
+      "grad_norm": 1.5058199167251587,
+      "learning_rate": 1.2009484291641969e-05,
+      "loss": 0.3521,
+      "step": 2024
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.942264437675476,
+      "learning_rate": 1.2005532503457815e-05,
+      "loss": 0.268,
+      "step": 2025
+    },
+    {
+      "epoch": 1.2005925925925927,
+      "grad_norm": 2.8799259662628174,
+      "learning_rate": 1.2001580715273661e-05,
+      "loss": 0.3177,
+      "step": 2026
+    },
+    {
+      "epoch": 1.2011851851851851,
+      "grad_norm": 2.1022861003875732,
+      "learning_rate": 1.1997628927089509e-05,
+      "loss": 0.2768,
+      "step": 2027
+    },
+    {
+      "epoch": 1.2017777777777778,
+      "grad_norm": 1.4454976320266724,
+      "learning_rate": 1.1993677138905355e-05,
+      "loss": 0.2098,
+      "step": 2028
+    },
+    {
+      "epoch": 1.2023703703703703,
+      "grad_norm": 1.9615153074264526,
+      "learning_rate": 1.1989725350721202e-05,
+      "loss": 0.315,
+      "step": 2029
+    },
+    {
+      "epoch": 1.202962962962963,
+      "grad_norm": 2.2764129638671875,
+      "learning_rate": 1.1985773562537048e-05,
+      "loss": 0.3353,
+      "step": 2030
+    },
+    {
+      "epoch": 1.2035555555555555,
+      "grad_norm": 1.9288355112075806,
+      "learning_rate": 1.1981821774352896e-05,
+      "loss": 0.2302,
+      "step": 2031
+    },
+    {
+      "epoch": 1.2041481481481482,
+      "grad_norm": 2.0382049083709717,
+      "learning_rate": 1.1977869986168742e-05,
+      "loss": 0.3299,
+      "step": 2032
+    },
+    {
+      "epoch": 1.2047407407407407,
+      "grad_norm": 1.872552514076233,
+      "learning_rate": 1.1973918197984588e-05,
+      "loss": 0.2832,
+      "step": 2033
+    },
+    {
+      "epoch": 1.2053333333333334,
+      "grad_norm": 1.8779420852661133,
+      "learning_rate": 1.1969966409800436e-05,
+      "loss": 0.1933,
+      "step": 2034
+    },
+    {
+      "epoch": 1.2059259259259258,
+      "grad_norm": 2.080366849899292,
+      "learning_rate": 1.1966014621616282e-05,
+      "loss": 0.3691,
+      "step": 2035
+    },
+    {
+      "epoch": 1.2065185185185185,
+      "grad_norm": 2.7163209915161133,
+      "learning_rate": 1.1962062833432129e-05,
+      "loss": 0.3289,
+      "step": 2036
+    },
+    {
+      "epoch": 1.207111111111111,
+      "grad_norm": 2.8525331020355225,
+      "learning_rate": 1.1958111045247975e-05,
+      "loss": 0.2777,
+      "step": 2037
+    },
+    {
+      "epoch": 1.2077037037037037,
+      "grad_norm": 2.1516590118408203,
+      "learning_rate": 1.1954159257063823e-05,
+      "loss": 0.3614,
+      "step": 2038
+    },
+    {
+      "epoch": 1.2082962962962962,
+      "grad_norm": 2.296645164489746,
+      "learning_rate": 1.1950207468879669e-05,
+      "loss": 0.4067,
+      "step": 2039
+    },
+    {
+      "epoch": 1.208888888888889,
+      "grad_norm": 1.327991247177124,
+      "learning_rate": 1.1946255680695515e-05,
+      "loss": 0.2333,
+      "step": 2040
+    },
+    {
+      "epoch": 1.2094814814814816,
+      "grad_norm": 1.721838116645813,
+      "learning_rate": 1.1942303892511362e-05,
+      "loss": 0.2366,
+      "step": 2041
+    },
+    {
+      "epoch": 1.210074074074074,
+      "grad_norm": 1.8459808826446533,
+      "learning_rate": 1.193835210432721e-05,
+      "loss": 0.3345,
+      "step": 2042
+    },
+    {
+      "epoch": 1.2106666666666666,
+      "grad_norm": 2.017134428024292,
+      "learning_rate": 1.1934400316143056e-05,
+      "loss": 0.2214,
+      "step": 2043
+    },
+    {
+      "epoch": 1.2112592592592593,
+      "grad_norm": 1.9770865440368652,
+      "learning_rate": 1.1930448527958902e-05,
+      "loss": 0.2493,
+      "step": 2044
+    },
+    {
+      "epoch": 1.211851851851852,
+      "grad_norm": 1.7503596544265747,
+      "learning_rate": 1.1926496739774748e-05,
+      "loss": 0.2888,
+      "step": 2045
+    },
+    {
+      "epoch": 1.2124444444444444,
+      "grad_norm": 1.6665823459625244,
+      "learning_rate": 1.1922544951590596e-05,
+      "loss": 0.2853,
+      "step": 2046
+    },
+    {
+      "epoch": 1.213037037037037,
+      "grad_norm": 1.6836087703704834,
+      "learning_rate": 1.1918593163406442e-05,
+      "loss": 0.2217,
+      "step": 2047
+    },
+    {
+      "epoch": 1.2136296296296296,
+      "grad_norm": 1.9016097784042358,
+      "learning_rate": 1.1914641375222289e-05,
+      "loss": 0.2703,
+      "step": 2048
+    },
+    {
+      "epoch": 1.2142222222222223,
+      "grad_norm": 2.6570065021514893,
+      "learning_rate": 1.1910689587038135e-05,
+      "loss": 0.3719,
+      "step": 2049
+    },
+    {
+      "epoch": 1.2148148148148148,
+      "grad_norm": 2.3471968173980713,
+      "learning_rate": 1.1906737798853983e-05,
+      "loss": 0.3342,
+      "step": 2050
+    },
+    {
+      "epoch": 1.2154074074074075,
+      "grad_norm": 2.426568031311035,
+      "learning_rate": 1.1902786010669829e-05,
+      "loss": 0.3576,
+      "step": 2051
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 2.399594783782959,
+      "learning_rate": 1.1898834222485675e-05,
+      "loss": 0.3331,
+      "step": 2052
+    },
+    {
+      "epoch": 1.2165925925925927,
+      "grad_norm": 1.7044233083724976,
+      "learning_rate": 1.1894882434301522e-05,
+      "loss": 0.2514,
+      "step": 2053
+    },
+    {
+      "epoch": 1.2171851851851851,
+      "grad_norm": 2.7388219833374023,
+      "learning_rate": 1.189093064611737e-05,
+      "loss": 0.2801,
+      "step": 2054
+    },
+    {
+      "epoch": 1.2177777777777778,
+      "grad_norm": 1.7341054677963257,
+      "learning_rate": 1.1886978857933216e-05,
+      "loss": 0.2316,
+      "step": 2055
+    },
+    {
+      "epoch": 1.2183703703703703,
+      "grad_norm": 3.0416414737701416,
+      "learning_rate": 1.1883027069749062e-05,
+      "loss": 0.3351,
+      "step": 2056
+    },
+    {
+      "epoch": 1.218962962962963,
+      "grad_norm": 1.6937378644943237,
+      "learning_rate": 1.187907528156491e-05,
+      "loss": 0.2767,
+      "step": 2057
+    },
+    {
+      "epoch": 1.2195555555555555,
+      "grad_norm": 1.8906358480453491,
+      "learning_rate": 1.1875123493380756e-05,
+      "loss": 0.3185,
+      "step": 2058
+    },
+    {
+      "epoch": 1.2201481481481482,
+      "grad_norm": 3.7203145027160645,
+      "learning_rate": 1.1871171705196602e-05,
+      "loss": 0.3257,
+      "step": 2059
+    },
+    {
+      "epoch": 1.2207407407407407,
+      "grad_norm": 4.838581085205078,
+      "learning_rate": 1.1867219917012449e-05,
+      "loss": 0.2736,
+      "step": 2060
+    },
+    {
+      "epoch": 1.2213333333333334,
+      "grad_norm": 2.214830160140991,
+      "learning_rate": 1.1863268128828297e-05,
+      "loss": 0.2871,
+      "step": 2061
+    },
+    {
+      "epoch": 1.2219259259259259,
+      "grad_norm": 1.7833261489868164,
+      "learning_rate": 1.1859316340644143e-05,
+      "loss": 0.3081,
+      "step": 2062
+    },
+    {
+      "epoch": 1.2225185185185186,
+      "grad_norm": 1.794026494026184,
+      "learning_rate": 1.1855364552459989e-05,
+      "loss": 0.3685,
+      "step": 2063
+    },
+    {
+      "epoch": 1.223111111111111,
+      "grad_norm": 1.863080620765686,
+      "learning_rate": 1.1851412764275835e-05,
+      "loss": 0.3566,
+      "step": 2064
+    },
+    {
+      "epoch": 1.2237037037037037,
+      "grad_norm": 2.0065689086914062,
+      "learning_rate": 1.1847460976091683e-05,
+      "loss": 0.2976,
+      "step": 2065
+    },
+    {
+      "epoch": 1.2242962962962962,
+      "grad_norm": 2.067603826522827,
+      "learning_rate": 1.184350918790753e-05,
+      "loss": 0.4211,
+      "step": 2066
+    },
+    {
+      "epoch": 1.224888888888889,
+      "grad_norm": 2.1765284538269043,
+      "learning_rate": 1.1839557399723376e-05,
+      "loss": 0.305,
+      "step": 2067
+    },
+    {
+      "epoch": 1.2254814814814814,
+      "grad_norm": 2.3467071056365967,
+      "learning_rate": 1.1835605611539222e-05,
+      "loss": 0.3041,
+      "step": 2068
+    },
+    {
+      "epoch": 1.226074074074074,
+      "grad_norm": 1.6216120719909668,
+      "learning_rate": 1.183165382335507e-05,
+      "loss": 0.255,
+      "step": 2069
+    },
+    {
+      "epoch": 1.2266666666666666,
+      "grad_norm": 2.6521317958831787,
+      "learning_rate": 1.1827702035170916e-05,
+      "loss": 0.3177,
+      "step": 2070
+    },
+    {
+      "epoch": 1.2272592592592593,
+      "grad_norm": 2.702881336212158,
+      "learning_rate": 1.1823750246986762e-05,
+      "loss": 0.2823,
+      "step": 2071
+    },
+    {
+      "epoch": 1.227851851851852,
+      "grad_norm": 3.3376212120056152,
+      "learning_rate": 1.1819798458802609e-05,
+      "loss": 0.3835,
+      "step": 2072
+    },
+    {
+      "epoch": 1.2284444444444444,
+      "grad_norm": 1.541361927986145,
+      "learning_rate": 1.1815846670618457e-05,
+      "loss": 0.2472,
+      "step": 2073
+    },
+    {
+      "epoch": 1.229037037037037,
+      "grad_norm": 1.9422844648361206,
+      "learning_rate": 1.1811894882434303e-05,
+      "loss": 0.2936,
+      "step": 2074
+    },
+    {
+      "epoch": 1.2296296296296296,
+      "grad_norm": 2.5493838787078857,
+      "learning_rate": 1.1807943094250149e-05,
+      "loss": 0.3804,
+      "step": 2075
+    },
+    {
+      "epoch": 1.2302222222222223,
+      "grad_norm": 2.2947051525115967,
+      "learning_rate": 1.1803991306065995e-05,
+      "loss": 0.3103,
+      "step": 2076
+    },
+    {
+      "epoch": 1.2308148148148148,
+      "grad_norm": 2.511087656021118,
+      "learning_rate": 1.1800039517881843e-05,
+      "loss": 0.2793,
+      "step": 2077
+    },
+    {
+      "epoch": 1.2314074074074073,
+      "grad_norm": 2.0471320152282715,
+      "learning_rate": 1.179608772969769e-05,
+      "loss": 0.324,
+      "step": 2078
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 2.6193580627441406,
+      "learning_rate": 1.1792135941513536e-05,
+      "loss": 0.3698,
+      "step": 2079
+    },
+    {
+      "epoch": 1.2325925925925927,
+      "grad_norm": 1.8163330554962158,
+      "learning_rate": 1.1788184153329384e-05,
+      "loss": 0.3439,
+      "step": 2080
+    },
+    {
+      "epoch": 1.2331851851851852,
+      "grad_norm": 2.3737170696258545,
+      "learning_rate": 1.178423236514523e-05,
+      "loss": 0.3113,
+      "step": 2081
+    },
+    {
+      "epoch": 1.2337777777777779,
+      "grad_norm": 2.9308528900146484,
+      "learning_rate": 1.1780280576961076e-05,
+      "loss": 0.3145,
+      "step": 2082
+    },
+    {
+      "epoch": 1.2343703703703703,
+      "grad_norm": 2.1148390769958496,
+      "learning_rate": 1.1776328788776922e-05,
+      "loss": 0.3186,
+      "step": 2083
+    },
+    {
+      "epoch": 1.234962962962963,
+      "grad_norm": 1.5778768062591553,
+      "learning_rate": 1.177237700059277e-05,
+      "loss": 0.229,
+      "step": 2084
+    },
+    {
+      "epoch": 1.2355555555555555,
+      "grad_norm": 3.3508002758026123,
+      "learning_rate": 1.1768425212408617e-05,
+      "loss": 0.471,
+      "step": 2085
+    },
+    {
+      "epoch": 1.2361481481481482,
+      "grad_norm": 2.1169941425323486,
+      "learning_rate": 1.1764473424224463e-05,
+      "loss": 0.3357,
+      "step": 2086
+    },
+    {
+      "epoch": 1.2367407407407407,
+      "grad_norm": 1.7952910661697388,
+      "learning_rate": 1.1760521636040309e-05,
+      "loss": 0.2877,
+      "step": 2087
+    },
+    {
+      "epoch": 1.2373333333333334,
+      "grad_norm": 2.762404203414917,
+      "learning_rate": 1.1756569847856157e-05,
+      "loss": 0.3817,
+      "step": 2088
+    },
+    {
+      "epoch": 1.2379259259259259,
+      "grad_norm": 2.0377609729766846,
+      "learning_rate": 1.1752618059672003e-05,
+      "loss": 0.2713,
+      "step": 2089
+    },
+    {
+      "epoch": 1.2385185185185186,
+      "grad_norm": 1.680978775024414,
+      "learning_rate": 1.174866627148785e-05,
+      "loss": 0.2952,
+      "step": 2090
+    },
+    {
+      "epoch": 1.239111111111111,
+      "grad_norm": 1.5777020454406738,
+      "learning_rate": 1.1744714483303696e-05,
+      "loss": 0.2566,
+      "step": 2091
+    },
+    {
+      "epoch": 1.2397037037037038,
+      "grad_norm": 2.025729179382324,
+      "learning_rate": 1.1740762695119544e-05,
+      "loss": 0.3988,
+      "step": 2092
+    },
+    {
+      "epoch": 1.2402962962962962,
+      "grad_norm": 1.6417675018310547,
+      "learning_rate": 1.173681090693539e-05,
+      "loss": 0.2977,
+      "step": 2093
+    },
+    {
+      "epoch": 1.240888888888889,
+      "grad_norm": 1.8698111772537231,
+      "learning_rate": 1.1732859118751236e-05,
+      "loss": 0.2963,
+      "step": 2094
+    },
+    {
+      "epoch": 1.2414814814814814,
+      "grad_norm": 2.3008062839508057,
+      "learning_rate": 1.1728907330567082e-05,
+      "loss": 0.2893,
+      "step": 2095
+    },
+    {
+      "epoch": 1.242074074074074,
+      "grad_norm": 2.0490002632141113,
+      "learning_rate": 1.172495554238293e-05,
+      "loss": 0.3006,
+      "step": 2096
+    },
+    {
+      "epoch": 1.2426666666666666,
+      "grad_norm": 2.1534337997436523,
+      "learning_rate": 1.1721003754198777e-05,
+      "loss": 0.2958,
+      "step": 2097
+    },
+    {
+      "epoch": 1.2432592592592593,
+      "grad_norm": 2.2468628883361816,
+      "learning_rate": 1.1717051966014623e-05,
+      "loss": 0.2997,
+      "step": 2098
+    },
+    {
+      "epoch": 1.2438518518518518,
+      "grad_norm": 1.7846730947494507,
+      "learning_rate": 1.1713100177830469e-05,
+      "loss": 0.2446,
+      "step": 2099
+    },
+    {
+      "epoch": 1.2444444444444445,
+      "grad_norm": 2.3387603759765625,
+      "learning_rate": 1.1709148389646317e-05,
+      "loss": 0.3373,
+      "step": 2100
+    },
+    {
+      "epoch": 1.245037037037037,
+      "grad_norm": 1.9664052724838257,
+      "learning_rate": 1.1705196601462163e-05,
+      "loss": 0.3145,
+      "step": 2101
+    },
+    {
+      "epoch": 1.2456296296296296,
+      "grad_norm": 3.1206490993499756,
+      "learning_rate": 1.170124481327801e-05,
+      "loss": 0.2965,
+      "step": 2102
+    },
+    {
+      "epoch": 1.2462222222222223,
+      "grad_norm": 2.5104029178619385,
+      "learning_rate": 1.1697293025093857e-05,
+      "loss": 0.341,
+      "step": 2103
+    },
+    {
+      "epoch": 1.2468148148148148,
+      "grad_norm": 1.6176241636276245,
+      "learning_rate": 1.1693341236909704e-05,
+      "loss": 0.2509,
+      "step": 2104
+    },
+    {
+      "epoch": 1.2474074074074073,
+      "grad_norm": 2.644590139389038,
+      "learning_rate": 1.168938944872555e-05,
+      "loss": 0.3698,
+      "step": 2105
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 2.1236438751220703,
+      "learning_rate": 1.1685437660541396e-05,
+      "loss": 0.3171,
+      "step": 2106
+    },
+    {
+      "epoch": 1.2485925925925927,
+      "grad_norm": 2.662517786026001,
+      "learning_rate": 1.1681485872357244e-05,
+      "loss": 0.2882,
+      "step": 2107
+    },
+    {
+      "epoch": 1.2491851851851852,
+      "grad_norm": 2.585818290710449,
+      "learning_rate": 1.167753408417309e-05,
+      "loss": 0.2949,
+      "step": 2108
+    },
+    {
+      "epoch": 1.2497777777777777,
+      "grad_norm": 1.334082841873169,
+      "learning_rate": 1.1673582295988937e-05,
+      "loss": 0.2073,
+      "step": 2109
+    },
+    {
+      "epoch": 1.2503703703703704,
+      "grad_norm": 3.113915205001831,
+      "learning_rate": 1.1669630507804783e-05,
+      "loss": 0.4224,
+      "step": 2110
+    },
+    {
+      "epoch": 1.250962962962963,
+      "grad_norm": 2.835284471511841,
+      "learning_rate": 1.166567871962063e-05,
+      "loss": 0.3723,
+      "step": 2111
+    },
+    {
+      "epoch": 1.2515555555555555,
+      "grad_norm": 2.0878255367279053,
+      "learning_rate": 1.1661726931436477e-05,
+      "loss": 0.3196,
+      "step": 2112
+    },
+    {
+      "epoch": 1.252148148148148,
+      "grad_norm": 1.7826154232025146,
+      "learning_rate": 1.1657775143252323e-05,
+      "loss": 0.2986,
+      "step": 2113
+    },
+    {
+      "epoch": 1.2527407407407407,
+      "grad_norm": 1.9010252952575684,
+      "learning_rate": 1.165382335506817e-05,
+      "loss": 0.2911,
+      "step": 2114
+    },
+    {
+      "epoch": 1.2533333333333334,
+      "grad_norm": 1.9874550104141235,
+      "learning_rate": 1.1649871566884017e-05,
+      "loss": 0.2671,
+      "step": 2115
+    },
+    {
+      "epoch": 1.2539259259259259,
+      "grad_norm": 2.6405653953552246,
+      "learning_rate": 1.1645919778699864e-05,
+      "loss": 0.3291,
+      "step": 2116
+    },
+    {
+      "epoch": 1.2545185185185186,
+      "grad_norm": 2.9206418991088867,
+      "learning_rate": 1.164196799051571e-05,
+      "loss": 0.3005,
+      "step": 2117
+    },
+    {
+      "epoch": 1.255111111111111,
+      "grad_norm": 2.5896198749542236,
+      "learning_rate": 1.1638016202331556e-05,
+      "loss": 0.4193,
+      "step": 2118
+    },
+    {
+      "epoch": 1.2557037037037038,
+      "grad_norm": 1.8946362733840942,
+      "learning_rate": 1.1634064414147404e-05,
+      "loss": 0.3548,
+      "step": 2119
+    },
+    {
+      "epoch": 1.2562962962962962,
+      "grad_norm": 1.9120051860809326,
+      "learning_rate": 1.163011262596325e-05,
+      "loss": 0.3878,
+      "step": 2120
+    },
+    {
+      "epoch": 1.256888888888889,
+      "grad_norm": 2.4515669345855713,
+      "learning_rate": 1.1626160837779097e-05,
+      "loss": 0.3067,
+      "step": 2121
+    },
+    {
+      "epoch": 1.2574814814814814,
+      "grad_norm": 2.7745296955108643,
+      "learning_rate": 1.1622209049594943e-05,
+      "loss": 0.2578,
+      "step": 2122
+    },
+    {
+      "epoch": 1.2580740740740741,
+      "grad_norm": 1.581356167793274,
+      "learning_rate": 1.161825726141079e-05,
+      "loss": 0.2921,
+      "step": 2123
+    },
+    {
+      "epoch": 1.2586666666666666,
+      "grad_norm": 2.0858731269836426,
+      "learning_rate": 1.1614305473226637e-05,
+      "loss": 0.2684,
+      "step": 2124
+    },
+    {
+      "epoch": 1.2592592592592593,
+      "grad_norm": 2.2976255416870117,
+      "learning_rate": 1.1610353685042483e-05,
+      "loss": 0.3198,
+      "step": 2125
+    },
+    {
+      "epoch": 1.2598518518518518,
+      "grad_norm": 1.9507085084915161,
+      "learning_rate": 1.160640189685833e-05,
+      "loss": 0.2686,
+      "step": 2126
+    },
+    {
+      "epoch": 1.2604444444444445,
+      "grad_norm": 1.4855600595474243,
+      "learning_rate": 1.1602450108674177e-05,
+      "loss": 0.2659,
+      "step": 2127
+    },
+    {
+      "epoch": 1.261037037037037,
+      "grad_norm": 1.6524900197982788,
+      "learning_rate": 1.1598498320490024e-05,
+      "loss": 0.2207,
+      "step": 2128
+    },
+    {
+      "epoch": 1.2616296296296297,
+      "grad_norm": 2.213207244873047,
+      "learning_rate": 1.159454653230587e-05,
+      "loss": 0.3257,
+      "step": 2129
+    },
+    {
+      "epoch": 1.2622222222222224,
+      "grad_norm": 2.4456489086151123,
+      "learning_rate": 1.1590594744121718e-05,
+      "loss": 0.3677,
+      "step": 2130
+    },
+    {
+      "epoch": 1.2628148148148148,
+      "grad_norm": 1.5804141759872437,
+      "learning_rate": 1.1586642955937564e-05,
+      "loss": 0.2717,
+      "step": 2131
+    },
+    {
+      "epoch": 1.2634074074074073,
+      "grad_norm": 2.133525848388672,
+      "learning_rate": 1.158269116775341e-05,
+      "loss": 0.3353,
+      "step": 2132
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 2.662773609161377,
+      "learning_rate": 1.1578739379569257e-05,
+      "loss": 0.3627,
+      "step": 2133
+    },
+    {
+      "epoch": 1.2645925925925927,
+      "grad_norm": 1.7749534845352173,
+      "learning_rate": 1.1574787591385104e-05,
+      "loss": 0.2761,
+      "step": 2134
+    },
+    {
+      "epoch": 1.2651851851851852,
+      "grad_norm": 2.442610263824463,
+      "learning_rate": 1.157083580320095e-05,
+      "loss": 0.4373,
+      "step": 2135
+    },
+    {
+      "epoch": 1.2657777777777777,
+      "grad_norm": 1.488034725189209,
+      "learning_rate": 1.1566884015016797e-05,
+      "loss": 0.271,
+      "step": 2136
+    },
+    {
+      "epoch": 1.2663703703703704,
+      "grad_norm": 3.44329571723938,
+      "learning_rate": 1.1562932226832643e-05,
+      "loss": 0.3998,
+      "step": 2137
+    },
+    {
+      "epoch": 1.266962962962963,
+      "grad_norm": 2.324214458465576,
+      "learning_rate": 1.1558980438648488e-05,
+      "loss": 0.3589,
+      "step": 2138
+    },
+    {
+      "epoch": 1.2675555555555555,
+      "grad_norm": 2.8454911708831787,
+      "learning_rate": 1.1555028650464336e-05,
+      "loss": 0.3148,
+      "step": 2139
+    },
+    {
+      "epoch": 1.268148148148148,
+      "grad_norm": 3.3585026264190674,
+      "learning_rate": 1.1551076862280182e-05,
+      "loss": 0.3354,
+      "step": 2140
+    },
+    {
+      "epoch": 1.2687407407407407,
+      "grad_norm": 2.0296404361724854,
+      "learning_rate": 1.1547125074096028e-05,
+      "loss": 0.2422,
+      "step": 2141
+    },
+    {
+      "epoch": 1.2693333333333334,
+      "grad_norm": 2.0767674446105957,
+      "learning_rate": 1.1543173285911874e-05,
+      "loss": 0.2943,
+      "step": 2142
+    },
+    {
+      "epoch": 1.269925925925926,
+      "grad_norm": 1.520619511604309,
+      "learning_rate": 1.1539221497727722e-05,
+      "loss": 0.2419,
+      "step": 2143
+    },
+    {
+      "epoch": 1.2705185185185184,
+      "grad_norm": 2.134883165359497,
+      "learning_rate": 1.1535269709543569e-05,
+      "loss": 0.3678,
+      "step": 2144
+    },
+    {
+      "epoch": 1.271111111111111,
+      "grad_norm": 1.9501959085464478,
+      "learning_rate": 1.1531317921359415e-05,
+      "loss": 0.3162,
+      "step": 2145
+    },
+    {
+      "epoch": 1.2717037037037038,
+      "grad_norm": 1.8297475576400757,
+      "learning_rate": 1.1527366133175261e-05,
+      "loss": 0.306,
+      "step": 2146
+    },
+    {
+      "epoch": 1.2722962962962963,
+      "grad_norm": 1.5692954063415527,
+      "learning_rate": 1.1523414344991109e-05,
+      "loss": 0.2332,
+      "step": 2147
+    },
+    {
+      "epoch": 1.272888888888889,
+      "grad_norm": 1.4784479141235352,
+      "learning_rate": 1.1519462556806955e-05,
+      "loss": 0.243,
+      "step": 2148
+    },
+    {
+      "epoch": 1.2734814814814814,
+      "grad_norm": 1.6035743951797485,
+      "learning_rate": 1.1515510768622802e-05,
+      "loss": 0.2612,
+      "step": 2149
+    },
+    {
+      "epoch": 1.2740740740740741,
+      "grad_norm": 2.4171814918518066,
+      "learning_rate": 1.1511558980438648e-05,
+      "loss": 0.4352,
+      "step": 2150
+    },
+    {
+      "epoch": 1.2746666666666666,
+      "grad_norm": 1.713697075843811,
+      "learning_rate": 1.1507607192254496e-05,
+      "loss": 0.3157,
+      "step": 2151
+    },
+    {
+      "epoch": 1.2752592592592593,
+      "grad_norm": 1.9956634044647217,
+      "learning_rate": 1.1503655404070342e-05,
+      "loss": 0.3989,
+      "step": 2152
+    },
+    {
+      "epoch": 1.2758518518518518,
+      "grad_norm": 1.9372994899749756,
+      "learning_rate": 1.1499703615886188e-05,
+      "loss": 0.2833,
+      "step": 2153
+    },
+    {
+      "epoch": 1.2764444444444445,
+      "grad_norm": 2.0907061100006104,
+      "learning_rate": 1.1495751827702034e-05,
+      "loss": 0.3236,
+      "step": 2154
+    },
+    {
+      "epoch": 1.277037037037037,
+      "grad_norm": 2.3225607872009277,
+      "learning_rate": 1.1491800039517882e-05,
+      "loss": 0.2735,
+      "step": 2155
+    },
+    {
+      "epoch": 1.2776296296296297,
+      "grad_norm": 2.6231842041015625,
+      "learning_rate": 1.1487848251333729e-05,
+      "loss": 0.3472,
+      "step": 2156
+    },
+    {
+      "epoch": 1.2782222222222221,
+      "grad_norm": 1.6670981645584106,
+      "learning_rate": 1.1483896463149575e-05,
+      "loss": 0.2764,
+      "step": 2157
+    },
+    {
+      "epoch": 1.2788148148148148,
+      "grad_norm": 1.8497653007507324,
+      "learning_rate": 1.1479944674965423e-05,
+      "loss": 0.2517,
+      "step": 2158
+    },
+    {
+      "epoch": 1.2794074074074073,
+      "grad_norm": 2.0361506938934326,
+      "learning_rate": 1.1475992886781269e-05,
+      "loss": 0.209,
+      "step": 2159
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7606617212295532,
+      "learning_rate": 1.1472041098597115e-05,
+      "loss": 0.2249,
+      "step": 2160
+    },
+    {
+      "epoch": 1.2805925925925927,
+      "grad_norm": 2.028604745864868,
+      "learning_rate": 1.1468089310412962e-05,
+      "loss": 0.3468,
+      "step": 2161
+    },
+    {
+      "epoch": 1.2811851851851852,
+      "grad_norm": 1.7160520553588867,
+      "learning_rate": 1.146413752222881e-05,
+      "loss": 0.2156,
+      "step": 2162
+    },
+    {
+      "epoch": 1.2817777777777777,
+      "grad_norm": 1.8842028379440308,
+      "learning_rate": 1.1460185734044656e-05,
+      "loss": 0.2375,
+      "step": 2163
+    },
+    {
+      "epoch": 1.2823703703703704,
+      "grad_norm": 2.4771029949188232,
+      "learning_rate": 1.1456233945860502e-05,
+      "loss": 0.3273,
+      "step": 2164
+    },
+    {
+      "epoch": 1.282962962962963,
+      "grad_norm": 1.6869632005691528,
+      "learning_rate": 1.1452282157676348e-05,
+      "loss": 0.2498,
+      "step": 2165
+    },
+    {
+      "epoch": 1.2835555555555556,
+      "grad_norm": 2.3799774646759033,
+      "learning_rate": 1.1448330369492196e-05,
+      "loss": 0.3785,
+      "step": 2166
+    },
+    {
+      "epoch": 1.284148148148148,
+      "grad_norm": 1.844546914100647,
+      "learning_rate": 1.1444378581308042e-05,
+      "loss": 0.2708,
+      "step": 2167
+    },
+    {
+      "epoch": 1.2847407407407407,
+      "grad_norm": 2.2456090450286865,
+      "learning_rate": 1.1440426793123889e-05,
+      "loss": 0.2777,
+      "step": 2168
+    },
+    {
+      "epoch": 1.2853333333333334,
+      "grad_norm": 2.3198747634887695,
+      "learning_rate": 1.1436475004939735e-05,
+      "loss": 0.3225,
+      "step": 2169
+    },
+    {
+      "epoch": 1.285925925925926,
+      "grad_norm": 2.1700236797332764,
+      "learning_rate": 1.1432523216755583e-05,
+      "loss": 0.3289,
+      "step": 2170
+    },
+    {
+      "epoch": 1.2865185185185184,
+      "grad_norm": 2.324221611022949,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.3136,
+      "step": 2171
+    },
+    {
+      "epoch": 1.287111111111111,
+      "grad_norm": 2.2977821826934814,
+      "learning_rate": 1.1424619640387275e-05,
+      "loss": 0.3487,
+      "step": 2172
+    },
+    {
+      "epoch": 1.2877037037037038,
+      "grad_norm": 1.9855068922042847,
+      "learning_rate": 1.1420667852203122e-05,
+      "loss": 0.3318,
+      "step": 2173
+    },
+    {
+      "epoch": 1.2882962962962963,
+      "grad_norm": 2.3068060874938965,
+      "learning_rate": 1.141671606401897e-05,
+      "loss": 0.2313,
+      "step": 2174
+    },
+    {
+      "epoch": 1.2888888888888888,
+      "grad_norm": 2.1035280227661133,
+      "learning_rate": 1.1412764275834816e-05,
+      "loss": 0.387,
+      "step": 2175
+    },
+    {
+      "epoch": 1.2894814814814815,
+      "grad_norm": 1.9942578077316284,
+      "learning_rate": 1.1408812487650662e-05,
+      "loss": 0.302,
+      "step": 2176
+    },
+    {
+      "epoch": 1.2900740740740742,
+      "grad_norm": 1.8097282648086548,
+      "learning_rate": 1.1404860699466508e-05,
+      "loss": 0.2772,
+      "step": 2177
+    },
+    {
+      "epoch": 1.2906666666666666,
+      "grad_norm": 1.4309965372085571,
+      "learning_rate": 1.1400908911282356e-05,
+      "loss": 0.2943,
+      "step": 2178
+    },
+    {
+      "epoch": 1.2912592592592593,
+      "grad_norm": 1.3964847326278687,
+      "learning_rate": 1.1396957123098202e-05,
+      "loss": 0.2561,
+      "step": 2179
+    },
+    {
+      "epoch": 1.2918518518518518,
+      "grad_norm": 2.2248010635375977,
+      "learning_rate": 1.1393005334914049e-05,
+      "loss": 0.301,
+      "step": 2180
+    },
+    {
+      "epoch": 1.2924444444444445,
+      "grad_norm": 2.7533466815948486,
+      "learning_rate": 1.1389053546729897e-05,
+      "loss": 0.3185,
+      "step": 2181
+    },
+    {
+      "epoch": 1.293037037037037,
+      "grad_norm": 1.9798941612243652,
+      "learning_rate": 1.1385101758545743e-05,
+      "loss": 0.2653,
+      "step": 2182
+    },
+    {
+      "epoch": 1.2936296296296297,
+      "grad_norm": 1.9669246673583984,
+      "learning_rate": 1.1381149970361589e-05,
+      "loss": 0.415,
+      "step": 2183
+    },
+    {
+      "epoch": 1.2942222222222222,
+      "grad_norm": 2.191131114959717,
+      "learning_rate": 1.1377198182177435e-05,
+      "loss": 0.284,
+      "step": 2184
+    },
+    {
+      "epoch": 1.2948148148148149,
+      "grad_norm": 1.893639326095581,
+      "learning_rate": 1.1373246393993283e-05,
+      "loss": 0.2574,
+      "step": 2185
+    },
+    {
+      "epoch": 1.2954074074074073,
+      "grad_norm": 2.849073886871338,
+      "learning_rate": 1.136929460580913e-05,
+      "loss": 0.3232,
+      "step": 2186
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 1.9835999011993408,
+      "learning_rate": 1.1365342817624976e-05,
+      "loss": 0.2756,
+      "step": 2187
+    },
+    {
+      "epoch": 1.2965925925925925,
+      "grad_norm": 2.779892921447754,
+      "learning_rate": 1.1361391029440822e-05,
+      "loss": 0.3448,
+      "step": 2188
+    },
+    {
+      "epoch": 1.2971851851851852,
+      "grad_norm": 2.1977016925811768,
+      "learning_rate": 1.135743924125667e-05,
+      "loss": 0.2757,
+      "step": 2189
+    },
+    {
+      "epoch": 1.2977777777777777,
+      "grad_norm": 2.21604061126709,
+      "learning_rate": 1.1353487453072516e-05,
+      "loss": 0.3075,
+      "step": 2190
+    },
+    {
+      "epoch": 1.2983703703703704,
+      "grad_norm": 2.1096129417419434,
+      "learning_rate": 1.1349535664888362e-05,
+      "loss": 0.3694,
+      "step": 2191
+    },
+    {
+      "epoch": 1.298962962962963,
+      "grad_norm": 4.045862197875977,
+      "learning_rate": 1.1345583876704209e-05,
+      "loss": 0.5158,
+      "step": 2192
+    },
+    {
+      "epoch": 1.2995555555555556,
+      "grad_norm": 2.126708507537842,
+      "learning_rate": 1.1341632088520057e-05,
+      "loss": 0.2368,
+      "step": 2193
+    },
+    {
+      "epoch": 1.300148148148148,
+      "grad_norm": 1.7964240312576294,
+      "learning_rate": 1.1337680300335903e-05,
+      "loss": 0.2351,
+      "step": 2194
+    },
+    {
+      "epoch": 1.3007407407407408,
+      "grad_norm": 2.9344048500061035,
+      "learning_rate": 1.1333728512151749e-05,
+      "loss": 0.5215,
+      "step": 2195
+    },
+    {
+      "epoch": 1.3013333333333335,
+      "grad_norm": 1.9139988422393799,
+      "learning_rate": 1.1329776723967595e-05,
+      "loss": 0.3661,
+      "step": 2196
+    },
+    {
+      "epoch": 1.301925925925926,
+      "grad_norm": 2.0840132236480713,
+      "learning_rate": 1.1325824935783443e-05,
+      "loss": 0.3664,
+      "step": 2197
+    },
+    {
+      "epoch": 1.3025185185185184,
+      "grad_norm": 2.048380136489868,
+      "learning_rate": 1.132187314759929e-05,
+      "loss": 0.2842,
+      "step": 2198
+    },
+    {
+      "epoch": 1.303111111111111,
+      "grad_norm": 1.7074178457260132,
+      "learning_rate": 1.1317921359415136e-05,
+      "loss": 0.2831,
+      "step": 2199
+    },
+    {
+      "epoch": 1.3037037037037038,
+      "grad_norm": 1.543402910232544,
+      "learning_rate": 1.1313969571230982e-05,
+      "loss": 0.3268,
+      "step": 2200
+    },
+    {
+      "epoch": 1.3042962962962963,
+      "grad_norm": 1.767120122909546,
+      "learning_rate": 1.131001778304683e-05,
+      "loss": 0.2834,
+      "step": 2201
+    },
+    {
+      "epoch": 1.3048888888888888,
+      "grad_norm": 1.7571955919265747,
+      "learning_rate": 1.1306065994862676e-05,
+      "loss": 0.2889,
+      "step": 2202
+    },
+    {
+      "epoch": 1.3054814814814815,
+      "grad_norm": 1.8186219930648804,
+      "learning_rate": 1.1302114206678522e-05,
+      "loss": 0.2409,
+      "step": 2203
+    },
+    {
+      "epoch": 1.3060740740740742,
+      "grad_norm": 4.659828186035156,
+      "learning_rate": 1.129816241849437e-05,
+      "loss": 0.323,
+      "step": 2204
+    },
+    {
+      "epoch": 1.3066666666666666,
+      "grad_norm": 2.0543107986450195,
+      "learning_rate": 1.1294210630310217e-05,
+      "loss": 0.3131,
+      "step": 2205
+    },
+    {
+      "epoch": 1.3072592592592591,
+      "grad_norm": 1.9805145263671875,
+      "learning_rate": 1.1290258842126063e-05,
+      "loss": 0.3538,
+      "step": 2206
+    },
+    {
+      "epoch": 1.3078518518518518,
+      "grad_norm": 1.6803957223892212,
+      "learning_rate": 1.1286307053941909e-05,
+      "loss": 0.3378,
+      "step": 2207
+    },
+    {
+      "epoch": 1.3084444444444445,
+      "grad_norm": 1.5613071918487549,
+      "learning_rate": 1.1282355265757757e-05,
+      "loss": 0.2535,
+      "step": 2208
+    },
+    {
+      "epoch": 1.309037037037037,
+      "grad_norm": 1.9051684141159058,
+      "learning_rate": 1.1278403477573603e-05,
+      "loss": 0.2645,
+      "step": 2209
+    },
+    {
+      "epoch": 1.3096296296296297,
+      "grad_norm": 1.5356380939483643,
+      "learning_rate": 1.127445168938945e-05,
+      "loss": 0.2947,
+      "step": 2210
+    },
+    {
+      "epoch": 1.3102222222222222,
+      "grad_norm": 1.5967742204666138,
+      "learning_rate": 1.1270499901205296e-05,
+      "loss": 0.2371,
+      "step": 2211
+    },
+    {
+      "epoch": 1.3108148148148149,
+      "grad_norm": 2.3283283710479736,
+      "learning_rate": 1.1266548113021144e-05,
+      "loss": 0.3552,
+      "step": 2212
+    },
+    {
+      "epoch": 1.3114074074074074,
+      "grad_norm": 1.7359204292297363,
+      "learning_rate": 1.126259632483699e-05,
+      "loss": 0.2702,
+      "step": 2213
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 1.9654388427734375,
+      "learning_rate": 1.1258644536652836e-05,
+      "loss": 0.2302,
+      "step": 2214
+    },
+    {
+      "epoch": 1.3125925925925925,
+      "grad_norm": 2.8144147396087646,
+      "learning_rate": 1.1254692748468682e-05,
+      "loss": 0.3317,
+      "step": 2215
+    },
+    {
+      "epoch": 1.3131851851851852,
+      "grad_norm": 5.712894439697266,
+      "learning_rate": 1.125074096028453e-05,
+      "loss": 0.5958,
+      "step": 2216
+    },
+    {
+      "epoch": 1.3137777777777777,
+      "grad_norm": 2.6768202781677246,
+      "learning_rate": 1.1246789172100377e-05,
+      "loss": 0.3452,
+      "step": 2217
+    },
+    {
+      "epoch": 1.3143703703703704,
+      "grad_norm": 2.615339756011963,
+      "learning_rate": 1.1242837383916223e-05,
+      "loss": 0.3614,
+      "step": 2218
+    },
+    {
+      "epoch": 1.314962962962963,
+      "grad_norm": 2.033229112625122,
+      "learning_rate": 1.1238885595732069e-05,
+      "loss": 0.3479,
+      "step": 2219
+    },
+    {
+      "epoch": 1.3155555555555556,
+      "grad_norm": 1.6108304262161255,
+      "learning_rate": 1.1234933807547917e-05,
+      "loss": 0.2119,
+      "step": 2220
+    },
+    {
+      "epoch": 1.316148148148148,
+      "grad_norm": 2.8189330101013184,
+      "learning_rate": 1.1230982019363763e-05,
+      "loss": 0.5639,
+      "step": 2221
+    },
+    {
+      "epoch": 1.3167407407407408,
+      "grad_norm": 1.8310495615005493,
+      "learning_rate": 1.122703023117961e-05,
+      "loss": 0.297,
+      "step": 2222
+    },
+    {
+      "epoch": 1.3173333333333335,
+      "grad_norm": 2.0458638668060303,
+      "learning_rate": 1.1223078442995456e-05,
+      "loss": 0.3041,
+      "step": 2223
+    },
+    {
+      "epoch": 1.317925925925926,
+      "grad_norm": 2.6484556198120117,
+      "learning_rate": 1.1219126654811304e-05,
+      "loss": 0.218,
+      "step": 2224
+    },
+    {
+      "epoch": 1.3185185185185184,
+      "grad_norm": 1.8325554132461548,
+      "learning_rate": 1.121517486662715e-05,
+      "loss": 0.3101,
+      "step": 2225
+    },
+    {
+      "epoch": 1.3191111111111111,
+      "grad_norm": 2.025631904602051,
+      "learning_rate": 1.1211223078442996e-05,
+      "loss": 0.2931,
+      "step": 2226
+    },
+    {
+      "epoch": 1.3197037037037038,
+      "grad_norm": 1.8689476251602173,
+      "learning_rate": 1.1207271290258844e-05,
+      "loss": 0.2519,
+      "step": 2227
+    },
+    {
+      "epoch": 1.3202962962962963,
+      "grad_norm": 1.788837194442749,
+      "learning_rate": 1.120331950207469e-05,
+      "loss": 0.4143,
+      "step": 2228
+    },
+    {
+      "epoch": 1.3208888888888888,
+      "grad_norm": 2.3122987747192383,
+      "learning_rate": 1.1199367713890537e-05,
+      "loss": 0.3929,
+      "step": 2229
+    },
+    {
+      "epoch": 1.3214814814814815,
+      "grad_norm": 2.09542179107666,
+      "learning_rate": 1.1195415925706383e-05,
+      "loss": 0.2486,
+      "step": 2230
+    },
+    {
+      "epoch": 1.3220740740740742,
+      "grad_norm": 1.783646583557129,
+      "learning_rate": 1.119146413752223e-05,
+      "loss": 0.284,
+      "step": 2231
+    },
+    {
+      "epoch": 1.3226666666666667,
+      "grad_norm": 1.9498192071914673,
+      "learning_rate": 1.1187512349338077e-05,
+      "loss": 0.2696,
+      "step": 2232
+    },
+    {
+      "epoch": 1.3232592592592591,
+      "grad_norm": 2.7122786045074463,
+      "learning_rate": 1.1183560561153923e-05,
+      "loss": 0.4472,
+      "step": 2233
+    },
+    {
+      "epoch": 1.3238518518518518,
+      "grad_norm": 3.30476450920105,
+      "learning_rate": 1.117960877296977e-05,
+      "loss": 0.4223,
+      "step": 2234
+    },
+    {
+      "epoch": 1.3244444444444445,
+      "grad_norm": 1.696821689605713,
+      "learning_rate": 1.1175656984785617e-05,
+      "loss": 0.2564,
+      "step": 2235
+    },
+    {
+      "epoch": 1.325037037037037,
+      "grad_norm": 2.3360939025878906,
+      "learning_rate": 1.1171705196601464e-05,
+      "loss": 0.351,
+      "step": 2236
+    },
+    {
+      "epoch": 1.3256296296296295,
+      "grad_norm": 1.6494330167770386,
+      "learning_rate": 1.116775340841731e-05,
+      "loss": 0.2522,
+      "step": 2237
+    },
+    {
+      "epoch": 1.3262222222222222,
+      "grad_norm": 2.3346569538116455,
+      "learning_rate": 1.1163801620233156e-05,
+      "loss": 0.3474,
+      "step": 2238
+    },
+    {
+      "epoch": 1.326814814814815,
+      "grad_norm": 1.5435643196105957,
+      "learning_rate": 1.1159849832049004e-05,
+      "loss": 0.2457,
+      "step": 2239
+    },
+    {
+      "epoch": 1.3274074074074074,
+      "grad_norm": 2.137354612350464,
+      "learning_rate": 1.115589804386485e-05,
+      "loss": 0.3124,
+      "step": 2240
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 1.9188755750656128,
+      "learning_rate": 1.1151946255680697e-05,
+      "loss": 0.3357,
+      "step": 2241
+    },
+    {
+      "epoch": 1.3285925925925925,
+      "grad_norm": 2.3746941089630127,
+      "learning_rate": 1.1147994467496543e-05,
+      "loss": 0.2279,
+      "step": 2242
+    },
+    {
+      "epoch": 1.3291851851851852,
+      "grad_norm": 1.9012064933776855,
+      "learning_rate": 1.114404267931239e-05,
+      "loss": 0.3201,
+      "step": 2243
+    },
+    {
+      "epoch": 1.3297777777777777,
+      "grad_norm": 1.6818475723266602,
+      "learning_rate": 1.1140090891128237e-05,
+      "loss": 0.2392,
+      "step": 2244
+    },
+    {
+      "epoch": 1.3303703703703704,
+      "grad_norm": 2.379401683807373,
+      "learning_rate": 1.1136139102944083e-05,
+      "loss": 0.4189,
+      "step": 2245
+    },
+    {
+      "epoch": 1.330962962962963,
+      "grad_norm": 3.9291107654571533,
+      "learning_rate": 1.113218731475993e-05,
+      "loss": 0.4473,
+      "step": 2246
+    },
+    {
+      "epoch": 1.3315555555555556,
+      "grad_norm": 2.2827446460723877,
+      "learning_rate": 1.1128235526575777e-05,
+      "loss": 0.3395,
+      "step": 2247
+    },
+    {
+      "epoch": 1.332148148148148,
+      "grad_norm": 2.006833553314209,
+      "learning_rate": 1.1124283738391624e-05,
+      "loss": 0.347,
+      "step": 2248
+    },
+    {
+      "epoch": 1.3327407407407408,
+      "grad_norm": 2.160987615585327,
+      "learning_rate": 1.112033195020747e-05,
+      "loss": 0.2902,
+      "step": 2249
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 2.6392831802368164,
+      "learning_rate": 1.1116380162023316e-05,
+      "loss": 0.2858,
+      "step": 2250
+    },
+    {
+      "epoch": 1.333925925925926,
+      "grad_norm": 2.5348901748657227,
+      "learning_rate": 1.1112428373839164e-05,
+      "loss": 0.2826,
+      "step": 2251
+    },
+    {
+      "epoch": 1.3345185185185184,
+      "grad_norm": 1.9208335876464844,
+      "learning_rate": 1.110847658565501e-05,
+      "loss": 0.2824,
+      "step": 2252
+    },
+    {
+      "epoch": 1.3351111111111111,
+      "grad_norm": 2.5806357860565186,
+      "learning_rate": 1.1104524797470857e-05,
+      "loss": 0.2649,
+      "step": 2253
+    },
+    {
+      "epoch": 1.3357037037037038,
+      "grad_norm": 1.8110727071762085,
+      "learning_rate": 1.1100573009286704e-05,
+      "loss": 0.3413,
+      "step": 2254
+    },
+    {
+      "epoch": 1.3362962962962963,
+      "grad_norm": 3.995388984680176,
+      "learning_rate": 1.109662122110255e-05,
+      "loss": 0.429,
+      "step": 2255
+    },
+    {
+      "epoch": 1.3368888888888888,
+      "grad_norm": 2.0386250019073486,
+      "learning_rate": 1.1092669432918397e-05,
+      "loss": 0.2981,
+      "step": 2256
+    },
+    {
+      "epoch": 1.3374814814814815,
+      "grad_norm": 1.9328155517578125,
+      "learning_rate": 1.1088717644734243e-05,
+      "loss": 0.2467,
+      "step": 2257
+    },
+    {
+      "epoch": 1.3380740740740742,
+      "grad_norm": 2.0091657638549805,
+      "learning_rate": 1.1084765856550091e-05,
+      "loss": 0.3141,
+      "step": 2258
+    },
+    {
+      "epoch": 1.3386666666666667,
+      "grad_norm": 1.53679358959198,
+      "learning_rate": 1.1080814068365937e-05,
+      "loss": 0.2853,
+      "step": 2259
+    },
+    {
+      "epoch": 1.3392592592592591,
+      "grad_norm": 2.5810415744781494,
+      "learning_rate": 1.1076862280181784e-05,
+      "loss": 0.41,
+      "step": 2260
+    },
+    {
+      "epoch": 1.3398518518518518,
+      "grad_norm": 1.7566720247268677,
+      "learning_rate": 1.107291049199763e-05,
+      "loss": 0.2389,
+      "step": 2261
+    },
+    {
+      "epoch": 1.3404444444444445,
+      "grad_norm": 2.3213961124420166,
+      "learning_rate": 1.1068958703813478e-05,
+      "loss": 0.4187,
+      "step": 2262
+    },
+    {
+      "epoch": 1.341037037037037,
+      "grad_norm": 2.5966761112213135,
+      "learning_rate": 1.1065006915629324e-05,
+      "loss": 0.2878,
+      "step": 2263
+    },
+    {
+      "epoch": 1.3416296296296295,
+      "grad_norm": 1.559181809425354,
+      "learning_rate": 1.106105512744517e-05,
+      "loss": 0.278,
+      "step": 2264
+    },
+    {
+      "epoch": 1.3422222222222222,
+      "grad_norm": 1.7931731939315796,
+      "learning_rate": 1.1057103339261017e-05,
+      "loss": 0.3606,
+      "step": 2265
+    },
+    {
+      "epoch": 1.342814814814815,
+      "grad_norm": 1.8341883420944214,
+      "learning_rate": 1.1053151551076864e-05,
+      "loss": 0.255,
+      "step": 2266
+    },
+    {
+      "epoch": 1.3434074074074074,
+      "grad_norm": 1.744694471359253,
+      "learning_rate": 1.104919976289271e-05,
+      "loss": 0.2784,
+      "step": 2267
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 1.8282978534698486,
+      "learning_rate": 1.1045247974708557e-05,
+      "loss": 0.4252,
+      "step": 2268
+    },
+    {
+      "epoch": 1.3445925925925926,
+      "grad_norm": 1.6006852388381958,
+      "learning_rate": 1.1041296186524403e-05,
+      "loss": 0.3585,
+      "step": 2269
+    },
+    {
+      "epoch": 1.3451851851851853,
+      "grad_norm": 1.525848627090454,
+      "learning_rate": 1.1037344398340251e-05,
+      "loss": 0.2529,
+      "step": 2270
+    },
+    {
+      "epoch": 1.3457777777777777,
+      "grad_norm": 1.6753060817718506,
+      "learning_rate": 1.1033392610156097e-05,
+      "loss": 0.2633,
+      "step": 2271
+    },
+    {
+      "epoch": 1.3463703703703704,
+      "grad_norm": 1.459890365600586,
+      "learning_rate": 1.1029440821971944e-05,
+      "loss": 0.2358,
+      "step": 2272
+    },
+    {
+      "epoch": 1.346962962962963,
+      "grad_norm": 1.3835829496383667,
+      "learning_rate": 1.102548903378779e-05,
+      "loss": 0.1888,
+      "step": 2273
+    },
+    {
+      "epoch": 1.3475555555555556,
+      "grad_norm": 1.748975157737732,
+      "learning_rate": 1.1021537245603638e-05,
+      "loss": 0.2856,
+      "step": 2274
+    },
+    {
+      "epoch": 1.348148148148148,
+      "grad_norm": 2.536123514175415,
+      "learning_rate": 1.1017585457419484e-05,
+      "loss": 0.3835,
+      "step": 2275
+    },
+    {
+      "epoch": 1.3487407407407408,
+      "grad_norm": 1.7122952938079834,
+      "learning_rate": 1.101363366923533e-05,
+      "loss": 0.2383,
+      "step": 2276
+    },
+    {
+      "epoch": 1.3493333333333333,
+      "grad_norm": 3.933701276779175,
+      "learning_rate": 1.1009681881051178e-05,
+      "loss": 0.4257,
+      "step": 2277
+    },
+    {
+      "epoch": 1.349925925925926,
+      "grad_norm": 1.9841924905776978,
+      "learning_rate": 1.1005730092867024e-05,
+      "loss": 0.2676,
+      "step": 2278
+    },
+    {
+      "epoch": 1.3505185185185185,
+      "grad_norm": 2.531139373779297,
+      "learning_rate": 1.100177830468287e-05,
+      "loss": 0.2677,
+      "step": 2279
+    },
+    {
+      "epoch": 1.3511111111111112,
+      "grad_norm": 2.8113410472869873,
+      "learning_rate": 1.0997826516498717e-05,
+      "loss": 0.2875,
+      "step": 2280
+    },
+    {
+      "epoch": 1.3517037037037036,
+      "grad_norm": 2.0232491493225098,
+      "learning_rate": 1.0993874728314565e-05,
+      "loss": 0.2725,
+      "step": 2281
+    },
+    {
+      "epoch": 1.3522962962962963,
+      "grad_norm": 2.258185625076294,
+      "learning_rate": 1.0989922940130411e-05,
+      "loss": 0.3291,
+      "step": 2282
+    },
+    {
+      "epoch": 1.3528888888888888,
+      "grad_norm": 1.773431420326233,
+      "learning_rate": 1.0985971151946257e-05,
+      "loss": 0.3357,
+      "step": 2283
+    },
+    {
+      "epoch": 1.3534814814814815,
+      "grad_norm": 3.830976963043213,
+      "learning_rate": 1.0982019363762104e-05,
+      "loss": 0.5692,
+      "step": 2284
+    },
+    {
+      "epoch": 1.3540740740740742,
+      "grad_norm": 1.67246413230896,
+      "learning_rate": 1.0978067575577952e-05,
+      "loss": 0.3304,
+      "step": 2285
+    },
+    {
+      "epoch": 1.3546666666666667,
+      "grad_norm": 1.578022837638855,
+      "learning_rate": 1.0974115787393798e-05,
+      "loss": 0.2921,
+      "step": 2286
+    },
+    {
+      "epoch": 1.3552592592592592,
+      "grad_norm": 1.8104262351989746,
+      "learning_rate": 1.0970163999209644e-05,
+      "loss": 0.2106,
+      "step": 2287
+    },
+    {
+      "epoch": 1.3558518518518519,
+      "grad_norm": 1.8533192873001099,
+      "learning_rate": 1.096621221102549e-05,
+      "loss": 0.1888,
+      "step": 2288
+    },
+    {
+      "epoch": 1.3564444444444446,
+      "grad_norm": 1.6631518602371216,
+      "learning_rate": 1.0962260422841338e-05,
+      "loss": 0.2996,
+      "step": 2289
+    },
+    {
+      "epoch": 1.357037037037037,
+      "grad_norm": 1.712837815284729,
+      "learning_rate": 1.0958308634657184e-05,
+      "loss": 0.2148,
+      "step": 2290
+    },
+    {
+      "epoch": 1.3576296296296295,
+      "grad_norm": 1.720829963684082,
+      "learning_rate": 1.095435684647303e-05,
+      "loss": 0.2704,
+      "step": 2291
+    },
+    {
+      "epoch": 1.3582222222222222,
+      "grad_norm": 2.3496875762939453,
+      "learning_rate": 1.0950405058288877e-05,
+      "loss": 0.2388,
+      "step": 2292
+    },
+    {
+      "epoch": 1.358814814814815,
+      "grad_norm": 1.705782413482666,
+      "learning_rate": 1.0946453270104725e-05,
+      "loss": 0.3124,
+      "step": 2293
+    },
+    {
+      "epoch": 1.3594074074074074,
+      "grad_norm": 1.654091238975525,
+      "learning_rate": 1.0942501481920571e-05,
+      "loss": 0.1996,
+      "step": 2294
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 2.240551233291626,
+      "learning_rate": 1.0938549693736417e-05,
+      "loss": 0.2329,
+      "step": 2295
+    },
+    {
+      "epoch": 1.3605925925925926,
+      "grad_norm": 4.9007792472839355,
+      "learning_rate": 1.0934597905552262e-05,
+      "loss": 0.4815,
+      "step": 2296
+    },
+    {
+      "epoch": 1.3611851851851853,
+      "grad_norm": 2.36930513381958,
+      "learning_rate": 1.0930646117368108e-05,
+      "loss": 0.2942,
+      "step": 2297
+    },
+    {
+      "epoch": 1.3617777777777778,
+      "grad_norm": 2.122377395629883,
+      "learning_rate": 1.0926694329183956e-05,
+      "loss": 0.3185,
+      "step": 2298
+    },
+    {
+      "epoch": 1.3623703703703702,
+      "grad_norm": 2.2160205841064453,
+      "learning_rate": 1.0922742540999802e-05,
+      "loss": 0.3511,
+      "step": 2299
+    },
+    {
+      "epoch": 1.362962962962963,
+      "grad_norm": 1.7915318012237549,
+      "learning_rate": 1.0918790752815649e-05,
+      "loss": 0.2317,
+      "step": 2300
+    },
+    {
+      "epoch": 1.3635555555555556,
+      "grad_norm": 2.3705191612243652,
+      "learning_rate": 1.0914838964631495e-05,
+      "loss": 0.2874,
+      "step": 2301
+    },
+    {
+      "epoch": 1.364148148148148,
+      "grad_norm": 1.9432179927825928,
+      "learning_rate": 1.0910887176447343e-05,
+      "loss": 0.2573,
+      "step": 2302
+    },
+    {
+      "epoch": 1.3647407407407408,
+      "grad_norm": 1.8327114582061768,
+      "learning_rate": 1.0906935388263189e-05,
+      "loss": 0.2367,
+      "step": 2303
+    },
+    {
+      "epoch": 1.3653333333333333,
+      "grad_norm": 1.8563377857208252,
+      "learning_rate": 1.0902983600079035e-05,
+      "loss": 0.2389,
+      "step": 2304
+    },
+    {
+      "epoch": 1.365925925925926,
+      "grad_norm": 2.318613052368164,
+      "learning_rate": 1.0899031811894883e-05,
+      "loss": 0.2984,
+      "step": 2305
+    },
+    {
+      "epoch": 1.3665185185185185,
+      "grad_norm": 1.999005675315857,
+      "learning_rate": 1.089508002371073e-05,
+      "loss": 0.3322,
+      "step": 2306
+    },
+    {
+      "epoch": 1.3671111111111112,
+      "grad_norm": 2.2920522689819336,
+      "learning_rate": 1.0891128235526576e-05,
+      "loss": 0.3012,
+      "step": 2307
+    },
+    {
+      "epoch": 1.3677037037037036,
+      "grad_norm": 1.9379653930664062,
+      "learning_rate": 1.0887176447342422e-05,
+      "loss": 0.3369,
+      "step": 2308
+    },
+    {
+      "epoch": 1.3682962962962963,
+      "grad_norm": 2.1408984661102295,
+      "learning_rate": 1.088322465915827e-05,
+      "loss": 0.2736,
+      "step": 2309
+    },
+    {
+      "epoch": 1.3688888888888888,
+      "grad_norm": 2.1763761043548584,
+      "learning_rate": 1.0879272870974116e-05,
+      "loss": 0.2717,
+      "step": 2310
+    },
+    {
+      "epoch": 1.3694814814814815,
+      "grad_norm": 3.595248222351074,
+      "learning_rate": 1.0875321082789962e-05,
+      "loss": 0.3555,
+      "step": 2311
+    },
+    {
+      "epoch": 1.370074074074074,
+      "grad_norm": 2.370877742767334,
+      "learning_rate": 1.0871369294605809e-05,
+      "loss": 0.4103,
+      "step": 2312
+    },
+    {
+      "epoch": 1.3706666666666667,
+      "grad_norm": 2.3401684761047363,
+      "learning_rate": 1.0867417506421656e-05,
+      "loss": 0.3513,
+      "step": 2313
+    },
+    {
+      "epoch": 1.3712592592592592,
+      "grad_norm": 2.8367679119110107,
+      "learning_rate": 1.0863465718237503e-05,
+      "loss": 0.3021,
+      "step": 2314
+    },
+    {
+      "epoch": 1.3718518518518519,
+      "grad_norm": 2.9580070972442627,
+      "learning_rate": 1.0859513930053349e-05,
+      "loss": 0.3352,
+      "step": 2315
+    },
+    {
+      "epoch": 1.3724444444444446,
+      "grad_norm": 2.739544153213501,
+      "learning_rate": 1.0855562141869195e-05,
+      "loss": 0.2698,
+      "step": 2316
+    },
+    {
+      "epoch": 1.373037037037037,
+      "grad_norm": 2.250401020050049,
+      "learning_rate": 1.0851610353685043e-05,
+      "loss": 0.2434,
+      "step": 2317
+    },
+    {
+      "epoch": 1.3736296296296295,
+      "grad_norm": 2.1196722984313965,
+      "learning_rate": 1.084765856550089e-05,
+      "loss": 0.3463,
+      "step": 2318
+    },
+    {
+      "epoch": 1.3742222222222222,
+      "grad_norm": 1.9983973503112793,
+      "learning_rate": 1.0843706777316736e-05,
+      "loss": 0.3361,
+      "step": 2319
+    },
+    {
+      "epoch": 1.374814814814815,
+      "grad_norm": 3.2092573642730713,
+      "learning_rate": 1.0839754989132582e-05,
+      "loss": 0.3741,
+      "step": 2320
+    },
+    {
+      "epoch": 1.3754074074074074,
+      "grad_norm": 2.7990615367889404,
+      "learning_rate": 1.083580320094843e-05,
+      "loss": 0.2198,
+      "step": 2321
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 2.2584826946258545,
+      "learning_rate": 1.0831851412764276e-05,
+      "loss": 0.2644,
+      "step": 2322
+    },
+    {
+      "epoch": 1.3765925925925926,
+      "grad_norm": 2.3690197467803955,
+      "learning_rate": 1.0827899624580122e-05,
+      "loss": 0.3434,
+      "step": 2323
+    },
+    {
+      "epoch": 1.3771851851851853,
+      "grad_norm": 2.221160411834717,
+      "learning_rate": 1.0823947836395969e-05,
+      "loss": 0.3124,
+      "step": 2324
+    },
+    {
+      "epoch": 1.3777777777777778,
+      "grad_norm": 2.1692607402801514,
+      "learning_rate": 1.0819996048211816e-05,
+      "loss": 0.3098,
+      "step": 2325
+    },
+    {
+      "epoch": 1.3783703703703702,
+      "grad_norm": 2.1512258052825928,
+      "learning_rate": 1.0816044260027663e-05,
+      "loss": 0.2891,
+      "step": 2326
+    },
+    {
+      "epoch": 1.378962962962963,
+      "grad_norm": 2.8442842960357666,
+      "learning_rate": 1.0812092471843509e-05,
+      "loss": 0.3569,
+      "step": 2327
+    },
+    {
+      "epoch": 1.3795555555555556,
+      "grad_norm": 1.954676628112793,
+      "learning_rate": 1.0808140683659357e-05,
+      "loss": 0.2719,
+      "step": 2328
+    },
+    {
+      "epoch": 1.3801481481481481,
+      "grad_norm": 1.93262779712677,
+      "learning_rate": 1.0804188895475203e-05,
+      "loss": 0.2757,
+      "step": 2329
+    },
+    {
+      "epoch": 1.3807407407407408,
+      "grad_norm": 1.9971131086349487,
+      "learning_rate": 1.080023710729105e-05,
+      "loss": 0.3854,
+      "step": 2330
+    },
+    {
+      "epoch": 1.3813333333333333,
+      "grad_norm": 1.5142337083816528,
+      "learning_rate": 1.0796285319106896e-05,
+      "loss": 0.2111,
+      "step": 2331
+    },
+    {
+      "epoch": 1.381925925925926,
+      "grad_norm": 1.7095754146575928,
+      "learning_rate": 1.0792333530922744e-05,
+      "loss": 0.283,
+      "step": 2332
+    },
+    {
+      "epoch": 1.3825185185185185,
+      "grad_norm": 2.027979850769043,
+      "learning_rate": 1.078838174273859e-05,
+      "loss": 0.276,
+      "step": 2333
+    },
+    {
+      "epoch": 1.3831111111111112,
+      "grad_norm": 4.3721795082092285,
+      "learning_rate": 1.0784429954554436e-05,
+      "loss": 0.3348,
+      "step": 2334
+    },
+    {
+      "epoch": 1.3837037037037037,
+      "grad_norm": 2.9374921321868896,
+      "learning_rate": 1.0780478166370282e-05,
+      "loss": 0.3633,
+      "step": 2335
+    },
+    {
+      "epoch": 1.3842962962962964,
+      "grad_norm": 3.966456413269043,
+      "learning_rate": 1.077652637818613e-05,
+      "loss": 0.4579,
+      "step": 2336
+    },
+    {
+      "epoch": 1.3848888888888888,
+      "grad_norm": 2.2341740131378174,
+      "learning_rate": 1.0772574590001976e-05,
+      "loss": 0.3429,
+      "step": 2337
+    },
+    {
+      "epoch": 1.3854814814814815,
+      "grad_norm": 5.239346981048584,
+      "learning_rate": 1.0768622801817823e-05,
+      "loss": 0.3986,
+      "step": 2338
+    },
+    {
+      "epoch": 1.386074074074074,
+      "grad_norm": 2.83630108833313,
+      "learning_rate": 1.0764671013633669e-05,
+      "loss": 0.2748,
+      "step": 2339
+    },
+    {
+      "epoch": 1.3866666666666667,
+      "grad_norm": 2.6694910526275635,
+      "learning_rate": 1.0760719225449517e-05,
+      "loss": 0.3045,
+      "step": 2340
+    },
+    {
+      "epoch": 1.3872592592592592,
+      "grad_norm": 3.96476149559021,
+      "learning_rate": 1.0756767437265363e-05,
+      "loss": 0.3607,
+      "step": 2341
+    },
+    {
+      "epoch": 1.387851851851852,
+      "grad_norm": 2.438718557357788,
+      "learning_rate": 1.075281564908121e-05,
+      "loss": 0.3401,
+      "step": 2342
+    },
+    {
+      "epoch": 1.3884444444444444,
+      "grad_norm": 2.144435167312622,
+      "learning_rate": 1.0748863860897056e-05,
+      "loss": 0.2716,
+      "step": 2343
+    },
+    {
+      "epoch": 1.389037037037037,
+      "grad_norm": 1.7316428422927856,
+      "learning_rate": 1.0744912072712904e-05,
+      "loss": 0.2866,
+      "step": 2344
+    },
+    {
+      "epoch": 1.3896296296296295,
+      "grad_norm": 1.4653562307357788,
+      "learning_rate": 1.074096028452875e-05,
+      "loss": 0.1976,
+      "step": 2345
+    },
+    {
+      "epoch": 1.3902222222222222,
+      "grad_norm": 1.9142284393310547,
+      "learning_rate": 1.0737008496344596e-05,
+      "loss": 0.373,
+      "step": 2346
+    },
+    {
+      "epoch": 1.390814814814815,
+      "grad_norm": 3.2299375534057617,
+      "learning_rate": 1.0733056708160442e-05,
+      "loss": 0.378,
+      "step": 2347
+    },
+    {
+      "epoch": 1.3914074074074074,
+      "grad_norm": 2.373948097229004,
+      "learning_rate": 1.072910491997629e-05,
+      "loss": 0.2604,
+      "step": 2348
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 2.6192829608917236,
+      "learning_rate": 1.0725153131792136e-05,
+      "loss": 0.3401,
+      "step": 2349
+    },
+    {
+      "epoch": 1.3925925925925926,
+      "grad_norm": 1.7720896005630493,
+      "learning_rate": 1.0721201343607983e-05,
+      "loss": 0.3161,
+      "step": 2350
+    },
+    {
+      "epoch": 1.3931851851851853,
+      "grad_norm": 1.8529200553894043,
+      "learning_rate": 1.071724955542383e-05,
+      "loss": 0.3083,
+      "step": 2351
+    },
+    {
+      "epoch": 1.3937777777777778,
+      "grad_norm": 1.7172199487686157,
+      "learning_rate": 1.0713297767239677e-05,
+      "loss": 0.24,
+      "step": 2352
+    },
+    {
+      "epoch": 1.3943703703703703,
+      "grad_norm": 2.3952276706695557,
+      "learning_rate": 1.0709345979055523e-05,
+      "loss": 0.3588,
+      "step": 2353
+    },
+    {
+      "epoch": 1.394962962962963,
+      "grad_norm": 2.1374807357788086,
+      "learning_rate": 1.070539419087137e-05,
+      "loss": 0.3649,
+      "step": 2354
+    },
+    {
+      "epoch": 1.3955555555555557,
+      "grad_norm": 1.9750853776931763,
+      "learning_rate": 1.0701442402687217e-05,
+      "loss": 0.3028,
+      "step": 2355
+    },
+    {
+      "epoch": 1.3961481481481481,
+      "grad_norm": 1.644826889038086,
+      "learning_rate": 1.0697490614503064e-05,
+      "loss": 0.2905,
+      "step": 2356
+    },
+    {
+      "epoch": 1.3967407407407406,
+      "grad_norm": 1.5585070848464966,
+      "learning_rate": 1.069353882631891e-05,
+      "loss": 0.2721,
+      "step": 2357
+    },
+    {
+      "epoch": 1.3973333333333333,
+      "grad_norm": 1.5995774269104004,
+      "learning_rate": 1.0689587038134756e-05,
+      "loss": 0.2225,
+      "step": 2358
+    },
+    {
+      "epoch": 1.397925925925926,
+      "grad_norm": 1.7873591184616089,
+      "learning_rate": 1.0685635249950604e-05,
+      "loss": 0.2868,
+      "step": 2359
+    },
+    {
+      "epoch": 1.3985185185185185,
+      "grad_norm": 2.450624704360962,
+      "learning_rate": 1.068168346176645e-05,
+      "loss": 0.3713,
+      "step": 2360
+    },
+    {
+      "epoch": 1.3991111111111112,
+      "grad_norm": 2.143658399581909,
+      "learning_rate": 1.0677731673582296e-05,
+      "loss": 0.2241,
+      "step": 2361
+    },
+    {
+      "epoch": 1.3997037037037037,
+      "grad_norm": 2.4094083309173584,
+      "learning_rate": 1.0673779885398143e-05,
+      "loss": 0.3126,
+      "step": 2362
+    },
+    {
+      "epoch": 1.4002962962962964,
+      "grad_norm": 2.0555620193481445,
+      "learning_rate": 1.066982809721399e-05,
+      "loss": 0.2554,
+      "step": 2363
+    },
+    {
+      "epoch": 1.4008888888888889,
+      "grad_norm": 2.5633842945098877,
+      "learning_rate": 1.0665876309029837e-05,
+      "loss": 0.2501,
+      "step": 2364
+    },
+    {
+      "epoch": 1.4014814814814816,
+      "grad_norm": 1.7919220924377441,
+      "learning_rate": 1.0661924520845683e-05,
+      "loss": 0.2036,
+      "step": 2365
+    },
+    {
+      "epoch": 1.402074074074074,
+      "grad_norm": 3.1917266845703125,
+      "learning_rate": 1.065797273266153e-05,
+      "loss": 0.5381,
+      "step": 2366
+    },
+    {
+      "epoch": 1.4026666666666667,
+      "grad_norm": 2.0949790477752686,
+      "learning_rate": 1.0654020944477377e-05,
+      "loss": 0.3371,
+      "step": 2367
+    },
+    {
+      "epoch": 1.4032592592592592,
+      "grad_norm": 1.8188904523849487,
+      "learning_rate": 1.0650069156293224e-05,
+      "loss": 0.284,
+      "step": 2368
+    },
+    {
+      "epoch": 1.403851851851852,
+      "grad_norm": 1.907219648361206,
+      "learning_rate": 1.064611736810907e-05,
+      "loss": 0.2873,
+      "step": 2369
+    },
+    {
+      "epoch": 1.4044444444444444,
+      "grad_norm": 1.939517855644226,
+      "learning_rate": 1.0642165579924916e-05,
+      "loss": 0.2571,
+      "step": 2370
+    },
+    {
+      "epoch": 1.405037037037037,
+      "grad_norm": 1.8702802658081055,
+      "learning_rate": 1.0638213791740764e-05,
+      "loss": 0.2012,
+      "step": 2371
+    },
+    {
+      "epoch": 1.4056296296296296,
+      "grad_norm": 1.7469837665557861,
+      "learning_rate": 1.063426200355661e-05,
+      "loss": 0.2767,
+      "step": 2372
+    },
+    {
+      "epoch": 1.4062222222222223,
+      "grad_norm": 2.73512864112854,
+      "learning_rate": 1.0630310215372456e-05,
+      "loss": 0.3299,
+      "step": 2373
+    },
+    {
+      "epoch": 1.4068148148148147,
+      "grad_norm": 2.5234274864196777,
+      "learning_rate": 1.0626358427188304e-05,
+      "loss": 0.3513,
+      "step": 2374
+    },
+    {
+      "epoch": 1.4074074074074074,
+      "grad_norm": 2.859783411026001,
+      "learning_rate": 1.062240663900415e-05,
+      "loss": 0.242,
+      "step": 2375
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 1.7839930057525635,
+      "learning_rate": 1.0618454850819997e-05,
+      "loss": 0.2668,
+      "step": 2376
+    },
+    {
+      "epoch": 1.4085925925925926,
+      "grad_norm": 2.360783100128174,
+      "learning_rate": 1.0614503062635843e-05,
+      "loss": 0.4008,
+      "step": 2377
+    },
+    {
+      "epoch": 1.4091851851851853,
+      "grad_norm": 1.8487942218780518,
+      "learning_rate": 1.0610551274451691e-05,
+      "loss": 0.2816,
+      "step": 2378
+    },
+    {
+      "epoch": 1.4097777777777778,
+      "grad_norm": 1.6002659797668457,
+      "learning_rate": 1.0606599486267537e-05,
+      "loss": 0.2296,
+      "step": 2379
+    },
+    {
+      "epoch": 1.4103703703703703,
+      "grad_norm": 3.3438892364501953,
+      "learning_rate": 1.0602647698083384e-05,
+      "loss": 0.4543,
+      "step": 2380
+    },
+    {
+      "epoch": 1.410962962962963,
+      "grad_norm": 2.7290656566619873,
+      "learning_rate": 1.059869590989923e-05,
+      "loss": 0.2692,
+      "step": 2381
+    },
+    {
+      "epoch": 1.4115555555555557,
+      "grad_norm": 3.306722402572632,
+      "learning_rate": 1.0594744121715078e-05,
+      "loss": 0.3997,
+      "step": 2382
+    },
+    {
+      "epoch": 1.4121481481481482,
+      "grad_norm": 1.9305033683776855,
+      "learning_rate": 1.0590792333530924e-05,
+      "loss": 0.257,
+      "step": 2383
+    },
+    {
+      "epoch": 1.4127407407407406,
+      "grad_norm": 1.9086958169937134,
+      "learning_rate": 1.058684054534677e-05,
+      "loss": 0.2243,
+      "step": 2384
+    },
+    {
+      "epoch": 1.4133333333333333,
+      "grad_norm": 1.854989767074585,
+      "learning_rate": 1.0582888757162616e-05,
+      "loss": 0.3012,
+      "step": 2385
+    },
+    {
+      "epoch": 1.413925925925926,
+      "grad_norm": 2.5557639598846436,
+      "learning_rate": 1.0578936968978464e-05,
+      "loss": 0.2724,
+      "step": 2386
+    },
+    {
+      "epoch": 1.4145185185185185,
+      "grad_norm": 1.9215487241744995,
+      "learning_rate": 1.057498518079431e-05,
+      "loss": 0.2582,
+      "step": 2387
+    },
+    {
+      "epoch": 1.415111111111111,
+      "grad_norm": 1.4485005140304565,
+      "learning_rate": 1.0571033392610157e-05,
+      "loss": 0.21,
+      "step": 2388
+    },
+    {
+      "epoch": 1.4157037037037037,
+      "grad_norm": 2.172549247741699,
+      "learning_rate": 1.0567081604426003e-05,
+      "loss": 0.2586,
+      "step": 2389
+    },
+    {
+      "epoch": 1.4162962962962964,
+      "grad_norm": 2.546903371810913,
+      "learning_rate": 1.0563129816241851e-05,
+      "loss": 0.3912,
+      "step": 2390
+    },
+    {
+      "epoch": 1.4168888888888889,
+      "grad_norm": 1.9813295602798462,
+      "learning_rate": 1.0559178028057697e-05,
+      "loss": 0.3001,
+      "step": 2391
+    },
+    {
+      "epoch": 1.4174814814814816,
+      "grad_norm": 1.903689980506897,
+      "learning_rate": 1.0555226239873544e-05,
+      "loss": 0.2308,
+      "step": 2392
+    },
+    {
+      "epoch": 1.418074074074074,
+      "grad_norm": 2.190202474594116,
+      "learning_rate": 1.055127445168939e-05,
+      "loss": 0.2709,
+      "step": 2393
+    },
+    {
+      "epoch": 1.4186666666666667,
+      "grad_norm": 1.7884316444396973,
+      "learning_rate": 1.0547322663505238e-05,
+      "loss": 0.2544,
+      "step": 2394
+    },
+    {
+      "epoch": 1.4192592592592592,
+      "grad_norm": 1.8002761602401733,
+      "learning_rate": 1.0543370875321084e-05,
+      "loss": 0.2667,
+      "step": 2395
+    },
+    {
+      "epoch": 1.419851851851852,
+      "grad_norm": 1.4127389192581177,
+      "learning_rate": 1.053941908713693e-05,
+      "loss": 0.1921,
+      "step": 2396
+    },
+    {
+      "epoch": 1.4204444444444444,
+      "grad_norm": 2.6571552753448486,
+      "learning_rate": 1.0535467298952776e-05,
+      "loss": 0.2508,
+      "step": 2397
+    },
+    {
+      "epoch": 1.421037037037037,
+      "grad_norm": 1.9515937566757202,
+      "learning_rate": 1.0531515510768624e-05,
+      "loss": 0.2976,
+      "step": 2398
+    },
+    {
+      "epoch": 1.4216296296296296,
+      "grad_norm": 2.1843278408050537,
+      "learning_rate": 1.052756372258447e-05,
+      "loss": 0.3129,
+      "step": 2399
+    },
+    {
+      "epoch": 1.4222222222222223,
+      "grad_norm": 1.763847827911377,
+      "learning_rate": 1.0523611934400317e-05,
+      "loss": 0.2886,
+      "step": 2400
+    },
+    {
+      "epoch": 1.4228148148148148,
+      "grad_norm": 2.74981689453125,
+      "learning_rate": 1.0519660146216165e-05,
+      "loss": 0.2289,
+      "step": 2401
+    },
+    {
+      "epoch": 1.4234074074074075,
+      "grad_norm": 2.531895875930786,
+      "learning_rate": 1.0515708358032011e-05,
+      "loss": 0.3964,
+      "step": 2402
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 2.326375722885132,
+      "learning_rate": 1.0511756569847857e-05,
+      "loss": 0.3363,
+      "step": 2403
+    },
+    {
+      "epoch": 1.4245925925925926,
+      "grad_norm": 2.2877330780029297,
+      "learning_rate": 1.0507804781663704e-05,
+      "loss": 0.2989,
+      "step": 2404
+    },
+    {
+      "epoch": 1.425185185185185,
+      "grad_norm": 2.776824712753296,
+      "learning_rate": 1.0503852993479551e-05,
+      "loss": 0.3882,
+      "step": 2405
+    },
+    {
+      "epoch": 1.4257777777777778,
+      "grad_norm": 2.1952104568481445,
+      "learning_rate": 1.0499901205295398e-05,
+      "loss": 0.2594,
+      "step": 2406
+    },
+    {
+      "epoch": 1.4263703703703703,
+      "grad_norm": 1.8058711290359497,
+      "learning_rate": 1.0495949417111244e-05,
+      "loss": 0.2649,
+      "step": 2407
+    },
+    {
+      "epoch": 1.426962962962963,
+      "grad_norm": 1.8542733192443848,
+      "learning_rate": 1.049199762892709e-05,
+      "loss": 0.3064,
+      "step": 2408
+    },
+    {
+      "epoch": 1.4275555555555557,
+      "grad_norm": 1.8550609350204468,
+      "learning_rate": 1.0488045840742938e-05,
+      "loss": 0.328,
+      "step": 2409
+    },
+    {
+      "epoch": 1.4281481481481482,
+      "grad_norm": 1.9533592462539673,
+      "learning_rate": 1.0484094052558784e-05,
+      "loss": 0.3184,
+      "step": 2410
+    },
+    {
+      "epoch": 1.4287407407407406,
+      "grad_norm": 2.1126861572265625,
+      "learning_rate": 1.048014226437463e-05,
+      "loss": 0.2529,
+      "step": 2411
+    },
+    {
+      "epoch": 1.4293333333333333,
+      "grad_norm": 2.1361725330352783,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 0.2704,
+      "step": 2412
+    },
+    {
+      "epoch": 1.429925925925926,
+      "grad_norm": 1.3936606645584106,
+      "learning_rate": 1.0472238688006325e-05,
+      "loss": 0.2033,
+      "step": 2413
+    },
+    {
+      "epoch": 1.4305185185185185,
+      "grad_norm": 1.520875334739685,
+      "learning_rate": 1.0468286899822171e-05,
+      "loss": 0.279,
+      "step": 2414
+    },
+    {
+      "epoch": 1.431111111111111,
+      "grad_norm": 1.6939860582351685,
+      "learning_rate": 1.0464335111638017e-05,
+      "loss": 0.2095,
+      "step": 2415
+    },
+    {
+      "epoch": 1.4317037037037037,
+      "grad_norm": 1.875057339668274,
+      "learning_rate": 1.0460383323453864e-05,
+      "loss": 0.2903,
+      "step": 2416
+    },
+    {
+      "epoch": 1.4322962962962964,
+      "grad_norm": 2.4372687339782715,
+      "learning_rate": 1.0456431535269711e-05,
+      "loss": 0.3222,
+      "step": 2417
+    },
+    {
+      "epoch": 1.4328888888888889,
+      "grad_norm": 2.399411201477051,
+      "learning_rate": 1.0452479747085558e-05,
+      "loss": 0.3428,
+      "step": 2418
+    },
+    {
+      "epoch": 1.4334814814814814,
+      "grad_norm": 2.031967878341675,
+      "learning_rate": 1.0448527958901404e-05,
+      "loss": 0.2833,
+      "step": 2419
+    },
+    {
+      "epoch": 1.434074074074074,
+      "grad_norm": 3.5370264053344727,
+      "learning_rate": 1.044457617071725e-05,
+      "loss": 0.4451,
+      "step": 2420
+    },
+    {
+      "epoch": 1.4346666666666668,
+      "grad_norm": 3.2940025329589844,
+      "learning_rate": 1.0440624382533098e-05,
+      "loss": 0.2964,
+      "step": 2421
+    },
+    {
+      "epoch": 1.4352592592592592,
+      "grad_norm": 1.8419616222381592,
+      "learning_rate": 1.0436672594348944e-05,
+      "loss": 0.2868,
+      "step": 2422
+    },
+    {
+      "epoch": 1.435851851851852,
+      "grad_norm": 1.6375725269317627,
+      "learning_rate": 1.043272080616479e-05,
+      "loss": 0.2001,
+      "step": 2423
+    },
+    {
+      "epoch": 1.4364444444444444,
+      "grad_norm": 2.8609585762023926,
+      "learning_rate": 1.0428769017980639e-05,
+      "loss": 0.3907,
+      "step": 2424
+    },
+    {
+      "epoch": 1.4370370370370371,
+      "grad_norm": 2.139777660369873,
+      "learning_rate": 1.0424817229796485e-05,
+      "loss": 0.2657,
+      "step": 2425
+    },
+    {
+      "epoch": 1.4376296296296296,
+      "grad_norm": 2.2758097648620605,
+      "learning_rate": 1.0420865441612331e-05,
+      "loss": 0.2785,
+      "step": 2426
+    },
+    {
+      "epoch": 1.4382222222222223,
+      "grad_norm": 2.5804178714752197,
+      "learning_rate": 1.0416913653428177e-05,
+      "loss": 0.3037,
+      "step": 2427
+    },
+    {
+      "epoch": 1.4388148148148148,
+      "grad_norm": 1.782004952430725,
+      "learning_rate": 1.0412961865244025e-05,
+      "loss": 0.2174,
+      "step": 2428
+    },
+    {
+      "epoch": 1.4394074074074075,
+      "grad_norm": 2.084263324737549,
+      "learning_rate": 1.0409010077059871e-05,
+      "loss": 0.3218,
+      "step": 2429
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.722407341003418,
+      "learning_rate": 1.0405058288875718e-05,
+      "loss": 0.3203,
+      "step": 2430
+    },
+    {
+      "epoch": 1.4405925925925926,
+      "grad_norm": 1.9336682558059692,
+      "learning_rate": 1.0401106500691564e-05,
+      "loss": 0.362,
+      "step": 2431
+    },
+    {
+      "epoch": 1.4411851851851851,
+      "grad_norm": 2.2461748123168945,
+      "learning_rate": 1.0397154712507412e-05,
+      "loss": 0.2708,
+      "step": 2432
+    },
+    {
+      "epoch": 1.4417777777777778,
+      "grad_norm": 1.4055596590042114,
+      "learning_rate": 1.0393202924323258e-05,
+      "loss": 0.2175,
+      "step": 2433
+    },
+    {
+      "epoch": 1.4423703703703703,
+      "grad_norm": 2.9104089736938477,
+      "learning_rate": 1.0389251136139104e-05,
+      "loss": 0.4098,
+      "step": 2434
+    },
+    {
+      "epoch": 1.442962962962963,
+      "grad_norm": 1.858819842338562,
+      "learning_rate": 1.038529934795495e-05,
+      "loss": 0.2846,
+      "step": 2435
+    },
+    {
+      "epoch": 1.4435555555555555,
+      "grad_norm": 1.5590959787368774,
+      "learning_rate": 1.0381347559770799e-05,
+      "loss": 0.2097,
+      "step": 2436
+    },
+    {
+      "epoch": 1.4441481481481482,
+      "grad_norm": 1.8224942684173584,
+      "learning_rate": 1.0377395771586645e-05,
+      "loss": 0.2908,
+      "step": 2437
+    },
+    {
+      "epoch": 1.4447407407407407,
+      "grad_norm": 1.6729803085327148,
+      "learning_rate": 1.0373443983402491e-05,
+      "loss": 0.3395,
+      "step": 2438
+    },
+    {
+      "epoch": 1.4453333333333334,
+      "grad_norm": 1.517288088798523,
+      "learning_rate": 1.0369492195218337e-05,
+      "loss": 0.2199,
+      "step": 2439
+    },
+    {
+      "epoch": 1.445925925925926,
+      "grad_norm": 2.6140952110290527,
+      "learning_rate": 1.0365540407034185e-05,
+      "loss": 0.3401,
+      "step": 2440
+    },
+    {
+      "epoch": 1.4465185185185185,
+      "grad_norm": 1.6827301979064941,
+      "learning_rate": 1.0361588618850031e-05,
+      "loss": 0.276,
+      "step": 2441
+    },
+    {
+      "epoch": 1.447111111111111,
+      "grad_norm": 1.6364465951919556,
+      "learning_rate": 1.0357636830665878e-05,
+      "loss": 0.3106,
+      "step": 2442
+    },
+    {
+      "epoch": 1.4477037037037037,
+      "grad_norm": 2.321970224380493,
+      "learning_rate": 1.0353685042481724e-05,
+      "loss": 0.3189,
+      "step": 2443
+    },
+    {
+      "epoch": 1.4482962962962964,
+      "grad_norm": 1.625611424446106,
+      "learning_rate": 1.0349733254297572e-05,
+      "loss": 0.2151,
+      "step": 2444
+    },
+    {
+      "epoch": 1.448888888888889,
+      "grad_norm": 1.7350125312805176,
+      "learning_rate": 1.0345781466113418e-05,
+      "loss": 0.2699,
+      "step": 2445
+    },
+    {
+      "epoch": 1.4494814814814814,
+      "grad_norm": 2.3516316413879395,
+      "learning_rate": 1.0341829677929264e-05,
+      "loss": 0.3685,
+      "step": 2446
+    },
+    {
+      "epoch": 1.450074074074074,
+      "grad_norm": 2.1124045848846436,
+      "learning_rate": 1.0337877889745112e-05,
+      "loss": 0.2475,
+      "step": 2447
+    },
+    {
+      "epoch": 1.4506666666666668,
+      "grad_norm": 2.0750977993011475,
+      "learning_rate": 1.0333926101560959e-05,
+      "loss": 0.2958,
+      "step": 2448
+    },
+    {
+      "epoch": 1.4512592592592592,
+      "grad_norm": 1.7762513160705566,
+      "learning_rate": 1.0329974313376805e-05,
+      "loss": 0.2675,
+      "step": 2449
+    },
+    {
+      "epoch": 1.4518518518518517,
+      "grad_norm": 1.6564881801605225,
+      "learning_rate": 1.0326022525192651e-05,
+      "loss": 0.2263,
+      "step": 2450
+    },
+    {
+      "epoch": 1.4524444444444444,
+      "grad_norm": 2.924825429916382,
+      "learning_rate": 1.0322070737008499e-05,
+      "loss": 0.3642,
+      "step": 2451
+    },
+    {
+      "epoch": 1.4530370370370371,
+      "grad_norm": 1.9329086542129517,
+      "learning_rate": 1.0318118948824345e-05,
+      "loss": 0.3284,
+      "step": 2452
+    },
+    {
+      "epoch": 1.4536296296296296,
+      "grad_norm": 2.8134090900421143,
+      "learning_rate": 1.0314167160640191e-05,
+      "loss": 0.2781,
+      "step": 2453
+    },
+    {
+      "epoch": 1.4542222222222223,
+      "grad_norm": 2.549788236618042,
+      "learning_rate": 1.0310215372456036e-05,
+      "loss": 0.2778,
+      "step": 2454
+    },
+    {
+      "epoch": 1.4548148148148148,
+      "grad_norm": 2.533801317214966,
+      "learning_rate": 1.0306263584271882e-05,
+      "loss": 0.3051,
+      "step": 2455
+    },
+    {
+      "epoch": 1.4554074074074075,
+      "grad_norm": 2.0454323291778564,
+      "learning_rate": 1.030231179608773e-05,
+      "loss": 0.3144,
+      "step": 2456
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 1.457763433456421,
+      "learning_rate": 1.0298360007903576e-05,
+      "loss": 0.251,
+      "step": 2457
+    },
+    {
+      "epoch": 1.4565925925925927,
+      "grad_norm": 1.7523553371429443,
+      "learning_rate": 1.0294408219719423e-05,
+      "loss": 0.2154,
+      "step": 2458
+    },
+    {
+      "epoch": 1.4571851851851851,
+      "grad_norm": 1.761773705482483,
+      "learning_rate": 1.0290456431535269e-05,
+      "loss": 0.2206,
+      "step": 2459
+    },
+    {
+      "epoch": 1.4577777777777778,
+      "grad_norm": 2.183199882507324,
+      "learning_rate": 1.0286504643351117e-05,
+      "loss": 0.2859,
+      "step": 2460
+    },
+    {
+      "epoch": 1.4583703703703703,
+      "grad_norm": 2.132707357406616,
+      "learning_rate": 1.0282552855166963e-05,
+      "loss": 0.2935,
+      "step": 2461
+    },
+    {
+      "epoch": 1.458962962962963,
+      "grad_norm": 1.8507558107376099,
+      "learning_rate": 1.027860106698281e-05,
+      "loss": 0.2431,
+      "step": 2462
+    },
+    {
+      "epoch": 1.4595555555555555,
+      "grad_norm": 1.5582311153411865,
+      "learning_rate": 1.0274649278798656e-05,
+      "loss": 0.2549,
+      "step": 2463
+    },
+    {
+      "epoch": 1.4601481481481482,
+      "grad_norm": 1.7385444641113281,
+      "learning_rate": 1.0270697490614504e-05,
+      "loss": 0.2947,
+      "step": 2464
+    },
+    {
+      "epoch": 1.4607407407407407,
+      "grad_norm": 2.0530834197998047,
+      "learning_rate": 1.026674570243035e-05,
+      "loss": 0.2841,
+      "step": 2465
+    },
+    {
+      "epoch": 1.4613333333333334,
+      "grad_norm": 1.6098973751068115,
+      "learning_rate": 1.0262793914246196e-05,
+      "loss": 0.2039,
+      "step": 2466
+    },
+    {
+      "epoch": 1.4619259259259259,
+      "grad_norm": 2.379680871963501,
+      "learning_rate": 1.0258842126062042e-05,
+      "loss": 0.3131,
+      "step": 2467
+    },
+    {
+      "epoch": 1.4625185185185186,
+      "grad_norm": 2.18462872505188,
+      "learning_rate": 1.025489033787789e-05,
+      "loss": 0.2445,
+      "step": 2468
+    },
+    {
+      "epoch": 1.463111111111111,
+      "grad_norm": 1.6083557605743408,
+      "learning_rate": 1.0250938549693736e-05,
+      "loss": 0.2494,
+      "step": 2469
+    },
+    {
+      "epoch": 1.4637037037037037,
+      "grad_norm": 2.119189739227295,
+      "learning_rate": 1.0246986761509583e-05,
+      "loss": 0.3057,
+      "step": 2470
+    },
+    {
+      "epoch": 1.4642962962962964,
+      "grad_norm": 2.392836093902588,
+      "learning_rate": 1.0243034973325429e-05,
+      "loss": 0.3711,
+      "step": 2471
+    },
+    {
+      "epoch": 1.464888888888889,
+      "grad_norm": 2.0722458362579346,
+      "learning_rate": 1.0239083185141277e-05,
+      "loss": 0.2941,
+      "step": 2472
+    },
+    {
+      "epoch": 1.4654814814814814,
+      "grad_norm": 2.2288243770599365,
+      "learning_rate": 1.0235131396957123e-05,
+      "loss": 0.2855,
+      "step": 2473
+    },
+    {
+      "epoch": 1.466074074074074,
+      "grad_norm": 2.136747360229492,
+      "learning_rate": 1.023117960877297e-05,
+      "loss": 0.309,
+      "step": 2474
+    },
+    {
+      "epoch": 1.4666666666666668,
+      "grad_norm": 2.216593027114868,
+      "learning_rate": 1.0227227820588817e-05,
+      "loss": 0.3684,
+      "step": 2475
+    },
+    {
+      "epoch": 1.4672592592592593,
+      "grad_norm": 2.045522451400757,
+      "learning_rate": 1.0223276032404664e-05,
+      "loss": 0.2674,
+      "step": 2476
+    },
+    {
+      "epoch": 1.4678518518518517,
+      "grad_norm": 3.316455364227295,
+      "learning_rate": 1.021932424422051e-05,
+      "loss": 0.2973,
+      "step": 2477
+    },
+    {
+      "epoch": 1.4684444444444444,
+      "grad_norm": 2.086486339569092,
+      "learning_rate": 1.0215372456036356e-05,
+      "loss": 0.3077,
+      "step": 2478
+    },
+    {
+      "epoch": 1.4690370370370371,
+      "grad_norm": 1.9176051616668701,
+      "learning_rate": 1.0211420667852204e-05,
+      "loss": 0.2514,
+      "step": 2479
+    },
+    {
+      "epoch": 1.4696296296296296,
+      "grad_norm": 3.183574676513672,
+      "learning_rate": 1.020746887966805e-05,
+      "loss": 0.3014,
+      "step": 2480
+    },
+    {
+      "epoch": 1.470222222222222,
+      "grad_norm": 1.997645616531372,
+      "learning_rate": 1.0203517091483896e-05,
+      "loss": 0.305,
+      "step": 2481
+    },
+    {
+      "epoch": 1.4708148148148148,
+      "grad_norm": 3.498404026031494,
+      "learning_rate": 1.0199565303299743e-05,
+      "loss": 0.3862,
+      "step": 2482
+    },
+    {
+      "epoch": 1.4714074074074075,
+      "grad_norm": 2.111151695251465,
+      "learning_rate": 1.019561351511559e-05,
+      "loss": 0.2773,
+      "step": 2483
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 1.8812440633773804,
+      "learning_rate": 1.0191661726931437e-05,
+      "loss": 0.2175,
+      "step": 2484
+    },
+    {
+      "epoch": 1.4725925925925927,
+      "grad_norm": 1.9070371389389038,
+      "learning_rate": 1.0187709938747283e-05,
+      "loss": 0.3605,
+      "step": 2485
+    },
+    {
+      "epoch": 1.4731851851851852,
+      "grad_norm": 2.4770705699920654,
+      "learning_rate": 1.018375815056313e-05,
+      "loss": 0.2577,
+      "step": 2486
+    },
+    {
+      "epoch": 1.4737777777777779,
+      "grad_norm": 2.087031126022339,
+      "learning_rate": 1.0179806362378977e-05,
+      "loss": 0.3431,
+      "step": 2487
+    },
+    {
+      "epoch": 1.4743703703703703,
+      "grad_norm": 2.344994068145752,
+      "learning_rate": 1.0175854574194824e-05,
+      "loss": 0.3484,
+      "step": 2488
+    },
+    {
+      "epoch": 1.474962962962963,
+      "grad_norm": 1.4907705783843994,
+      "learning_rate": 1.017190278601067e-05,
+      "loss": 0.2417,
+      "step": 2489
+    },
+    {
+      "epoch": 1.4755555555555555,
+      "grad_norm": 2.3776488304138184,
+      "learning_rate": 1.0167950997826516e-05,
+      "loss": 0.3445,
+      "step": 2490
+    },
+    {
+      "epoch": 1.4761481481481482,
+      "grad_norm": 2.405644416809082,
+      "learning_rate": 1.0163999209642364e-05,
+      "loss": 0.4147,
+      "step": 2491
+    },
+    {
+      "epoch": 1.4767407407407407,
+      "grad_norm": 1.8785308599472046,
+      "learning_rate": 1.016004742145821e-05,
+      "loss": 0.245,
+      "step": 2492
+    },
+    {
+      "epoch": 1.4773333333333334,
+      "grad_norm": 1.829829454421997,
+      "learning_rate": 1.0156095633274056e-05,
+      "loss": 0.2766,
+      "step": 2493
+    },
+    {
+      "epoch": 1.4779259259259259,
+      "grad_norm": 2.046363115310669,
+      "learning_rate": 1.0152143845089903e-05,
+      "loss": 0.3414,
+      "step": 2494
+    },
+    {
+      "epoch": 1.4785185185185186,
+      "grad_norm": 1.7704737186431885,
+      "learning_rate": 1.014819205690575e-05,
+      "loss": 0.2109,
+      "step": 2495
+    },
+    {
+      "epoch": 1.479111111111111,
+      "grad_norm": 2.4176595211029053,
+      "learning_rate": 1.0144240268721597e-05,
+      "loss": 0.3065,
+      "step": 2496
+    },
+    {
+      "epoch": 1.4797037037037037,
+      "grad_norm": 2.5289337635040283,
+      "learning_rate": 1.0140288480537443e-05,
+      "loss": 0.3246,
+      "step": 2497
+    },
+    {
+      "epoch": 1.4802962962962962,
+      "grad_norm": 1.8737410306930542,
+      "learning_rate": 1.0136336692353291e-05,
+      "loss": 0.2923,
+      "step": 2498
+    },
+    {
+      "epoch": 1.480888888888889,
+      "grad_norm": 1.662710189819336,
+      "learning_rate": 1.0132384904169137e-05,
+      "loss": 0.2261,
+      "step": 2499
+    },
+    {
+      "epoch": 1.4814814814814814,
+      "grad_norm": 1.6063086986541748,
+      "learning_rate": 1.0128433115984984e-05,
+      "loss": 0.2109,
+      "step": 2500
+    },
+    {
+      "epoch": 1.482074074074074,
+      "grad_norm": 3.7506885528564453,
+      "learning_rate": 1.012448132780083e-05,
+      "loss": 0.258,
+      "step": 2501
+    },
+    {
+      "epoch": 1.4826666666666668,
+      "grad_norm": 1.8277090787887573,
+      "learning_rate": 1.0120529539616678e-05,
+      "loss": 0.287,
+      "step": 2502
+    },
+    {
+      "epoch": 1.4832592592592593,
+      "grad_norm": 2.378685235977173,
+      "learning_rate": 1.0116577751432524e-05,
+      "loss": 0.2321,
+      "step": 2503
+    },
+    {
+      "epoch": 1.4838518518518518,
+      "grad_norm": 1.4849727153778076,
+      "learning_rate": 1.011262596324837e-05,
+      "loss": 0.2362,
+      "step": 2504
+    },
+    {
+      "epoch": 1.4844444444444445,
+      "grad_norm": 1.9681180715560913,
+      "learning_rate": 1.0108674175064216e-05,
+      "loss": 0.2525,
+      "step": 2505
+    },
+    {
+      "epoch": 1.4850370370370372,
+      "grad_norm": 2.0627360343933105,
+      "learning_rate": 1.0104722386880064e-05,
+      "loss": 0.3433,
+      "step": 2506
+    },
+    {
+      "epoch": 1.4856296296296296,
+      "grad_norm": 2.2477810382843018,
+      "learning_rate": 1.010077059869591e-05,
+      "loss": 0.3156,
+      "step": 2507
+    },
+    {
+      "epoch": 1.4862222222222221,
+      "grad_norm": 2.7742116451263428,
+      "learning_rate": 1.0096818810511757e-05,
+      "loss": 0.2677,
+      "step": 2508
+    },
+    {
+      "epoch": 1.4868148148148148,
+      "grad_norm": 2.69305682182312,
+      "learning_rate": 1.0092867022327603e-05,
+      "loss": 0.2026,
+      "step": 2509
+    },
+    {
+      "epoch": 1.4874074074074075,
+      "grad_norm": 2.333003282546997,
+      "learning_rate": 1.0088915234143451e-05,
+      "loss": 0.2918,
+      "step": 2510
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 2.6414763927459717,
+      "learning_rate": 1.0084963445959297e-05,
+      "loss": 0.3291,
+      "step": 2511
+    },
+    {
+      "epoch": 1.4885925925925925,
+      "grad_norm": 2.2082483768463135,
+      "learning_rate": 1.0081011657775144e-05,
+      "loss": 0.2765,
+      "step": 2512
+    },
+    {
+      "epoch": 1.4891851851851852,
+      "grad_norm": 2.12756085395813,
+      "learning_rate": 1.007705986959099e-05,
+      "loss": 0.342,
+      "step": 2513
+    },
+    {
+      "epoch": 1.4897777777777779,
+      "grad_norm": 2.0906476974487305,
+      "learning_rate": 1.0073108081406838e-05,
+      "loss": 0.2542,
+      "step": 2514
+    },
+    {
+      "epoch": 1.4903703703703703,
+      "grad_norm": 1.8887280225753784,
+      "learning_rate": 1.0069156293222684e-05,
+      "loss": 0.3161,
+      "step": 2515
+    },
+    {
+      "epoch": 1.490962962962963,
+      "grad_norm": 2.130493640899658,
+      "learning_rate": 1.006520450503853e-05,
+      "loss": 0.2518,
+      "step": 2516
+    },
+    {
+      "epoch": 1.4915555555555555,
+      "grad_norm": 2.4801807403564453,
+      "learning_rate": 1.0061252716854376e-05,
+      "loss": 0.3473,
+      "step": 2517
+    },
+    {
+      "epoch": 1.4921481481481482,
+      "grad_norm": 1.2939742803573608,
+      "learning_rate": 1.0057300928670224e-05,
+      "loss": 0.1919,
+      "step": 2518
+    },
+    {
+      "epoch": 1.4927407407407407,
+      "grad_norm": 1.7417817115783691,
+      "learning_rate": 1.005334914048607e-05,
+      "loss": 0.2583,
+      "step": 2519
+    },
+    {
+      "epoch": 1.4933333333333334,
+      "grad_norm": 2.520718574523926,
+      "learning_rate": 1.0049397352301917e-05,
+      "loss": 0.3119,
+      "step": 2520
+    },
+    {
+      "epoch": 1.4939259259259259,
+      "grad_norm": 2.0422861576080322,
+      "learning_rate": 1.0045445564117765e-05,
+      "loss": 0.2803,
+      "step": 2521
+    },
+    {
+      "epoch": 1.4945185185185186,
+      "grad_norm": 2.4748852252960205,
+      "learning_rate": 1.0041493775933611e-05,
+      "loss": 0.2893,
+      "step": 2522
+    },
+    {
+      "epoch": 1.495111111111111,
+      "grad_norm": 2.804664373397827,
+      "learning_rate": 1.0037541987749457e-05,
+      "loss": 0.4192,
+      "step": 2523
+    },
+    {
+      "epoch": 1.4957037037037038,
+      "grad_norm": 2.690760374069214,
+      "learning_rate": 1.0033590199565304e-05,
+      "loss": 0.3361,
+      "step": 2524
+    },
+    {
+      "epoch": 1.4962962962962962,
+      "grad_norm": 1.8734771013259888,
+      "learning_rate": 1.0029638411381151e-05,
+      "loss": 0.2435,
+      "step": 2525
+    },
+    {
+      "epoch": 1.496888888888889,
+      "grad_norm": 2.0335495471954346,
+      "learning_rate": 1.0025686623196998e-05,
+      "loss": 0.3471,
+      "step": 2526
+    },
+    {
+      "epoch": 1.4974814814814814,
+      "grad_norm": 2.623274087905884,
+      "learning_rate": 1.0021734835012844e-05,
+      "loss": 0.3338,
+      "step": 2527
+    },
+    {
+      "epoch": 1.4980740740740741,
+      "grad_norm": 1.5604110956192017,
+      "learning_rate": 1.001778304682869e-05,
+      "loss": 0.3058,
+      "step": 2528
+    },
+    {
+      "epoch": 1.4986666666666666,
+      "grad_norm": 1.873975396156311,
+      "learning_rate": 1.0013831258644538e-05,
+      "loss": 0.2971,
+      "step": 2529
+    },
+    {
+      "epoch": 1.4992592592592593,
+      "grad_norm": 2.8412981033325195,
+      "learning_rate": 1.0009879470460384e-05,
+      "loss": 0.5108,
+      "step": 2530
+    },
+    {
+      "epoch": 1.4998518518518518,
+      "grad_norm": 3.59525203704834,
+      "learning_rate": 1.000592768227623e-05,
+      "loss": 0.3316,
+      "step": 2531
+    },
+    {
+      "epoch": 1.5004444444444445,
+      "grad_norm": 2.6374852657318115,
+      "learning_rate": 1.0001975894092077e-05,
+      "loss": 0.2013,
+      "step": 2532
+    },
+    {
+      "epoch": 1.5010370370370372,
+      "grad_norm": 2.3971266746520996,
+      "learning_rate": 9.998024105907925e-06,
+      "loss": 0.3369,
+      "step": 2533
+    },
+    {
+      "epoch": 1.5016296296296296,
+      "grad_norm": 3.3115265369415283,
+      "learning_rate": 9.994072317723771e-06,
+      "loss": 0.3806,
+      "step": 2534
+    },
+    {
+      "epoch": 1.5022222222222221,
+      "grad_norm": 1.8655791282653809,
+      "learning_rate": 9.990120529539617e-06,
+      "loss": 0.2036,
+      "step": 2535
+    },
+    {
+      "epoch": 1.5028148148148148,
+      "grad_norm": 1.9425292015075684,
+      "learning_rate": 9.986168741355464e-06,
+      "loss": 0.2426,
+      "step": 2536
+    },
+    {
+      "epoch": 1.5034074074074075,
+      "grad_norm": 2.032224178314209,
+      "learning_rate": 9.982216953171311e-06,
+      "loss": 0.2665,
+      "step": 2537
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 2.848874807357788,
+      "learning_rate": 9.978265164987158e-06,
+      "loss": 0.2863,
+      "step": 2538
+    },
+    {
+      "epoch": 1.5045925925925925,
+      "grad_norm": 2.469555377960205,
+      "learning_rate": 9.974313376803004e-06,
+      "loss": 0.2403,
+      "step": 2539
+    },
+    {
+      "epoch": 1.5051851851851852,
+      "grad_norm": 2.697622299194336,
+      "learning_rate": 9.97036158861885e-06,
+      "loss": 0.3026,
+      "step": 2540
+    },
+    {
+      "epoch": 1.5057777777777779,
+      "grad_norm": 1.7998290061950684,
+      "learning_rate": 9.966409800434698e-06,
+      "loss": 0.2982,
+      "step": 2541
+    },
+    {
+      "epoch": 1.5063703703703704,
+      "grad_norm": 1.5836988687515259,
+      "learning_rate": 9.962458012250544e-06,
+      "loss": 0.2634,
+      "step": 2542
+    },
+    {
+      "epoch": 1.5069629629629628,
+      "grad_norm": 3.286146402359009,
+      "learning_rate": 9.95850622406639e-06,
+      "loss": 0.2905,
+      "step": 2543
+    },
+    {
+      "epoch": 1.5075555555555555,
+      "grad_norm": 2.373324155807495,
+      "learning_rate": 9.954554435882237e-06,
+      "loss": 0.2157,
+      "step": 2544
+    },
+    {
+      "epoch": 1.5081481481481482,
+      "grad_norm": 2.099912643432617,
+      "learning_rate": 9.950602647698085e-06,
+      "loss": 0.3763,
+      "step": 2545
+    },
+    {
+      "epoch": 1.5087407407407407,
+      "grad_norm": 2.426046133041382,
+      "learning_rate": 9.946650859513931e-06,
+      "loss": 0.2956,
+      "step": 2546
+    },
+    {
+      "epoch": 1.5093333333333332,
+      "grad_norm": 2.247101306915283,
+      "learning_rate": 9.942699071329777e-06,
+      "loss": 0.2176,
+      "step": 2547
+    },
+    {
+      "epoch": 1.509925925925926,
+      "grad_norm": 1.8053836822509766,
+      "learning_rate": 9.938747283145625e-06,
+      "loss": 0.2307,
+      "step": 2548
+    },
+    {
+      "epoch": 1.5105185185185186,
+      "grad_norm": 1.8379324674606323,
+      "learning_rate": 9.934795494961471e-06,
+      "loss": 0.3164,
+      "step": 2549
+    },
+    {
+      "epoch": 1.511111111111111,
+      "grad_norm": 1.8471317291259766,
+      "learning_rate": 9.930843706777318e-06,
+      "loss": 0.2496,
+      "step": 2550
+    },
+    {
+      "epoch": 1.5117037037037035,
+      "grad_norm": 1.6880301237106323,
+      "learning_rate": 9.926891918593164e-06,
+      "loss": 0.2883,
+      "step": 2551
+    },
+    {
+      "epoch": 1.5122962962962962,
+      "grad_norm": 3.2706222534179688,
+      "learning_rate": 9.922940130409012e-06,
+      "loss": 0.2042,
+      "step": 2552
+    },
+    {
+      "epoch": 1.512888888888889,
+      "grad_norm": 3.9777514934539795,
+      "learning_rate": 9.918988342224858e-06,
+      "loss": 0.3292,
+      "step": 2553
+    },
+    {
+      "epoch": 1.5134814814814814,
+      "grad_norm": 1.7576566934585571,
+      "learning_rate": 9.915036554040704e-06,
+      "loss": 0.301,
+      "step": 2554
+    },
+    {
+      "epoch": 1.5140740740740741,
+      "grad_norm": 1.816907525062561,
+      "learning_rate": 9.91108476585655e-06,
+      "loss": 0.2192,
+      "step": 2555
+    },
+    {
+      "epoch": 1.5146666666666668,
+      "grad_norm": 1.6303327083587646,
+      "learning_rate": 9.907132977672399e-06,
+      "loss": 0.2214,
+      "step": 2556
+    },
+    {
+      "epoch": 1.5152592592592593,
+      "grad_norm": 1.368719458580017,
+      "learning_rate": 9.903181189488245e-06,
+      "loss": 0.2185,
+      "step": 2557
+    },
+    {
+      "epoch": 1.5158518518518518,
+      "grad_norm": 2.5178732872009277,
+      "learning_rate": 9.899229401304091e-06,
+      "loss": 0.3428,
+      "step": 2558
+    },
+    {
+      "epoch": 1.5164444444444445,
+      "grad_norm": 2.0861001014709473,
+      "learning_rate": 9.895277613119937e-06,
+      "loss": 0.2559,
+      "step": 2559
+    },
+    {
+      "epoch": 1.5170370370370372,
+      "grad_norm": 3.6666531562805176,
+      "learning_rate": 9.891325824935785e-06,
+      "loss": 0.2633,
+      "step": 2560
+    },
+    {
+      "epoch": 1.5176296296296297,
+      "grad_norm": 2.0326833724975586,
+      "learning_rate": 9.887374036751631e-06,
+      "loss": 0.2739,
+      "step": 2561
+    },
+    {
+      "epoch": 1.5182222222222221,
+      "grad_norm": 1.583383560180664,
+      "learning_rate": 9.883422248567478e-06,
+      "loss": 0.2293,
+      "step": 2562
+    },
+    {
+      "epoch": 1.5188148148148148,
+      "grad_norm": 2.160085439682007,
+      "learning_rate": 9.879470460383324e-06,
+      "loss": 0.2906,
+      "step": 2563
+    },
+    {
+      "epoch": 1.5194074074074075,
+      "grad_norm": 2.4415204524993896,
+      "learning_rate": 9.875518672199172e-06,
+      "loss": 0.3529,
+      "step": 2564
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.773586630821228,
+      "learning_rate": 9.871566884015018e-06,
+      "loss": 0.2874,
+      "step": 2565
+    },
+    {
+      "epoch": 1.5205925925925925,
+      "grad_norm": 1.685492753982544,
+      "learning_rate": 9.867615095830864e-06,
+      "loss": 0.2316,
+      "step": 2566
+    },
+    {
+      "epoch": 1.5211851851851852,
+      "grad_norm": 1.88389253616333,
+      "learning_rate": 9.86366330764671e-06,
+      "loss": 0.2988,
+      "step": 2567
+    },
+    {
+      "epoch": 1.521777777777778,
+      "grad_norm": 2.2108607292175293,
+      "learning_rate": 9.859711519462559e-06,
+      "loss": 0.2856,
+      "step": 2568
+    },
+    {
+      "epoch": 1.5223703703703704,
+      "grad_norm": 2.0331039428710938,
+      "learning_rate": 9.855759731278405e-06,
+      "loss": 0.2904,
+      "step": 2569
+    },
+    {
+      "epoch": 1.5229629629629629,
+      "grad_norm": 1.9677793979644775,
+      "learning_rate": 9.851807943094251e-06,
+      "loss": 0.2731,
+      "step": 2570
+    },
+    {
+      "epoch": 1.5235555555555556,
+      "grad_norm": 1.6250696182250977,
+      "learning_rate": 9.847856154910099e-06,
+      "loss": 0.2486,
+      "step": 2571
+    },
+    {
+      "epoch": 1.5241481481481483,
+      "grad_norm": 2.280285358428955,
+      "learning_rate": 9.843904366725945e-06,
+      "loss": 0.2662,
+      "step": 2572
+    },
+    {
+      "epoch": 1.5247407407407407,
+      "grad_norm": 1.4999984502792358,
+      "learning_rate": 9.83995257854179e-06,
+      "loss": 0.1716,
+      "step": 2573
+    },
+    {
+      "epoch": 1.5253333333333332,
+      "grad_norm": 2.1147375106811523,
+      "learning_rate": 9.836000790357638e-06,
+      "loss": 0.3807,
+      "step": 2574
+    },
+    {
+      "epoch": 1.525925925925926,
+      "grad_norm": 1.8736424446105957,
+      "learning_rate": 9.832049002173484e-06,
+      "loss": 0.297,
+      "step": 2575
+    },
+    {
+      "epoch": 1.5265185185185186,
+      "grad_norm": 1.7065719366073608,
+      "learning_rate": 9.82809721398933e-06,
+      "loss": 0.1461,
+      "step": 2576
+    },
+    {
+      "epoch": 1.527111111111111,
+      "grad_norm": 1.6719976663589478,
+      "learning_rate": 9.824145425805176e-06,
+      "loss": 0.2411,
+      "step": 2577
+    },
+    {
+      "epoch": 1.5277037037037036,
+      "grad_norm": 2.1352367401123047,
+      "learning_rate": 9.820193637621024e-06,
+      "loss": 0.2666,
+      "step": 2578
+    },
+    {
+      "epoch": 1.5282962962962963,
+      "grad_norm": 2.3025596141815186,
+      "learning_rate": 9.81624184943687e-06,
+      "loss": 0.286,
+      "step": 2579
+    },
+    {
+      "epoch": 1.528888888888889,
+      "grad_norm": 2.4366211891174316,
+      "learning_rate": 9.812290061252717e-06,
+      "loss": 0.3615,
+      "step": 2580
+    },
+    {
+      "epoch": 1.5294814814814814,
+      "grad_norm": 1.075118064880371,
+      "learning_rate": 9.808338273068563e-06,
+      "loss": 0.1765,
+      "step": 2581
+    },
+    {
+      "epoch": 1.530074074074074,
+      "grad_norm": 1.4529460668563843,
+      "learning_rate": 9.804386484884411e-06,
+      "loss": 0.208,
+      "step": 2582
+    },
+    {
+      "epoch": 1.5306666666666666,
+      "grad_norm": 1.7463024854660034,
+      "learning_rate": 9.800434696700257e-06,
+      "loss": 0.234,
+      "step": 2583
+    },
+    {
+      "epoch": 1.5312592592592593,
+      "grad_norm": 1.876061201095581,
+      "learning_rate": 9.796482908516104e-06,
+      "loss": 0.1787,
+      "step": 2584
+    },
+    {
+      "epoch": 1.5318518518518518,
+      "grad_norm": 1.9467613697052002,
+      "learning_rate": 9.792531120331951e-06,
+      "loss": 0.2917,
+      "step": 2585
+    },
+    {
+      "epoch": 1.5324444444444445,
+      "grad_norm": 1.906068205833435,
+      "learning_rate": 9.788579332147798e-06,
+      "loss": 0.3659,
+      "step": 2586
+    },
+    {
+      "epoch": 1.5330370370370372,
+      "grad_norm": 2.8988609313964844,
+      "learning_rate": 9.784627543963644e-06,
+      "loss": 0.346,
+      "step": 2587
+    },
+    {
+      "epoch": 1.5336296296296297,
+      "grad_norm": 2.1930248737335205,
+      "learning_rate": 9.78067575577949e-06,
+      "loss": 0.3785,
+      "step": 2588
+    },
+    {
+      "epoch": 1.5342222222222222,
+      "grad_norm": 2.012111186981201,
+      "learning_rate": 9.776723967595338e-06,
+      "loss": 0.1605,
+      "step": 2589
+    },
+    {
+      "epoch": 1.5348148148148149,
+      "grad_norm": 1.9958025217056274,
+      "learning_rate": 9.772772179411184e-06,
+      "loss": 0.2422,
+      "step": 2590
+    },
+    {
+      "epoch": 1.5354074074074076,
+      "grad_norm": 1.579609990119934,
+      "learning_rate": 9.76882039122703e-06,
+      "loss": 0.1805,
+      "step": 2591
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 2.01224946975708,
+      "learning_rate": 9.764868603042877e-06,
+      "loss": 0.2519,
+      "step": 2592
+    },
+    {
+      "epoch": 1.5365925925925925,
+      "grad_norm": 2.4626107215881348,
+      "learning_rate": 9.760916814858725e-06,
+      "loss": 0.3127,
+      "step": 2593
+    },
+    {
+      "epoch": 1.5371851851851852,
+      "grad_norm": 1.856805682182312,
+      "learning_rate": 9.756965026674571e-06,
+      "loss": 0.2886,
+      "step": 2594
+    },
+    {
+      "epoch": 1.537777777777778,
+      "grad_norm": 1.7355167865753174,
+      "learning_rate": 9.753013238490417e-06,
+      "loss": 0.2548,
+      "step": 2595
+    },
+    {
+      "epoch": 1.5383703703703704,
+      "grad_norm": 1.8150662183761597,
+      "learning_rate": 9.749061450306263e-06,
+      "loss": 0.2005,
+      "step": 2596
+    },
+    {
+      "epoch": 1.5389629629629629,
+      "grad_norm": 2.428880453109741,
+      "learning_rate": 9.745109662122111e-06,
+      "loss": 0.2848,
+      "step": 2597
+    },
+    {
+      "epoch": 1.5395555555555556,
+      "grad_norm": 1.7132172584533691,
+      "learning_rate": 9.741157873937958e-06,
+      "loss": 0.1887,
+      "step": 2598
+    },
+    {
+      "epoch": 1.5401481481481483,
+      "grad_norm": 2.9255125522613525,
+      "learning_rate": 9.737206085753804e-06,
+      "loss": 0.3499,
+      "step": 2599
+    },
+    {
+      "epoch": 1.5407407407407407,
+      "grad_norm": 2.2438950538635254,
+      "learning_rate": 9.73325429756965e-06,
+      "loss": 0.2307,
+      "step": 2600
+    },
+    {
+      "epoch": 1.5413333333333332,
+      "grad_norm": 2.595252513885498,
+      "learning_rate": 9.729302509385498e-06,
+      "loss": 0.2225,
+      "step": 2601
+    },
+    {
+      "epoch": 1.541925925925926,
+      "grad_norm": 2.055124521255493,
+      "learning_rate": 9.725350721201344e-06,
+      "loss": 0.2926,
+      "step": 2602
+    },
+    {
+      "epoch": 1.5425185185185186,
+      "grad_norm": 2.1323089599609375,
+      "learning_rate": 9.72139893301719e-06,
+      "loss": 0.2275,
+      "step": 2603
+    },
+    {
+      "epoch": 1.543111111111111,
+      "grad_norm": 1.8309077024459839,
+      "learning_rate": 9.717447144833037e-06,
+      "loss": 0.3574,
+      "step": 2604
+    },
+    {
+      "epoch": 1.5437037037037036,
+      "grad_norm": 1.6672463417053223,
+      "learning_rate": 9.713495356648885e-06,
+      "loss": 0.2291,
+      "step": 2605
+    },
+    {
+      "epoch": 1.5442962962962963,
+      "grad_norm": 1.955834150314331,
+      "learning_rate": 9.709543568464731e-06,
+      "loss": 0.2709,
+      "step": 2606
+    },
+    {
+      "epoch": 1.544888888888889,
+      "grad_norm": 1.9834613800048828,
+      "learning_rate": 9.705591780280577e-06,
+      "loss": 0.2747,
+      "step": 2607
+    },
+    {
+      "epoch": 1.5454814814814815,
+      "grad_norm": 1.6157745122909546,
+      "learning_rate": 9.701639992096425e-06,
+      "loss": 0.21,
+      "step": 2608
+    },
+    {
+      "epoch": 1.546074074074074,
+      "grad_norm": 1.7380040884017944,
+      "learning_rate": 9.697688203912271e-06,
+      "loss": 0.2832,
+      "step": 2609
+    },
+    {
+      "epoch": 1.5466666666666666,
+      "grad_norm": 1.6477042436599731,
+      "learning_rate": 9.693736415728118e-06,
+      "loss": 0.242,
+      "step": 2610
+    },
+    {
+      "epoch": 1.5472592592592593,
+      "grad_norm": 1.5338919162750244,
+      "learning_rate": 9.689784627543964e-06,
+      "loss": 0.1609,
+      "step": 2611
+    },
+    {
+      "epoch": 1.5478518518518518,
+      "grad_norm": 1.6246225833892822,
+      "learning_rate": 9.685832839359812e-06,
+      "loss": 0.2959,
+      "step": 2612
+    },
+    {
+      "epoch": 1.5484444444444443,
+      "grad_norm": 4.746414661407471,
+      "learning_rate": 9.681881051175658e-06,
+      "loss": 0.4797,
+      "step": 2613
+    },
+    {
+      "epoch": 1.549037037037037,
+      "grad_norm": 2.0302021503448486,
+      "learning_rate": 9.677929262991504e-06,
+      "loss": 0.35,
+      "step": 2614
+    },
+    {
+      "epoch": 1.5496296296296297,
+      "grad_norm": 1.8330540657043457,
+      "learning_rate": 9.67397747480735e-06,
+      "loss": 0.2869,
+      "step": 2615
+    },
+    {
+      "epoch": 1.5502222222222222,
+      "grad_norm": 1.951296091079712,
+      "learning_rate": 9.670025686623199e-06,
+      "loss": 0.3399,
+      "step": 2616
+    },
+    {
+      "epoch": 1.5508148148148149,
+      "grad_norm": 1.561105728149414,
+      "learning_rate": 9.666073898439045e-06,
+      "loss": 0.2563,
+      "step": 2617
+    },
+    {
+      "epoch": 1.5514074074074076,
+      "grad_norm": 2.552166700363159,
+      "learning_rate": 9.662122110254891e-06,
+      "loss": 0.2887,
+      "step": 2618
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 1.863504409790039,
+      "learning_rate": 9.658170322070737e-06,
+      "loss": 0.2852,
+      "step": 2619
+    },
+    {
+      "epoch": 1.5525925925925925,
+      "grad_norm": 1.6533026695251465,
+      "learning_rate": 9.654218533886585e-06,
+      "loss": 0.1966,
+      "step": 2620
+    },
+    {
+      "epoch": 1.5531851851851852,
+      "grad_norm": 1.5940282344818115,
+      "learning_rate": 9.650266745702431e-06,
+      "loss": 0.2866,
+      "step": 2621
+    },
+    {
+      "epoch": 1.553777777777778,
+      "grad_norm": 1.6983367204666138,
+      "learning_rate": 9.646314957518278e-06,
+      "loss": 0.2609,
+      "step": 2622
+    },
+    {
+      "epoch": 1.5543703703703704,
+      "grad_norm": 1.8362126350402832,
+      "learning_rate": 9.642363169334124e-06,
+      "loss": 0.2816,
+      "step": 2623
+    },
+    {
+      "epoch": 1.5549629629629629,
+      "grad_norm": 1.9165643453598022,
+      "learning_rate": 9.638411381149972e-06,
+      "loss": 0.3526,
+      "step": 2624
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 2.34714937210083,
+      "learning_rate": 9.634459592965818e-06,
+      "loss": 0.2966,
+      "step": 2625
+    },
+    {
+      "epoch": 1.5561481481481483,
+      "grad_norm": 1.4744517803192139,
+      "learning_rate": 9.630507804781664e-06,
+      "loss": 0.2214,
+      "step": 2626
+    },
+    {
+      "epoch": 1.5567407407407408,
+      "grad_norm": 1.7968119382858276,
+      "learning_rate": 9.62655601659751e-06,
+      "loss": 0.3292,
+      "step": 2627
+    },
+    {
+      "epoch": 1.5573333333333332,
+      "grad_norm": 2.09110951423645,
+      "learning_rate": 9.622604228413359e-06,
+      "loss": 0.2712,
+      "step": 2628
+    },
+    {
+      "epoch": 1.557925925925926,
+      "grad_norm": 2.8744876384735107,
+      "learning_rate": 9.618652440229205e-06,
+      "loss": 0.399,
+      "step": 2629
+    },
+    {
+      "epoch": 1.5585185185185186,
+      "grad_norm": 3.2395853996276855,
+      "learning_rate": 9.614700652045051e-06,
+      "loss": 0.2718,
+      "step": 2630
+    },
+    {
+      "epoch": 1.5591111111111111,
+      "grad_norm": 1.6491354703903198,
+      "learning_rate": 9.610748863860899e-06,
+      "loss": 0.2074,
+      "step": 2631
+    },
+    {
+      "epoch": 1.5597037037037036,
+      "grad_norm": 2.5144665241241455,
+      "learning_rate": 9.606797075676745e-06,
+      "loss": 0.2923,
+      "step": 2632
+    },
+    {
+      "epoch": 1.5602962962962963,
+      "grad_norm": 2.188006639480591,
+      "learning_rate": 9.602845287492591e-06,
+      "loss": 0.2907,
+      "step": 2633
+    },
+    {
+      "epoch": 1.560888888888889,
+      "grad_norm": 1.8710626363754272,
+      "learning_rate": 9.598893499308438e-06,
+      "loss": 0.2028,
+      "step": 2634
+    },
+    {
+      "epoch": 1.5614814814814815,
+      "grad_norm": 1.9143848419189453,
+      "learning_rate": 9.594941711124286e-06,
+      "loss": 0.2597,
+      "step": 2635
+    },
+    {
+      "epoch": 1.562074074074074,
+      "grad_norm": 2.374401807785034,
+      "learning_rate": 9.590989922940132e-06,
+      "loss": 0.3652,
+      "step": 2636
+    },
+    {
+      "epoch": 1.5626666666666666,
+      "grad_norm": 2.1201283931732178,
+      "learning_rate": 9.587038134755978e-06,
+      "loss": 0.2968,
+      "step": 2637
+    },
+    {
+      "epoch": 1.5632592592592593,
+      "grad_norm": 3.1292943954467773,
+      "learning_rate": 9.583086346571824e-06,
+      "loss": 0.3283,
+      "step": 2638
+    },
+    {
+      "epoch": 1.5638518518518518,
+      "grad_norm": 2.051957368850708,
+      "learning_rate": 9.579134558387672e-06,
+      "loss": 0.1955,
+      "step": 2639
+    },
+    {
+      "epoch": 1.5644444444444443,
+      "grad_norm": 1.6408557891845703,
+      "learning_rate": 9.575182770203519e-06,
+      "loss": 0.2246,
+      "step": 2640
+    },
+    {
+      "epoch": 1.565037037037037,
+      "grad_norm": 1.7697851657867432,
+      "learning_rate": 9.571230982019365e-06,
+      "loss": 0.2954,
+      "step": 2641
+    },
+    {
+      "epoch": 1.5656296296296297,
+      "grad_norm": 2.3257150650024414,
+      "learning_rate": 9.567279193835211e-06,
+      "loss": 0.2938,
+      "step": 2642
+    },
+    {
+      "epoch": 1.5662222222222222,
+      "grad_norm": 1.8210402727127075,
+      "learning_rate": 9.563327405651059e-06,
+      "loss": 0.2475,
+      "step": 2643
+    },
+    {
+      "epoch": 1.5668148148148147,
+      "grad_norm": 1.8313543796539307,
+      "learning_rate": 9.559375617466905e-06,
+      "loss": 0.2412,
+      "step": 2644
+    },
+    {
+      "epoch": 1.5674074074074074,
+      "grad_norm": 2.1651611328125,
+      "learning_rate": 9.555423829282751e-06,
+      "loss": 0.1577,
+      "step": 2645
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 1.9500426054000854,
+      "learning_rate": 9.551472041098598e-06,
+      "loss": 0.2612,
+      "step": 2646
+    },
+    {
+      "epoch": 1.5685925925925925,
+      "grad_norm": 2.2464005947113037,
+      "learning_rate": 9.547520252914446e-06,
+      "loss": 0.2351,
+      "step": 2647
+    },
+    {
+      "epoch": 1.5691851851851852,
+      "grad_norm": 1.6943529844284058,
+      "learning_rate": 9.543568464730292e-06,
+      "loss": 0.2522,
+      "step": 2648
+    },
+    {
+      "epoch": 1.569777777777778,
+      "grad_norm": 1.9848434925079346,
+      "learning_rate": 9.539616676546138e-06,
+      "loss": 0.3052,
+      "step": 2649
+    },
+    {
+      "epoch": 1.5703703703703704,
+      "grad_norm": 2.047006368637085,
+      "learning_rate": 9.535664888361984e-06,
+      "loss": 0.3169,
+      "step": 2650
+    },
+    {
+      "epoch": 1.570962962962963,
+      "grad_norm": 3.248735189437866,
+      "learning_rate": 9.531713100177832e-06,
+      "loss": 0.3233,
+      "step": 2651
+    },
+    {
+      "epoch": 1.5715555555555556,
+      "grad_norm": 3.1153500080108643,
+      "learning_rate": 9.527761311993677e-06,
+      "loss": 0.2902,
+      "step": 2652
+    },
+    {
+      "epoch": 1.5721481481481483,
+      "grad_norm": 3.707437038421631,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.4399,
+      "step": 2653
+    },
+    {
+      "epoch": 1.5727407407407408,
+      "grad_norm": 1.9192004203796387,
+      "learning_rate": 9.519857735625371e-06,
+      "loss": 0.3564,
+      "step": 2654
+    },
+    {
+      "epoch": 1.5733333333333333,
+      "grad_norm": 1.7692315578460693,
+      "learning_rate": 9.515905947441217e-06,
+      "loss": 0.225,
+      "step": 2655
+    },
+    {
+      "epoch": 1.573925925925926,
+      "grad_norm": 1.9881722927093506,
+      "learning_rate": 9.511954159257063e-06,
+      "loss": 0.2497,
+      "step": 2656
+    },
+    {
+      "epoch": 1.5745185185185187,
+      "grad_norm": 1.7664867639541626,
+      "learning_rate": 9.508002371072911e-06,
+      "loss": 0.3097,
+      "step": 2657
+    },
+    {
+      "epoch": 1.5751111111111111,
+      "grad_norm": 2.4972119331359863,
+      "learning_rate": 9.504050582888758e-06,
+      "loss": 0.2366,
+      "step": 2658
+    },
+    {
+      "epoch": 1.5757037037037036,
+      "grad_norm": 2.293471336364746,
+      "learning_rate": 9.500098794704604e-06,
+      "loss": 0.2415,
+      "step": 2659
+    },
+    {
+      "epoch": 1.5762962962962963,
+      "grad_norm": 2.8387184143066406,
+      "learning_rate": 9.49614700652045e-06,
+      "loss": 0.2818,
+      "step": 2660
+    },
+    {
+      "epoch": 1.576888888888889,
+      "grad_norm": 2.396261215209961,
+      "learning_rate": 9.492195218336298e-06,
+      "loss": 0.3157,
+      "step": 2661
+    },
+    {
+      "epoch": 1.5774814814814815,
+      "grad_norm": 1.4193440675735474,
+      "learning_rate": 9.488243430152144e-06,
+      "loss": 0.2118,
+      "step": 2662
+    },
+    {
+      "epoch": 1.578074074074074,
+      "grad_norm": 1.8652853965759277,
+      "learning_rate": 9.48429164196799e-06,
+      "loss": 0.2656,
+      "step": 2663
+    },
+    {
+      "epoch": 1.5786666666666667,
+      "grad_norm": 2.355194091796875,
+      "learning_rate": 9.480339853783837e-06,
+      "loss": 0.2389,
+      "step": 2664
+    },
+    {
+      "epoch": 1.5792592592592594,
+      "grad_norm": 2.046435594558716,
+      "learning_rate": 9.476388065599685e-06,
+      "loss": 0.168,
+      "step": 2665
+    },
+    {
+      "epoch": 1.5798518518518518,
+      "grad_norm": 2.8499550819396973,
+      "learning_rate": 9.472436277415531e-06,
+      "loss": 0.3122,
+      "step": 2666
+    },
+    {
+      "epoch": 1.5804444444444443,
+      "grad_norm": 1.6256639957427979,
+      "learning_rate": 9.468484489231377e-06,
+      "loss": 0.2586,
+      "step": 2667
+    },
+    {
+      "epoch": 1.581037037037037,
+      "grad_norm": 2.3510031700134277,
+      "learning_rate": 9.464532701047225e-06,
+      "loss": 0.3077,
+      "step": 2668
+    },
+    {
+      "epoch": 1.5816296296296297,
+      "grad_norm": 2.5695157051086426,
+      "learning_rate": 9.460580912863071e-06,
+      "loss": 0.404,
+      "step": 2669
+    },
+    {
+      "epoch": 1.5822222222222222,
+      "grad_norm": 2.0862252712249756,
+      "learning_rate": 9.456629124678918e-06,
+      "loss": 0.2525,
+      "step": 2670
+    },
+    {
+      "epoch": 1.5828148148148147,
+      "grad_norm": 2.889094352722168,
+      "learning_rate": 9.452677336494764e-06,
+      "loss": 0.4816,
+      "step": 2671
+    },
+    {
+      "epoch": 1.5834074074074074,
+      "grad_norm": 2.302417039871216,
+      "learning_rate": 9.448725548310612e-06,
+      "loss": 0.3256,
+      "step": 2672
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 2.7622551918029785,
+      "learning_rate": 9.444773760126458e-06,
+      "loss": 0.2875,
+      "step": 2673
+    },
+    {
+      "epoch": 1.5845925925925926,
+      "grad_norm": 2.2517290115356445,
+      "learning_rate": 9.440821971942304e-06,
+      "loss": 0.2212,
+      "step": 2674
+    },
+    {
+      "epoch": 1.585185185185185,
+      "grad_norm": 2.8492624759674072,
+      "learning_rate": 9.43687018375815e-06,
+      "loss": 0.416,
+      "step": 2675
+    },
+    {
+      "epoch": 1.5857777777777777,
+      "grad_norm": 2.7426257133483887,
+      "learning_rate": 9.432918395573998e-06,
+      "loss": 0.2111,
+      "step": 2676
+    },
+    {
+      "epoch": 1.5863703703703704,
+      "grad_norm": 2.5752298831939697,
+      "learning_rate": 9.428966607389845e-06,
+      "loss": 0.255,
+      "step": 2677
+    },
+    {
+      "epoch": 1.586962962962963,
+      "grad_norm": 1.8272169828414917,
+      "learning_rate": 9.425014819205691e-06,
+      "loss": 0.221,
+      "step": 2678
+    },
+    {
+      "epoch": 1.5875555555555556,
+      "grad_norm": 2.17390513420105,
+      "learning_rate": 9.421063031021537e-06,
+      "loss": 0.2199,
+      "step": 2679
+    },
+    {
+      "epoch": 1.5881481481481483,
+      "grad_norm": 4.5538010597229,
+      "learning_rate": 9.417111242837385e-06,
+      "loss": 0.3441,
+      "step": 2680
+    },
+    {
+      "epoch": 1.5887407407407408,
+      "grad_norm": 1.6002103090286255,
+      "learning_rate": 9.413159454653231e-06,
+      "loss": 0.251,
+      "step": 2681
+    },
+    {
+      "epoch": 1.5893333333333333,
+      "grad_norm": 2.467226028442383,
+      "learning_rate": 9.409207666469078e-06,
+      "loss": 0.2692,
+      "step": 2682
+    },
+    {
+      "epoch": 1.589925925925926,
+      "grad_norm": 2.2987334728240967,
+      "learning_rate": 9.405255878284924e-06,
+      "loss": 0.239,
+      "step": 2683
+    },
+    {
+      "epoch": 1.5905185185185187,
+      "grad_norm": 2.1408519744873047,
+      "learning_rate": 9.401304090100772e-06,
+      "loss": 0.4375,
+      "step": 2684
+    },
+    {
+      "epoch": 1.5911111111111111,
+      "grad_norm": 2.185520648956299,
+      "learning_rate": 9.397352301916618e-06,
+      "loss": 0.3,
+      "step": 2685
+    },
+    {
+      "epoch": 1.5917037037037036,
+      "grad_norm": 2.3359196186065674,
+      "learning_rate": 9.393400513732464e-06,
+      "loss": 0.3713,
+      "step": 2686
+    },
+    {
+      "epoch": 1.5922962962962963,
+      "grad_norm": 1.7439159154891968,
+      "learning_rate": 9.38944872554831e-06,
+      "loss": 0.3668,
+      "step": 2687
+    },
+    {
+      "epoch": 1.592888888888889,
+      "grad_norm": 1.597812294960022,
+      "learning_rate": 9.385496937364158e-06,
+      "loss": 0.2533,
+      "step": 2688
+    },
+    {
+      "epoch": 1.5934814814814815,
+      "grad_norm": 1.7350304126739502,
+      "learning_rate": 9.381545149180005e-06,
+      "loss": 0.2264,
+      "step": 2689
+    },
+    {
+      "epoch": 1.594074074074074,
+      "grad_norm": 1.8854881525039673,
+      "learning_rate": 9.377593360995851e-06,
+      "loss": 0.2707,
+      "step": 2690
+    },
+    {
+      "epoch": 1.5946666666666667,
+      "grad_norm": 1.8703038692474365,
+      "learning_rate": 9.373641572811697e-06,
+      "loss": 0.3092,
+      "step": 2691
+    },
+    {
+      "epoch": 1.5952592592592594,
+      "grad_norm": 2.329282522201538,
+      "learning_rate": 9.369689784627545e-06,
+      "loss": 0.3897,
+      "step": 2692
+    },
+    {
+      "epoch": 1.5958518518518519,
+      "grad_norm": 1.6264687776565552,
+      "learning_rate": 9.365737996443391e-06,
+      "loss": 0.2216,
+      "step": 2693
+    },
+    {
+      "epoch": 1.5964444444444443,
+      "grad_norm": 1.9029139280319214,
+      "learning_rate": 9.361786208259238e-06,
+      "loss": 0.273,
+      "step": 2694
+    },
+    {
+      "epoch": 1.597037037037037,
+      "grad_norm": 1.7720500230789185,
+      "learning_rate": 9.357834420075086e-06,
+      "loss": 0.2995,
+      "step": 2695
+    },
+    {
+      "epoch": 1.5976296296296297,
+      "grad_norm": 2.015205144882202,
+      "learning_rate": 9.353882631890932e-06,
+      "loss": 0.2155,
+      "step": 2696
+    },
+    {
+      "epoch": 1.5982222222222222,
+      "grad_norm": 1.725473165512085,
+      "learning_rate": 9.349930843706778e-06,
+      "loss": 0.34,
+      "step": 2697
+    },
+    {
+      "epoch": 1.5988148148148147,
+      "grad_norm": 1.5868672132492065,
+      "learning_rate": 9.345979055522624e-06,
+      "loss": 0.3233,
+      "step": 2698
+    },
+    {
+      "epoch": 1.5994074074074074,
+      "grad_norm": 3.045706033706665,
+      "learning_rate": 9.342027267338472e-06,
+      "loss": 0.3605,
+      "step": 2699
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.510543942451477,
+      "learning_rate": 9.338075479154318e-06,
+      "loss": 0.1862,
+      "step": 2700
+    },
+    {
+      "epoch": 1.6005925925925926,
+      "grad_norm": 2.4151723384857178,
+      "learning_rate": 9.334123690970165e-06,
+      "loss": 0.4155,
+      "step": 2701
+    },
+    {
+      "epoch": 1.601185185185185,
+      "grad_norm": 2.116975784301758,
+      "learning_rate": 9.330171902786011e-06,
+      "loss": 0.3158,
+      "step": 2702
+    },
+    {
+      "epoch": 1.6017777777777777,
+      "grad_norm": 2.0077459812164307,
+      "learning_rate": 9.326220114601859e-06,
+      "loss": 0.1834,
+      "step": 2703
+    },
+    {
+      "epoch": 1.6023703703703704,
+      "grad_norm": 2.080587863922119,
+      "learning_rate": 9.322268326417705e-06,
+      "loss": 0.3201,
+      "step": 2704
+    },
+    {
+      "epoch": 1.602962962962963,
+      "grad_norm": 2.3796770572662354,
+      "learning_rate": 9.318316538233551e-06,
+      "loss": 0.3068,
+      "step": 2705
+    },
+    {
+      "epoch": 1.6035555555555554,
+      "grad_norm": 2.2364022731781006,
+      "learning_rate": 9.314364750049398e-06,
+      "loss": 0.3405,
+      "step": 2706
+    },
+    {
+      "epoch": 1.604148148148148,
+      "grad_norm": 2.5778064727783203,
+      "learning_rate": 9.310412961865246e-06,
+      "loss": 0.2494,
+      "step": 2707
+    },
+    {
+      "epoch": 1.6047407407407408,
+      "grad_norm": 2.4267241954803467,
+      "learning_rate": 9.306461173681092e-06,
+      "loss": 0.2335,
+      "step": 2708
+    },
+    {
+      "epoch": 1.6053333333333333,
+      "grad_norm": 1.8919293880462646,
+      "learning_rate": 9.302509385496938e-06,
+      "loss": 0.289,
+      "step": 2709
+    },
+    {
+      "epoch": 1.605925925925926,
+      "grad_norm": 2.256136178970337,
+      "learning_rate": 9.298557597312784e-06,
+      "loss": 0.2394,
+      "step": 2710
+    },
+    {
+      "epoch": 1.6065185185185187,
+      "grad_norm": 2.1327717304229736,
+      "learning_rate": 9.294605809128632e-06,
+      "loss": 0.2695,
+      "step": 2711
+    },
+    {
+      "epoch": 1.6071111111111112,
+      "grad_norm": 1.6749356985092163,
+      "learning_rate": 9.290654020944478e-06,
+      "loss": 0.2475,
+      "step": 2712
+    },
+    {
+      "epoch": 1.6077037037037036,
+      "grad_norm": 1.8579556941986084,
+      "learning_rate": 9.286702232760325e-06,
+      "loss": 0.2343,
+      "step": 2713
+    },
+    {
+      "epoch": 1.6082962962962963,
+      "grad_norm": 2.0034310817718506,
+      "learning_rate": 9.282750444576171e-06,
+      "loss": 0.3798,
+      "step": 2714
+    },
+    {
+      "epoch": 1.608888888888889,
+      "grad_norm": 1.7144393920898438,
+      "learning_rate": 9.278798656392019e-06,
+      "loss": 0.2674,
+      "step": 2715
+    },
+    {
+      "epoch": 1.6094814814814815,
+      "grad_norm": 1.8259952068328857,
+      "learning_rate": 9.274846868207865e-06,
+      "loss": 0.3056,
+      "step": 2716
+    },
+    {
+      "epoch": 1.610074074074074,
+      "grad_norm": 2.114086866378784,
+      "learning_rate": 9.270895080023711e-06,
+      "loss": 0.2462,
+      "step": 2717
+    },
+    {
+      "epoch": 1.6106666666666667,
+      "grad_norm": 1.769028663635254,
+      "learning_rate": 9.26694329183956e-06,
+      "loss": 0.2156,
+      "step": 2718
+    },
+    {
+      "epoch": 1.6112592592592594,
+      "grad_norm": 2.1561636924743652,
+      "learning_rate": 9.262991503655406e-06,
+      "loss": 0.3056,
+      "step": 2719
+    },
+    {
+      "epoch": 1.6118518518518519,
+      "grad_norm": 1.5451669692993164,
+      "learning_rate": 9.259039715471252e-06,
+      "loss": 0.2649,
+      "step": 2720
+    },
+    {
+      "epoch": 1.6124444444444443,
+      "grad_norm": 1.4852455854415894,
+      "learning_rate": 9.255087927287098e-06,
+      "loss": 0.2074,
+      "step": 2721
+    },
+    {
+      "epoch": 1.613037037037037,
+      "grad_norm": 1.8847637176513672,
+      "learning_rate": 9.251136139102946e-06,
+      "loss": 0.2611,
+      "step": 2722
+    },
+    {
+      "epoch": 1.6136296296296297,
+      "grad_norm": 3.128976583480835,
+      "learning_rate": 9.247184350918792e-06,
+      "loss": 0.3008,
+      "step": 2723
+    },
+    {
+      "epoch": 1.6142222222222222,
+      "grad_norm": 3.4000329971313477,
+      "learning_rate": 9.243232562734638e-06,
+      "loss": 0.5107,
+      "step": 2724
+    },
+    {
+      "epoch": 1.6148148148148147,
+      "grad_norm": 1.9700267314910889,
+      "learning_rate": 9.239280774550485e-06,
+      "loss": 0.3186,
+      "step": 2725
+    },
+    {
+      "epoch": 1.6154074074074074,
+      "grad_norm": 1.7896177768707275,
+      "learning_rate": 9.235328986366333e-06,
+      "loss": 0.1878,
+      "step": 2726
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 2.8685576915740967,
+      "learning_rate": 9.231377198182179e-06,
+      "loss": 0.3007,
+      "step": 2727
+    },
+    {
+      "epoch": 1.6165925925925926,
+      "grad_norm": 1.7246830463409424,
+      "learning_rate": 9.227425409998025e-06,
+      "loss": 0.2364,
+      "step": 2728
+    },
+    {
+      "epoch": 1.617185185185185,
+      "grad_norm": 1.5073856115341187,
+      "learning_rate": 9.223473621813871e-06,
+      "loss": 0.2171,
+      "step": 2729
+    },
+    {
+      "epoch": 1.6177777777777778,
+      "grad_norm": 1.9667011499404907,
+      "learning_rate": 9.21952183362972e-06,
+      "loss": 0.2437,
+      "step": 2730
+    },
+    {
+      "epoch": 1.6183703703703705,
+      "grad_norm": 2.6252083778381348,
+      "learning_rate": 9.215570045445564e-06,
+      "loss": 0.3419,
+      "step": 2731
+    },
+    {
+      "epoch": 1.618962962962963,
+      "grad_norm": 2.018165111541748,
+      "learning_rate": 9.211618257261412e-06,
+      "loss": 0.2351,
+      "step": 2732
+    },
+    {
+      "epoch": 1.6195555555555554,
+      "grad_norm": 2.3121752738952637,
+      "learning_rate": 9.207666469077258e-06,
+      "loss": 0.3058,
+      "step": 2733
+    },
+    {
+      "epoch": 1.6201481481481481,
+      "grad_norm": 1.680524468421936,
+      "learning_rate": 9.203714680893104e-06,
+      "loss": 0.2465,
+      "step": 2734
+    },
+    {
+      "epoch": 1.6207407407407408,
+      "grad_norm": 1.922727346420288,
+      "learning_rate": 9.19976289270895e-06,
+      "loss": 0.2814,
+      "step": 2735
+    },
+    {
+      "epoch": 1.6213333333333333,
+      "grad_norm": 2.6292102336883545,
+      "learning_rate": 9.195811104524798e-06,
+      "loss": 0.3424,
+      "step": 2736
+    },
+    {
+      "epoch": 1.6219259259259258,
+      "grad_norm": 2.798360824584961,
+      "learning_rate": 9.191859316340645e-06,
+      "loss": 0.3748,
+      "step": 2737
+    },
+    {
+      "epoch": 1.6225185185185185,
+      "grad_norm": 2.002873420715332,
+      "learning_rate": 9.187907528156491e-06,
+      "loss": 0.2738,
+      "step": 2738
+    },
+    {
+      "epoch": 1.6231111111111112,
+      "grad_norm": 1.488196849822998,
+      "learning_rate": 9.183955739972337e-06,
+      "loss": 0.2188,
+      "step": 2739
+    },
+    {
+      "epoch": 1.6237037037037036,
+      "grad_norm": 2.3493969440460205,
+      "learning_rate": 9.180003951788185e-06,
+      "loss": 0.2532,
+      "step": 2740
+    },
+    {
+      "epoch": 1.6242962962962963,
+      "grad_norm": 2.6168313026428223,
+      "learning_rate": 9.176052163604031e-06,
+      "loss": 0.2558,
+      "step": 2741
+    },
+    {
+      "epoch": 1.624888888888889,
+      "grad_norm": 2.6312050819396973,
+      "learning_rate": 9.172100375419878e-06,
+      "loss": 0.3138,
+      "step": 2742
+    },
+    {
+      "epoch": 1.6254814814814815,
+      "grad_norm": 2.0851619243621826,
+      "learning_rate": 9.168148587235724e-06,
+      "loss": 0.2612,
+      "step": 2743
+    },
+    {
+      "epoch": 1.626074074074074,
+      "grad_norm": 2.1912548542022705,
+      "learning_rate": 9.164196799051572e-06,
+      "loss": 0.3579,
+      "step": 2744
+    },
+    {
+      "epoch": 1.6266666666666667,
+      "grad_norm": 1.6997731924057007,
+      "learning_rate": 9.160245010867418e-06,
+      "loss": 0.2791,
+      "step": 2745
+    },
+    {
+      "epoch": 1.6272592592592594,
+      "grad_norm": 1.8256407976150513,
+      "learning_rate": 9.156293222683264e-06,
+      "loss": 0.2331,
+      "step": 2746
+    },
+    {
+      "epoch": 1.6278518518518519,
+      "grad_norm": 1.733689308166504,
+      "learning_rate": 9.15234143449911e-06,
+      "loss": 0.2941,
+      "step": 2747
+    },
+    {
+      "epoch": 1.6284444444444444,
+      "grad_norm": 2.1536190509796143,
+      "learning_rate": 9.148389646314958e-06,
+      "loss": 0.2636,
+      "step": 2748
+    },
+    {
+      "epoch": 1.629037037037037,
+      "grad_norm": 1.868902325630188,
+      "learning_rate": 9.144437858130805e-06,
+      "loss": 0.2639,
+      "step": 2749
+    },
+    {
+      "epoch": 1.6296296296296298,
+      "grad_norm": 2.6236228942871094,
+      "learning_rate": 9.140486069946651e-06,
+      "loss": 0.194,
+      "step": 2750
+    },
+    {
+      "epoch": 1.6302222222222222,
+      "grad_norm": 1.9091511964797974,
+      "learning_rate": 9.136534281762497e-06,
+      "loss": 0.2583,
+      "step": 2751
+    },
+    {
+      "epoch": 1.6308148148148147,
+      "grad_norm": 2.8464579582214355,
+      "learning_rate": 9.132582493578345e-06,
+      "loss": 0.2515,
+      "step": 2752
+    },
+    {
+      "epoch": 1.6314074074074074,
+      "grad_norm": 2.4935619831085205,
+      "learning_rate": 9.128630705394191e-06,
+      "loss": 0.3707,
+      "step": 2753
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 2.4409542083740234,
+      "learning_rate": 9.124678917210038e-06,
+      "loss": 0.2822,
+      "step": 2754
+    },
+    {
+      "epoch": 1.6325925925925926,
+      "grad_norm": 2.1792993545532227,
+      "learning_rate": 9.120727129025886e-06,
+      "loss": 0.3119,
+      "step": 2755
+    },
+    {
+      "epoch": 1.633185185185185,
+      "grad_norm": 2.088217258453369,
+      "learning_rate": 9.116775340841732e-06,
+      "loss": 0.2949,
+      "step": 2756
+    },
+    {
+      "epoch": 1.6337777777777778,
+      "grad_norm": 1.9866514205932617,
+      "learning_rate": 9.112823552657578e-06,
+      "loss": 0.2866,
+      "step": 2757
+    },
+    {
+      "epoch": 1.6343703703703705,
+      "grad_norm": 2.8304965496063232,
+      "learning_rate": 9.108871764473424e-06,
+      "loss": 0.3199,
+      "step": 2758
+    },
+    {
+      "epoch": 1.634962962962963,
+      "grad_norm": 1.8003642559051514,
+      "learning_rate": 9.104919976289272e-06,
+      "loss": 0.2371,
+      "step": 2759
+    },
+    {
+      "epoch": 1.6355555555555554,
+      "grad_norm": 2.26478910446167,
+      "learning_rate": 9.100968188105118e-06,
+      "loss": 0.3024,
+      "step": 2760
+    },
+    {
+      "epoch": 1.6361481481481481,
+      "grad_norm": 2.17110276222229,
+      "learning_rate": 9.097016399920965e-06,
+      "loss": 0.2829,
+      "step": 2761
+    },
+    {
+      "epoch": 1.6367407407407408,
+      "grad_norm": 1.6669994592666626,
+      "learning_rate": 9.093064611736811e-06,
+      "loss": 0.218,
+      "step": 2762
+    },
+    {
+      "epoch": 1.6373333333333333,
+      "grad_norm": 1.3842778205871582,
+      "learning_rate": 9.089112823552659e-06,
+      "loss": 0.2121,
+      "step": 2763
+    },
+    {
+      "epoch": 1.6379259259259258,
+      "grad_norm": 2.311875343322754,
+      "learning_rate": 9.085161035368505e-06,
+      "loss": 0.2662,
+      "step": 2764
+    },
+    {
+      "epoch": 1.6385185185185185,
+      "grad_norm": 1.9529178142547607,
+      "learning_rate": 9.081209247184351e-06,
+      "loss": 0.2635,
+      "step": 2765
+    },
+    {
+      "epoch": 1.6391111111111112,
+      "grad_norm": 1.8715184926986694,
+      "learning_rate": 9.077257459000198e-06,
+      "loss": 0.2345,
+      "step": 2766
+    },
+    {
+      "epoch": 1.6397037037037037,
+      "grad_norm": 2.564044237136841,
+      "learning_rate": 9.073305670816046e-06,
+      "loss": 0.2554,
+      "step": 2767
+    },
+    {
+      "epoch": 1.6402962962962961,
+      "grad_norm": 2.0011000633239746,
+      "learning_rate": 9.069353882631892e-06,
+      "loss": 0.2688,
+      "step": 2768
+    },
+    {
+      "epoch": 1.6408888888888888,
+      "grad_norm": 2.1106300354003906,
+      "learning_rate": 9.065402094447738e-06,
+      "loss": 0.2409,
+      "step": 2769
+    },
+    {
+      "epoch": 1.6414814814814815,
+      "grad_norm": 1.9562993049621582,
+      "learning_rate": 9.061450306263584e-06,
+      "loss": 0.3446,
+      "step": 2770
+    },
+    {
+      "epoch": 1.642074074074074,
+      "grad_norm": 1.9370026588439941,
+      "learning_rate": 9.057498518079432e-06,
+      "loss": 0.2658,
+      "step": 2771
+    },
+    {
+      "epoch": 1.6426666666666667,
+      "grad_norm": 2.853555679321289,
+      "learning_rate": 9.053546729895278e-06,
+      "loss": 0.3142,
+      "step": 2772
+    },
+    {
+      "epoch": 1.6432592592592594,
+      "grad_norm": 2.479185104370117,
+      "learning_rate": 9.049594941711125e-06,
+      "loss": 0.3293,
+      "step": 2773
+    },
+    {
+      "epoch": 1.643851851851852,
+      "grad_norm": 1.8049688339233398,
+      "learning_rate": 9.045643153526971e-06,
+      "loss": 0.1815,
+      "step": 2774
+    },
+    {
+      "epoch": 1.6444444444444444,
+      "grad_norm": 1.5344862937927246,
+      "learning_rate": 9.041691365342819e-06,
+      "loss": 0.1936,
+      "step": 2775
+    },
+    {
+      "epoch": 1.645037037037037,
+      "grad_norm": 3.5519931316375732,
+      "learning_rate": 9.037739577158665e-06,
+      "loss": 0.4549,
+      "step": 2776
+    },
+    {
+      "epoch": 1.6456296296296298,
+      "grad_norm": 1.8131327629089355,
+      "learning_rate": 9.033787788974511e-06,
+      "loss": 0.2412,
+      "step": 2777
+    },
+    {
+      "epoch": 1.6462222222222223,
+      "grad_norm": 2.1116180419921875,
+      "learning_rate": 9.02983600079036e-06,
+      "loss": 0.2715,
+      "step": 2778
+    },
+    {
+      "epoch": 1.6468148148148147,
+      "grad_norm": 3.540597438812256,
+      "learning_rate": 9.025884212606206e-06,
+      "loss": 0.3223,
+      "step": 2779
+    },
+    {
+      "epoch": 1.6474074074074074,
+      "grad_norm": 2.0901296138763428,
+      "learning_rate": 9.021932424422052e-06,
+      "loss": 0.2132,
+      "step": 2780
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 1.994560956954956,
+      "learning_rate": 9.017980636237898e-06,
+      "loss": 0.2477,
+      "step": 2781
+    },
+    {
+      "epoch": 1.6485925925925926,
+      "grad_norm": 1.9730842113494873,
+      "learning_rate": 9.014028848053746e-06,
+      "loss": 0.2817,
+      "step": 2782
+    },
+    {
+      "epoch": 1.649185185185185,
+      "grad_norm": 1.853010654449463,
+      "learning_rate": 9.010077059869592e-06,
+      "loss": 0.2104,
+      "step": 2783
+    },
+    {
+      "epoch": 1.6497777777777778,
+      "grad_norm": 2.406782388687134,
+      "learning_rate": 9.006125271685438e-06,
+      "loss": 0.3178,
+      "step": 2784
+    },
+    {
+      "epoch": 1.6503703703703705,
+      "grad_norm": 1.95796799659729,
+      "learning_rate": 9.002173483501285e-06,
+      "loss": 0.2889,
+      "step": 2785
+    },
+    {
+      "epoch": 1.650962962962963,
+      "grad_norm": 1.7796316146850586,
+      "learning_rate": 8.998221695317133e-06,
+      "loss": 0.3037,
+      "step": 2786
+    },
+    {
+      "epoch": 1.6515555555555554,
+      "grad_norm": 1.7297028303146362,
+      "learning_rate": 8.994269907132979e-06,
+      "loss": 0.2208,
+      "step": 2787
+    },
+    {
+      "epoch": 1.6521481481481481,
+      "grad_norm": 1.918816328048706,
+      "learning_rate": 8.990318118948825e-06,
+      "loss": 0.2239,
+      "step": 2788
+    },
+    {
+      "epoch": 1.6527407407407408,
+      "grad_norm": 2.2494218349456787,
+      "learning_rate": 8.986366330764671e-06,
+      "loss": 0.4433,
+      "step": 2789
+    },
+    {
+      "epoch": 1.6533333333333333,
+      "grad_norm": 2.181358814239502,
+      "learning_rate": 8.98241454258052e-06,
+      "loss": 0.182,
+      "step": 2790
+    },
+    {
+      "epoch": 1.6539259259259258,
+      "grad_norm": 1.385337471961975,
+      "learning_rate": 8.978462754396366e-06,
+      "loss": 0.1567,
+      "step": 2791
+    },
+    {
+      "epoch": 1.6545185185185185,
+      "grad_norm": 1.8122811317443848,
+      "learning_rate": 8.974510966212212e-06,
+      "loss": 0.2624,
+      "step": 2792
+    },
+    {
+      "epoch": 1.6551111111111112,
+      "grad_norm": 1.625289797782898,
+      "learning_rate": 8.970559178028058e-06,
+      "loss": 0.2221,
+      "step": 2793
+    },
+    {
+      "epoch": 1.6557037037037037,
+      "grad_norm": 1.9137665033340454,
+      "learning_rate": 8.966607389843906e-06,
+      "loss": 0.1759,
+      "step": 2794
+    },
+    {
+      "epoch": 1.6562962962962962,
+      "grad_norm": 2.5283968448638916,
+      "learning_rate": 8.962655601659752e-06,
+      "loss": 0.3014,
+      "step": 2795
+    },
+    {
+      "epoch": 1.6568888888888889,
+      "grad_norm": 2.5585131645202637,
+      "learning_rate": 8.958703813475598e-06,
+      "loss": 0.3573,
+      "step": 2796
+    },
+    {
+      "epoch": 1.6574814814814816,
+      "grad_norm": 2.2342865467071533,
+      "learning_rate": 8.954752025291445e-06,
+      "loss": 0.2655,
+      "step": 2797
+    },
+    {
+      "epoch": 1.658074074074074,
+      "grad_norm": 1.8365260362625122,
+      "learning_rate": 8.950800237107293e-06,
+      "loss": 0.3088,
+      "step": 2798
+    },
+    {
+      "epoch": 1.6586666666666665,
+      "grad_norm": 1.7509589195251465,
+      "learning_rate": 8.946848448923139e-06,
+      "loss": 0.2617,
+      "step": 2799
+    },
+    {
+      "epoch": 1.6592592592592592,
+      "grad_norm": 2.1518285274505615,
+      "learning_rate": 8.942896660738985e-06,
+      "loss": 0.2395,
+      "step": 2800
+    },
+    {
+      "epoch": 1.659851851851852,
+      "grad_norm": 3.1609017848968506,
+      "learning_rate": 8.938944872554833e-06,
+      "loss": 0.3714,
+      "step": 2801
+    },
+    {
+      "epoch": 1.6604444444444444,
+      "grad_norm": 1.862982988357544,
+      "learning_rate": 8.93499308437068e-06,
+      "loss": 0.2346,
+      "step": 2802
+    },
+    {
+      "epoch": 1.661037037037037,
+      "grad_norm": 2.4717488288879395,
+      "learning_rate": 8.931041296186526e-06,
+      "loss": 0.3862,
+      "step": 2803
+    },
+    {
+      "epoch": 1.6616296296296298,
+      "grad_norm": 1.7801570892333984,
+      "learning_rate": 8.927089508002372e-06,
+      "loss": 0.2076,
+      "step": 2804
+    },
+    {
+      "epoch": 1.6622222222222223,
+      "grad_norm": 1.6935722827911377,
+      "learning_rate": 8.92313771981822e-06,
+      "loss": 0.217,
+      "step": 2805
+    },
+    {
+      "epoch": 1.6628148148148147,
+      "grad_norm": 1.9519463777542114,
+      "learning_rate": 8.919185931634066e-06,
+      "loss": 0.2233,
+      "step": 2806
+    },
+    {
+      "epoch": 1.6634074074074074,
+      "grad_norm": 1.5704416036605835,
+      "learning_rate": 8.915234143449912e-06,
+      "loss": 0.2032,
+      "step": 2807
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 3.1859982013702393,
+      "learning_rate": 8.911282355265758e-06,
+      "loss": 0.2623,
+      "step": 2808
+    },
+    {
+      "epoch": 1.6645925925925926,
+      "grad_norm": 2.5697007179260254,
+      "learning_rate": 8.907330567081606e-06,
+      "loss": 0.2665,
+      "step": 2809
+    },
+    {
+      "epoch": 1.665185185185185,
+      "grad_norm": 4.606372356414795,
+      "learning_rate": 8.903378778897451e-06,
+      "loss": 0.3937,
+      "step": 2810
+    },
+    {
+      "epoch": 1.6657777777777778,
+      "grad_norm": 2.4921252727508545,
+      "learning_rate": 8.899426990713297e-06,
+      "loss": 0.3604,
+      "step": 2811
+    },
+    {
+      "epoch": 1.6663703703703705,
+      "grad_norm": 1.8225489854812622,
+      "learning_rate": 8.895475202529145e-06,
+      "loss": 0.2159,
+      "step": 2812
+    },
+    {
+      "epoch": 1.666962962962963,
+      "grad_norm": 1.746298909187317,
+      "learning_rate": 8.891523414344991e-06,
+      "loss": 0.3018,
+      "step": 2813
+    },
+    {
+      "epoch": 1.6675555555555555,
+      "grad_norm": 1.607102394104004,
+      "learning_rate": 8.887571626160838e-06,
+      "loss": 0.287,
+      "step": 2814
+    },
+    {
+      "epoch": 1.6681481481481482,
+      "grad_norm": 2.5210325717926025,
+      "learning_rate": 8.883619837976686e-06,
+      "loss": 0.252,
+      "step": 2815
+    },
+    {
+      "epoch": 1.6687407407407409,
+      "grad_norm": 2.6342251300811768,
+      "learning_rate": 8.879668049792532e-06,
+      "loss": 0.2757,
+      "step": 2816
+    },
+    {
+      "epoch": 1.6693333333333333,
+      "grad_norm": 2.301706075668335,
+      "learning_rate": 8.875716261608378e-06,
+      "loss": 0.2931,
+      "step": 2817
+    },
+    {
+      "epoch": 1.6699259259259258,
+      "grad_norm": 1.6071490049362183,
+      "learning_rate": 8.871764473424224e-06,
+      "loss": 0.2541,
+      "step": 2818
+    },
+    {
+      "epoch": 1.6705185185185185,
+      "grad_norm": 1.9216601848602295,
+      "learning_rate": 8.867812685240072e-06,
+      "loss": 0.2676,
+      "step": 2819
+    },
+    {
+      "epoch": 1.6711111111111112,
+      "grad_norm": 3.1636924743652344,
+      "learning_rate": 8.863860897055918e-06,
+      "loss": 0.3353,
+      "step": 2820
+    },
+    {
+      "epoch": 1.6717037037037037,
+      "grad_norm": 2.190472364425659,
+      "learning_rate": 8.859909108871765e-06,
+      "loss": 0.3348,
+      "step": 2821
+    },
+    {
+      "epoch": 1.6722962962962962,
+      "grad_norm": 2.4025726318359375,
+      "learning_rate": 8.855957320687611e-06,
+      "loss": 0.2796,
+      "step": 2822
+    },
+    {
+      "epoch": 1.6728888888888889,
+      "grad_norm": 2.15029239654541,
+      "learning_rate": 8.852005532503459e-06,
+      "loss": 0.2733,
+      "step": 2823
+    },
+    {
+      "epoch": 1.6734814814814816,
+      "grad_norm": 1.8581126928329468,
+      "learning_rate": 8.848053744319305e-06,
+      "loss": 0.1846,
+      "step": 2824
+    },
+    {
+      "epoch": 1.674074074074074,
+      "grad_norm": 2.535572052001953,
+      "learning_rate": 8.844101956135151e-06,
+      "loss": 0.2149,
+      "step": 2825
+    },
+    {
+      "epoch": 1.6746666666666665,
+      "grad_norm": 2.1250803470611572,
+      "learning_rate": 8.840150167950998e-06,
+      "loss": 0.3084,
+      "step": 2826
+    },
+    {
+      "epoch": 1.6752592592592592,
+      "grad_norm": 2.25244402885437,
+      "learning_rate": 8.836198379766846e-06,
+      "loss": 0.331,
+      "step": 2827
+    },
+    {
+      "epoch": 1.675851851851852,
+      "grad_norm": 1.9986907243728638,
+      "learning_rate": 8.832246591582692e-06,
+      "loss": 0.2028,
+      "step": 2828
+    },
+    {
+      "epoch": 1.6764444444444444,
+      "grad_norm": 1.9361169338226318,
+      "learning_rate": 8.828294803398538e-06,
+      "loss": 0.2529,
+      "step": 2829
+    },
+    {
+      "epoch": 1.6770370370370369,
+      "grad_norm": 1.7154847383499146,
+      "learning_rate": 8.824343015214384e-06,
+      "loss": 0.2338,
+      "step": 2830
+    },
+    {
+      "epoch": 1.6776296296296296,
+      "grad_norm": 3.820802688598633,
+      "learning_rate": 8.820391227030232e-06,
+      "loss": 0.4046,
+      "step": 2831
+    },
+    {
+      "epoch": 1.6782222222222223,
+      "grad_norm": 1.904059886932373,
+      "learning_rate": 8.816439438846078e-06,
+      "loss": 0.2689,
+      "step": 2832
+    },
+    {
+      "epoch": 1.6788148148148148,
+      "grad_norm": 2.1077003479003906,
+      "learning_rate": 8.812487650661925e-06,
+      "loss": 0.336,
+      "step": 2833
+    },
+    {
+      "epoch": 1.6794074074074075,
+      "grad_norm": 1.7072348594665527,
+      "learning_rate": 8.808535862477771e-06,
+      "loss": 0.2319,
+      "step": 2834
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 1.7677385807037354,
+      "learning_rate": 8.804584074293619e-06,
+      "loss": 0.2166,
+      "step": 2835
+    },
+    {
+      "epoch": 1.6805925925925926,
+      "grad_norm": 2.062427282333374,
+      "learning_rate": 8.800632286109465e-06,
+      "loss": 0.3167,
+      "step": 2836
+    },
+    {
+      "epoch": 1.6811851851851851,
+      "grad_norm": 1.310509443283081,
+      "learning_rate": 8.796680497925311e-06,
+      "loss": 0.1488,
+      "step": 2837
+    },
+    {
+      "epoch": 1.6817777777777778,
+      "grad_norm": 2.4233663082122803,
+      "learning_rate": 8.792728709741158e-06,
+      "loss": 0.2905,
+      "step": 2838
+    },
+    {
+      "epoch": 1.6823703703703705,
+      "grad_norm": 1.706667423248291,
+      "learning_rate": 8.788776921557006e-06,
+      "loss": 0.2491,
+      "step": 2839
+    },
+    {
+      "epoch": 1.682962962962963,
+      "grad_norm": 2.1374008655548096,
+      "learning_rate": 8.784825133372852e-06,
+      "loss": 0.2996,
+      "step": 2840
+    },
+    {
+      "epoch": 1.6835555555555555,
+      "grad_norm": 1.6799591779708862,
+      "learning_rate": 8.780873345188698e-06,
+      "loss": 0.1944,
+      "step": 2841
+    },
+    {
+      "epoch": 1.6841481481481482,
+      "grad_norm": 2.2824366092681885,
+      "learning_rate": 8.776921557004546e-06,
+      "loss": 0.3228,
+      "step": 2842
+    },
+    {
+      "epoch": 1.6847407407407409,
+      "grad_norm": 1.4887539148330688,
+      "learning_rate": 8.772969768820392e-06,
+      "loss": 0.2214,
+      "step": 2843
+    },
+    {
+      "epoch": 1.6853333333333333,
+      "grad_norm": 1.9967490434646606,
+      "learning_rate": 8.769017980636238e-06,
+      "loss": 0.3388,
+      "step": 2844
+    },
+    {
+      "epoch": 1.6859259259259258,
+      "grad_norm": 1.6254206895828247,
+      "learning_rate": 8.765066192452085e-06,
+      "loss": 0.2319,
+      "step": 2845
+    },
+    {
+      "epoch": 1.6865185185185185,
+      "grad_norm": 2.1348068714141846,
+      "learning_rate": 8.761114404267933e-06,
+      "loss": 0.3529,
+      "step": 2846
+    },
+    {
+      "epoch": 1.6871111111111112,
+      "grad_norm": 1.6326017379760742,
+      "learning_rate": 8.757162616083779e-06,
+      "loss": 0.2269,
+      "step": 2847
+    },
+    {
+      "epoch": 1.6877037037037037,
+      "grad_norm": 1.7441834211349487,
+      "learning_rate": 8.753210827899625e-06,
+      "loss": 0.1887,
+      "step": 2848
+    },
+    {
+      "epoch": 1.6882962962962962,
+      "grad_norm": 2.0708303451538086,
+      "learning_rate": 8.749259039715471e-06,
+      "loss": 0.2068,
+      "step": 2849
+    },
+    {
+      "epoch": 1.6888888888888889,
+      "grad_norm": 2.4660208225250244,
+      "learning_rate": 8.74530725153132e-06,
+      "loss": 0.2022,
+      "step": 2850
+    },
+    {
+      "epoch": 1.6894814814814816,
+      "grad_norm": 2.1973893642425537,
+      "learning_rate": 8.741355463347166e-06,
+      "loss": 0.2389,
+      "step": 2851
+    },
+    {
+      "epoch": 1.690074074074074,
+      "grad_norm": 1.9153367280960083,
+      "learning_rate": 8.737403675163012e-06,
+      "loss": 0.2229,
+      "step": 2852
+    },
+    {
+      "epoch": 1.6906666666666665,
+      "grad_norm": 2.3082966804504395,
+      "learning_rate": 8.733451886978858e-06,
+      "loss": 0.2353,
+      "step": 2853
+    },
+    {
+      "epoch": 1.6912592592592592,
+      "grad_norm": 2.2462527751922607,
+      "learning_rate": 8.729500098794706e-06,
+      "loss": 0.2943,
+      "step": 2854
+    },
+    {
+      "epoch": 1.691851851851852,
+      "grad_norm": 2.0578770637512207,
+      "learning_rate": 8.725548310610552e-06,
+      "loss": 0.2361,
+      "step": 2855
+    },
+    {
+      "epoch": 1.6924444444444444,
+      "grad_norm": 2.0418195724487305,
+      "learning_rate": 8.721596522426398e-06,
+      "loss": 0.1903,
+      "step": 2856
+    },
+    {
+      "epoch": 1.693037037037037,
+      "grad_norm": 2.42622971534729,
+      "learning_rate": 8.717644734242245e-06,
+      "loss": 0.2753,
+      "step": 2857
+    },
+    {
+      "epoch": 1.6936296296296296,
+      "grad_norm": 1.743391990661621,
+      "learning_rate": 8.713692946058093e-06,
+      "loss": 0.2844,
+      "step": 2858
+    },
+    {
+      "epoch": 1.6942222222222223,
+      "grad_norm": 1.551126480102539,
+      "learning_rate": 8.709741157873939e-06,
+      "loss": 0.173,
+      "step": 2859
+    },
+    {
+      "epoch": 1.6948148148148148,
+      "grad_norm": 2.2807254791259766,
+      "learning_rate": 8.705789369689785e-06,
+      "loss": 0.2946,
+      "step": 2860
+    },
+    {
+      "epoch": 1.6954074074074073,
+      "grad_norm": 2.253286361694336,
+      "learning_rate": 8.701837581505631e-06,
+      "loss": 0.341,
+      "step": 2861
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 2.3700857162475586,
+      "learning_rate": 8.69788579332148e-06,
+      "loss": 0.3365,
+      "step": 2862
+    },
+    {
+      "epoch": 1.6965925925925927,
+      "grad_norm": 2.628281593322754,
+      "learning_rate": 8.693934005137326e-06,
+      "loss": 0.3553,
+      "step": 2863
+    },
+    {
+      "epoch": 1.6971851851851851,
+      "grad_norm": 2.3594257831573486,
+      "learning_rate": 8.689982216953172e-06,
+      "loss": 0.3051,
+      "step": 2864
+    },
+    {
+      "epoch": 1.6977777777777778,
+      "grad_norm": 2.216641902923584,
+      "learning_rate": 8.68603042876902e-06,
+      "loss": 0.2729,
+      "step": 2865
+    },
+    {
+      "epoch": 1.6983703703703705,
+      "grad_norm": 2.51208758354187,
+      "learning_rate": 8.682078640584866e-06,
+      "loss": 0.3234,
+      "step": 2866
+    },
+    {
+      "epoch": 1.698962962962963,
+      "grad_norm": 1.8584808111190796,
+      "learning_rate": 8.678126852400712e-06,
+      "loss": 0.3049,
+      "step": 2867
+    },
+    {
+      "epoch": 1.6995555555555555,
+      "grad_norm": 2.0900657176971436,
+      "learning_rate": 8.674175064216558e-06,
+      "loss": 0.3251,
+      "step": 2868
+    },
+    {
+      "epoch": 1.7001481481481482,
+      "grad_norm": 2.4790258407592773,
+      "learning_rate": 8.670223276032406e-06,
+      "loss": 0.276,
+      "step": 2869
+    },
+    {
+      "epoch": 1.7007407407407409,
+      "grad_norm": 2.2566518783569336,
+      "learning_rate": 8.666271487848253e-06,
+      "loss": 0.4074,
+      "step": 2870
+    },
+    {
+      "epoch": 1.7013333333333334,
+      "grad_norm": 2.0057289600372314,
+      "learning_rate": 8.662319699664099e-06,
+      "loss": 0.278,
+      "step": 2871
+    },
+    {
+      "epoch": 1.7019259259259258,
+      "grad_norm": 2.800534248352051,
+      "learning_rate": 8.658367911479945e-06,
+      "loss": 0.2821,
+      "step": 2872
+    },
+    {
+      "epoch": 1.7025185185185185,
+      "grad_norm": 1.7894115447998047,
+      "learning_rate": 8.654416123295793e-06,
+      "loss": 0.1826,
+      "step": 2873
+    },
+    {
+      "epoch": 1.7031111111111112,
+      "grad_norm": 1.9180102348327637,
+      "learning_rate": 8.65046433511164e-06,
+      "loss": 0.3676,
+      "step": 2874
+    },
+    {
+      "epoch": 1.7037037037037037,
+      "grad_norm": 2.375399351119995,
+      "learning_rate": 8.646512546927486e-06,
+      "loss": 0.2943,
+      "step": 2875
+    },
+    {
+      "epoch": 1.7042962962962962,
+      "grad_norm": 1.9785064458847046,
+      "learning_rate": 8.642560758743332e-06,
+      "loss": 0.2483,
+      "step": 2876
+    },
+    {
+      "epoch": 1.704888888888889,
+      "grad_norm": 3.2112808227539062,
+      "learning_rate": 8.63860897055918e-06,
+      "loss": 0.4526,
+      "step": 2877
+    },
+    {
+      "epoch": 1.7054814814814816,
+      "grad_norm": 2.065176010131836,
+      "learning_rate": 8.634657182375026e-06,
+      "loss": 0.2448,
+      "step": 2878
+    },
+    {
+      "epoch": 1.706074074074074,
+      "grad_norm": 2.117818593978882,
+      "learning_rate": 8.630705394190872e-06,
+      "loss": 0.2334,
+      "step": 2879
+    },
+    {
+      "epoch": 1.7066666666666666,
+      "grad_norm": 2.058741807937622,
+      "learning_rate": 8.626753606006718e-06,
+      "loss": 0.3307,
+      "step": 2880
+    },
+    {
+      "epoch": 1.7072592592592593,
+      "grad_norm": 1.7736985683441162,
+      "learning_rate": 8.622801817822566e-06,
+      "loss": 0.2131,
+      "step": 2881
+    },
+    {
+      "epoch": 1.707851851851852,
+      "grad_norm": 1.9215459823608398,
+      "learning_rate": 8.618850029638413e-06,
+      "loss": 0.3121,
+      "step": 2882
+    },
+    {
+      "epoch": 1.7084444444444444,
+      "grad_norm": 2.0311477184295654,
+      "learning_rate": 8.614898241454259e-06,
+      "loss": 0.2634,
+      "step": 2883
+    },
+    {
+      "epoch": 1.709037037037037,
+      "grad_norm": 2.4358582496643066,
+      "learning_rate": 8.610946453270105e-06,
+      "loss": 0.2941,
+      "step": 2884
+    },
+    {
+      "epoch": 1.7096296296296296,
+      "grad_norm": 2.7107431888580322,
+      "learning_rate": 8.606994665085953e-06,
+      "loss": 0.4133,
+      "step": 2885
+    },
+    {
+      "epoch": 1.7102222222222223,
+      "grad_norm": 1.6590828895568848,
+      "learning_rate": 8.6030428769018e-06,
+      "loss": 0.2716,
+      "step": 2886
+    },
+    {
+      "epoch": 1.7108148148148148,
+      "grad_norm": 2.043215274810791,
+      "learning_rate": 8.599091088717646e-06,
+      "loss": 0.2558,
+      "step": 2887
+    },
+    {
+      "epoch": 1.7114074074074073,
+      "grad_norm": 2.3831119537353516,
+      "learning_rate": 8.595139300533493e-06,
+      "loss": 0.3039,
+      "step": 2888
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 2.7713935375213623,
+      "learning_rate": 8.591187512349338e-06,
+      "loss": 0.4481,
+      "step": 2889
+    },
+    {
+      "epoch": 1.7125925925925927,
+      "grad_norm": 1.983169436454773,
+      "learning_rate": 8.587235724165184e-06,
+      "loss": 0.2495,
+      "step": 2890
+    },
+    {
+      "epoch": 1.7131851851851851,
+      "grad_norm": 6.111391544342041,
+      "learning_rate": 8.583283935981032e-06,
+      "loss": 0.6122,
+      "step": 2891
+    },
+    {
+      "epoch": 1.7137777777777776,
+      "grad_norm": 2.0180039405822754,
+      "learning_rate": 8.579332147796878e-06,
+      "loss": 0.3148,
+      "step": 2892
+    },
+    {
+      "epoch": 1.7143703703703703,
+      "grad_norm": 2.7580714225769043,
+      "learning_rate": 8.575380359612725e-06,
+      "loss": 0.3372,
+      "step": 2893
+    },
+    {
+      "epoch": 1.714962962962963,
+      "grad_norm": 1.7227360010147095,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.2058,
+      "step": 2894
+    },
+    {
+      "epoch": 1.7155555555555555,
+      "grad_norm": 1.9132333993911743,
+      "learning_rate": 8.567476783244419e-06,
+      "loss": 0.2188,
+      "step": 2895
+    },
+    {
+      "epoch": 1.7161481481481482,
+      "grad_norm": 2.9416844844818115,
+      "learning_rate": 8.563524995060265e-06,
+      "loss": 0.3046,
+      "step": 2896
+    },
+    {
+      "epoch": 1.716740740740741,
+      "grad_norm": 1.5822827816009521,
+      "learning_rate": 8.559573206876111e-06,
+      "loss": 0.1701,
+      "step": 2897
+    },
+    {
+      "epoch": 1.7173333333333334,
+      "grad_norm": 1.6582123041152954,
+      "learning_rate": 8.555621418691958e-06,
+      "loss": 0.2643,
+      "step": 2898
+    },
+    {
+      "epoch": 1.7179259259259259,
+      "grad_norm": 1.8421573638916016,
+      "learning_rate": 8.551669630507806e-06,
+      "loss": 0.2903,
+      "step": 2899
+    },
+    {
+      "epoch": 1.7185185185185186,
+      "grad_norm": 2.6566593647003174,
+      "learning_rate": 8.547717842323652e-06,
+      "loss": 0.2844,
+      "step": 2900
+    },
+    {
+      "epoch": 1.7191111111111113,
+      "grad_norm": 1.963384985923767,
+      "learning_rate": 8.543766054139498e-06,
+      "loss": 0.2325,
+      "step": 2901
+    },
+    {
+      "epoch": 1.7197037037037037,
+      "grad_norm": 2.8680315017700195,
+      "learning_rate": 8.539814265955346e-06,
+      "loss": 0.2599,
+      "step": 2902
+    },
+    {
+      "epoch": 1.7202962962962962,
+      "grad_norm": 2.8816235065460205,
+      "learning_rate": 8.535862477771192e-06,
+      "loss": 0.3923,
+      "step": 2903
+    },
+    {
+      "epoch": 1.720888888888889,
+      "grad_norm": 1.6320559978485107,
+      "learning_rate": 8.531910689587038e-06,
+      "loss": 0.2347,
+      "step": 2904
+    },
+    {
+      "epoch": 1.7214814814814816,
+      "grad_norm": 1.5693448781967163,
+      "learning_rate": 8.527958901402885e-06,
+      "loss": 0.2242,
+      "step": 2905
+    },
+    {
+      "epoch": 1.722074074074074,
+      "grad_norm": 2.0983541011810303,
+      "learning_rate": 8.524007113218733e-06,
+      "loss": 0.336,
+      "step": 2906
+    },
+    {
+      "epoch": 1.7226666666666666,
+      "grad_norm": 1.8689796924591064,
+      "learning_rate": 8.520055325034579e-06,
+      "loss": 0.2132,
+      "step": 2907
+    },
+    {
+      "epoch": 1.7232592592592593,
+      "grad_norm": 2.421116590499878,
+      "learning_rate": 8.516103536850425e-06,
+      "loss": 0.2959,
+      "step": 2908
+    },
+    {
+      "epoch": 1.723851851851852,
+      "grad_norm": 1.5992075204849243,
+      "learning_rate": 8.512151748666271e-06,
+      "loss": 0.1645,
+      "step": 2909
+    },
+    {
+      "epoch": 1.7244444444444444,
+      "grad_norm": 1.6104823350906372,
+      "learning_rate": 8.50819996048212e-06,
+      "loss": 0.1927,
+      "step": 2910
+    },
+    {
+      "epoch": 1.725037037037037,
+      "grad_norm": 2.315183639526367,
+      "learning_rate": 8.504248172297966e-06,
+      "loss": 0.2462,
+      "step": 2911
+    },
+    {
+      "epoch": 1.7256296296296296,
+      "grad_norm": 2.8115692138671875,
+      "learning_rate": 8.500296384113812e-06,
+      "loss": 0.3622,
+      "step": 2912
+    },
+    {
+      "epoch": 1.7262222222222223,
+      "grad_norm": 2.1504404544830322,
+      "learning_rate": 8.496344595929658e-06,
+      "loss": 0.2956,
+      "step": 2913
+    },
+    {
+      "epoch": 1.7268148148148148,
+      "grad_norm": 2.1159141063690186,
+      "learning_rate": 8.492392807745506e-06,
+      "loss": 0.2547,
+      "step": 2914
+    },
+    {
+      "epoch": 1.7274074074074073,
+      "grad_norm": 2.3364527225494385,
+      "learning_rate": 8.488441019561352e-06,
+      "loss": 0.1937,
+      "step": 2915
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 1.8078352212905884,
+      "learning_rate": 8.484489231377198e-06,
+      "loss": 0.2498,
+      "step": 2916
+    },
+    {
+      "epoch": 1.7285925925925927,
+      "grad_norm": 2.0641448497772217,
+      "learning_rate": 8.480537443193045e-06,
+      "loss": 0.2382,
+      "step": 2917
+    },
+    {
+      "epoch": 1.7291851851851852,
+      "grad_norm": 3.466825008392334,
+      "learning_rate": 8.476585655008893e-06,
+      "loss": 0.2264,
+      "step": 2918
+    },
+    {
+      "epoch": 1.7297777777777776,
+      "grad_norm": 2.99130916595459,
+      "learning_rate": 8.472633866824739e-06,
+      "loss": 0.286,
+      "step": 2919
+    },
+    {
+      "epoch": 1.7303703703703703,
+      "grad_norm": 2.4810690879821777,
+      "learning_rate": 8.468682078640585e-06,
+      "loss": 0.1685,
+      "step": 2920
+    },
+    {
+      "epoch": 1.730962962962963,
+      "grad_norm": 3.098957061767578,
+      "learning_rate": 8.464730290456431e-06,
+      "loss": 0.4656,
+      "step": 2921
+    },
+    {
+      "epoch": 1.7315555555555555,
+      "grad_norm": 2.275949001312256,
+      "learning_rate": 8.46077850227228e-06,
+      "loss": 0.3514,
+      "step": 2922
+    },
+    {
+      "epoch": 1.732148148148148,
+      "grad_norm": 2.3620734214782715,
+      "learning_rate": 8.456826714088126e-06,
+      "loss": 0.332,
+      "step": 2923
+    },
+    {
+      "epoch": 1.7327407407407407,
+      "grad_norm": 1.7329065799713135,
+      "learning_rate": 8.452874925903972e-06,
+      "loss": 0.2221,
+      "step": 2924
+    },
+    {
+      "epoch": 1.7333333333333334,
+      "grad_norm": 2.2002100944519043,
+      "learning_rate": 8.44892313771982e-06,
+      "loss": 0.2994,
+      "step": 2925
+    },
+    {
+      "epoch": 1.7339259259259259,
+      "grad_norm": 2.2748970985412598,
+      "learning_rate": 8.444971349535666e-06,
+      "loss": 0.3383,
+      "step": 2926
+    },
+    {
+      "epoch": 1.7345185185185186,
+      "grad_norm": 1.6602650880813599,
+      "learning_rate": 8.441019561351512e-06,
+      "loss": 0.2007,
+      "step": 2927
+    },
+    {
+      "epoch": 1.7351111111111113,
+      "grad_norm": 1.9247047901153564,
+      "learning_rate": 8.437067773167358e-06,
+      "loss": 0.2801,
+      "step": 2928
+    },
+    {
+      "epoch": 1.7357037037037037,
+      "grad_norm": 1.5850194692611694,
+      "learning_rate": 8.433115984983206e-06,
+      "loss": 0.2047,
+      "step": 2929
+    },
+    {
+      "epoch": 1.7362962962962962,
+      "grad_norm": 1.4122494459152222,
+      "learning_rate": 8.429164196799053e-06,
+      "loss": 0.1994,
+      "step": 2930
+    },
+    {
+      "epoch": 1.736888888888889,
+      "grad_norm": 2.364133358001709,
+      "learning_rate": 8.425212408614899e-06,
+      "loss": 0.3973,
+      "step": 2931
+    },
+    {
+      "epoch": 1.7374814814814816,
+      "grad_norm": 1.847122311592102,
+      "learning_rate": 8.421260620430745e-06,
+      "loss": 0.2008,
+      "step": 2932
+    },
+    {
+      "epoch": 1.738074074074074,
+      "grad_norm": 2.0557382106781006,
+      "learning_rate": 8.417308832246593e-06,
+      "loss": 0.2939,
+      "step": 2933
+    },
+    {
+      "epoch": 1.7386666666666666,
+      "grad_norm": 1.988776683807373,
+      "learning_rate": 8.41335704406244e-06,
+      "loss": 0.3454,
+      "step": 2934
+    },
+    {
+      "epoch": 1.7392592592592593,
+      "grad_norm": 2.1146645545959473,
+      "learning_rate": 8.409405255878286e-06,
+      "loss": 0.2798,
+      "step": 2935
+    },
+    {
+      "epoch": 1.739851851851852,
+      "grad_norm": 1.9327672719955444,
+      "learning_rate": 8.405453467694132e-06,
+      "loss": 0.2415,
+      "step": 2936
+    },
+    {
+      "epoch": 1.7404444444444445,
+      "grad_norm": 2.176947832107544,
+      "learning_rate": 8.40150167950998e-06,
+      "loss": 0.3286,
+      "step": 2937
+    },
+    {
+      "epoch": 1.741037037037037,
+      "grad_norm": 1.69938063621521,
+      "learning_rate": 8.397549891325826e-06,
+      "loss": 0.2778,
+      "step": 2938
+    },
+    {
+      "epoch": 1.7416296296296296,
+      "grad_norm": 2.343254566192627,
+      "learning_rate": 8.393598103141672e-06,
+      "loss": 0.3148,
+      "step": 2939
+    },
+    {
+      "epoch": 1.7422222222222223,
+      "grad_norm": 2.3646512031555176,
+      "learning_rate": 8.389646314957518e-06,
+      "loss": 0.3097,
+      "step": 2940
+    },
+    {
+      "epoch": 1.7428148148148148,
+      "grad_norm": 2.1847479343414307,
+      "learning_rate": 8.385694526773366e-06,
+      "loss": 0.3603,
+      "step": 2941
+    },
+    {
+      "epoch": 1.7434074074074073,
+      "grad_norm": 1.7832756042480469,
+      "learning_rate": 8.381742738589213e-06,
+      "loss": 0.2357,
+      "step": 2942
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 2.0364773273468018,
+      "learning_rate": 8.377790950405059e-06,
+      "loss": 0.2305,
+      "step": 2943
+    },
+    {
+      "epoch": 1.7445925925925927,
+      "grad_norm": 1.95876944065094,
+      "learning_rate": 8.373839162220905e-06,
+      "loss": 0.2819,
+      "step": 2944
+    },
+    {
+      "epoch": 1.7451851851851852,
+      "grad_norm": 1.429614782333374,
+      "learning_rate": 8.369887374036753e-06,
+      "loss": 0.171,
+      "step": 2945
+    },
+    {
+      "epoch": 1.7457777777777777,
+      "grad_norm": 2.6756839752197266,
+      "learning_rate": 8.3659355858526e-06,
+      "loss": 0.2434,
+      "step": 2946
+    },
+    {
+      "epoch": 1.7463703703703704,
+      "grad_norm": 2.2092795372009277,
+      "learning_rate": 8.361983797668446e-06,
+      "loss": 0.289,
+      "step": 2947
+    },
+    {
+      "epoch": 1.746962962962963,
+      "grad_norm": 1.7894587516784668,
+      "learning_rate": 8.358032009484293e-06,
+      "loss": 0.1948,
+      "step": 2948
+    },
+    {
+      "epoch": 1.7475555555555555,
+      "grad_norm": 1.7683628797531128,
+      "learning_rate": 8.35408022130014e-06,
+      "loss": 0.2119,
+      "step": 2949
+    },
+    {
+      "epoch": 1.748148148148148,
+      "grad_norm": 2.608316659927368,
+      "learning_rate": 8.350128433115986e-06,
+      "loss": 0.3478,
+      "step": 2950
+    },
+    {
+      "epoch": 1.7487407407407407,
+      "grad_norm": 2.577648878097534,
+      "learning_rate": 8.346176644931832e-06,
+      "loss": 0.2625,
+      "step": 2951
+    },
+    {
+      "epoch": 1.7493333333333334,
+      "grad_norm": 2.7357189655303955,
+      "learning_rate": 8.34222485674768e-06,
+      "loss": 0.3469,
+      "step": 2952
+    },
+    {
+      "epoch": 1.7499259259259259,
+      "grad_norm": 2.435128688812256,
+      "learning_rate": 8.338273068563526e-06,
+      "loss": 0.3689,
+      "step": 2953
+    },
+    {
+      "epoch": 1.7505185185185184,
+      "grad_norm": 3.0442380905151367,
+      "learning_rate": 8.334321280379373e-06,
+      "loss": 0.3639,
+      "step": 2954
+    },
+    {
+      "epoch": 1.751111111111111,
+      "grad_norm": 2.3630573749542236,
+      "learning_rate": 8.330369492195219e-06,
+      "loss": 0.3756,
+      "step": 2955
+    },
+    {
+      "epoch": 1.7517037037037038,
+      "grad_norm": 2.4501821994781494,
+      "learning_rate": 8.326417704011067e-06,
+      "loss": 0.2499,
+      "step": 2956
+    },
+    {
+      "epoch": 1.7522962962962962,
+      "grad_norm": 2.023219585418701,
+      "learning_rate": 8.322465915826913e-06,
+      "loss": 0.2409,
+      "step": 2957
+    },
+    {
+      "epoch": 1.752888888888889,
+      "grad_norm": 3.176694393157959,
+      "learning_rate": 8.31851412764276e-06,
+      "loss": 0.2139,
+      "step": 2958
+    },
+    {
+      "epoch": 1.7534814814814816,
+      "grad_norm": 2.305670976638794,
+      "learning_rate": 8.314562339458606e-06,
+      "loss": 0.2927,
+      "step": 2959
+    },
+    {
+      "epoch": 1.7540740740740741,
+      "grad_norm": 2.238044261932373,
+      "learning_rate": 8.310610551274453e-06,
+      "loss": 0.2721,
+      "step": 2960
+    },
+    {
+      "epoch": 1.7546666666666666,
+      "grad_norm": 1.803071141242981,
+      "learning_rate": 8.3066587630903e-06,
+      "loss": 0.2879,
+      "step": 2961
+    },
+    {
+      "epoch": 1.7552592592592593,
+      "grad_norm": 1.948067307472229,
+      "learning_rate": 8.302706974906146e-06,
+      "loss": 0.2293,
+      "step": 2962
+    },
+    {
+      "epoch": 1.755851851851852,
+      "grad_norm": 1.7984262704849243,
+      "learning_rate": 8.298755186721992e-06,
+      "loss": 0.2166,
+      "step": 2963
+    },
+    {
+      "epoch": 1.7564444444444445,
+      "grad_norm": 1.872652292251587,
+      "learning_rate": 8.29480339853784e-06,
+      "loss": 0.2489,
+      "step": 2964
+    },
+    {
+      "epoch": 1.757037037037037,
+      "grad_norm": 1.9965965747833252,
+      "learning_rate": 8.290851610353686e-06,
+      "loss": 0.2413,
+      "step": 2965
+    },
+    {
+      "epoch": 1.7576296296296297,
+      "grad_norm": 1.6477789878845215,
+      "learning_rate": 8.286899822169533e-06,
+      "loss": 0.2138,
+      "step": 2966
+    },
+    {
+      "epoch": 1.7582222222222224,
+      "grad_norm": 2.6680285930633545,
+      "learning_rate": 8.282948033985379e-06,
+      "loss": 0.332,
+      "step": 2967
+    },
+    {
+      "epoch": 1.7588148148148148,
+      "grad_norm": 2.2727818489074707,
+      "learning_rate": 8.278996245801225e-06,
+      "loss": 0.2222,
+      "step": 2968
+    },
+    {
+      "epoch": 1.7594074074074073,
+      "grad_norm": 2.7599010467529297,
+      "learning_rate": 8.275044457617071e-06,
+      "loss": 0.2909,
+      "step": 2969
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8476340770721436,
+      "learning_rate": 8.27109266943292e-06,
+      "loss": 0.2374,
+      "step": 2970
+    },
+    {
+      "epoch": 1.7605925925925927,
+      "grad_norm": 3.072150707244873,
+      "learning_rate": 8.267140881248766e-06,
+      "loss": 0.4001,
+      "step": 2971
+    },
+    {
+      "epoch": 1.7611851851851852,
+      "grad_norm": 2.039106607437134,
+      "learning_rate": 8.263189093064612e-06,
+      "loss": 0.2581,
+      "step": 2972
+    },
+    {
+      "epoch": 1.7617777777777777,
+      "grad_norm": 1.4507689476013184,
+      "learning_rate": 8.259237304880458e-06,
+      "loss": 0.1995,
+      "step": 2973
+    },
+    {
+      "epoch": 1.7623703703703704,
+      "grad_norm": 2.490492820739746,
+      "learning_rate": 8.255285516696306e-06,
+      "loss": 0.4062,
+      "step": 2974
+    },
+    {
+      "epoch": 1.762962962962963,
+      "grad_norm": 2.366874933242798,
+      "learning_rate": 8.251333728512152e-06,
+      "loss": 0.2692,
+      "step": 2975
+    },
+    {
+      "epoch": 1.7635555555555555,
+      "grad_norm": 3.0400590896606445,
+      "learning_rate": 8.247381940327998e-06,
+      "loss": 0.2578,
+      "step": 2976
+    },
+    {
+      "epoch": 1.764148148148148,
+      "grad_norm": 2.5713517665863037,
+      "learning_rate": 8.243430152143845e-06,
+      "loss": 0.2071,
+      "step": 2977
+    },
+    {
+      "epoch": 1.7647407407407407,
+      "grad_norm": 1.9291075468063354,
+      "learning_rate": 8.239478363959693e-06,
+      "loss": 0.2121,
+      "step": 2978
+    },
+    {
+      "epoch": 1.7653333333333334,
+      "grad_norm": 2.528790235519409,
+      "learning_rate": 8.235526575775539e-06,
+      "loss": 0.26,
+      "step": 2979
+    },
+    {
+      "epoch": 1.765925925925926,
+      "grad_norm": 1.9279264211654663,
+      "learning_rate": 8.231574787591385e-06,
+      "loss": 0.2812,
+      "step": 2980
+    },
+    {
+      "epoch": 1.7665185185185184,
+      "grad_norm": 2.457944631576538,
+      "learning_rate": 8.227622999407231e-06,
+      "loss": 0.1982,
+      "step": 2981
+    },
+    {
+      "epoch": 1.767111111111111,
+      "grad_norm": 2.3146069049835205,
+      "learning_rate": 8.22367121122308e-06,
+      "loss": 0.3177,
+      "step": 2982
+    },
+    {
+      "epoch": 1.7677037037037038,
+      "grad_norm": 3.0191195011138916,
+      "learning_rate": 8.219719423038925e-06,
+      "loss": 0.346,
+      "step": 2983
+    },
+    {
+      "epoch": 1.7682962962962963,
+      "grad_norm": 1.8780372142791748,
+      "learning_rate": 8.215767634854772e-06,
+      "loss": 0.2511,
+      "step": 2984
+    },
+    {
+      "epoch": 1.7688888888888887,
+      "grad_norm": 2.280994176864624,
+      "learning_rate": 8.211815846670618e-06,
+      "loss": 0.4493,
+      "step": 2985
+    },
+    {
+      "epoch": 1.7694814814814814,
+      "grad_norm": 1.911859154701233,
+      "learning_rate": 8.207864058486466e-06,
+      "loss": 0.2489,
+      "step": 2986
+    },
+    {
+      "epoch": 1.7700740740740741,
+      "grad_norm": 2.2034168243408203,
+      "learning_rate": 8.203912270302312e-06,
+      "loss": 0.2552,
+      "step": 2987
+    },
+    {
+      "epoch": 1.7706666666666666,
+      "grad_norm": 2.1232247352600098,
+      "learning_rate": 8.199960482118158e-06,
+      "loss": 0.2662,
+      "step": 2988
+    },
+    {
+      "epoch": 1.7712592592592593,
+      "grad_norm": 1.919485092163086,
+      "learning_rate": 8.196008693934006e-06,
+      "loss": 0.2406,
+      "step": 2989
+    },
+    {
+      "epoch": 1.771851851851852,
+      "grad_norm": 2.682098627090454,
+      "learning_rate": 8.192056905749853e-06,
+      "loss": 0.2665,
+      "step": 2990
+    },
+    {
+      "epoch": 1.7724444444444445,
+      "grad_norm": 2.4153554439544678,
+      "learning_rate": 8.188105117565699e-06,
+      "loss": 0.3441,
+      "step": 2991
+    },
+    {
+      "epoch": 1.773037037037037,
+      "grad_norm": 1.5661629438400269,
+      "learning_rate": 8.184153329381545e-06,
+      "loss": 0.1985,
+      "step": 2992
+    },
+    {
+      "epoch": 1.7736296296296297,
+      "grad_norm": 2.7832863330841064,
+      "learning_rate": 8.180201541197393e-06,
+      "loss": 0.303,
+      "step": 2993
+    },
+    {
+      "epoch": 1.7742222222222224,
+      "grad_norm": 1.8000779151916504,
+      "learning_rate": 8.17624975301324e-06,
+      "loss": 0.2073,
+      "step": 2994
+    },
+    {
+      "epoch": 1.7748148148148148,
+      "grad_norm": 1.5725908279418945,
+      "learning_rate": 8.172297964829085e-06,
+      "loss": 0.203,
+      "step": 2995
+    },
+    {
+      "epoch": 1.7754074074074073,
+      "grad_norm": 2.0886497497558594,
+      "learning_rate": 8.168346176644932e-06,
+      "loss": 0.3373,
+      "step": 2996
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 1.9843171834945679,
+      "learning_rate": 8.16439438846078e-06,
+      "loss": 0.3581,
+      "step": 2997
+    },
+    {
+      "epoch": 1.7765925925925927,
+      "grad_norm": 2.2643091678619385,
+      "learning_rate": 8.160442600276626e-06,
+      "loss": 0.2732,
+      "step": 2998
+    },
+    {
+      "epoch": 1.7771851851851852,
+      "grad_norm": 2.187854051589966,
+      "learning_rate": 8.156490812092472e-06,
+      "loss": 0.3721,
+      "step": 2999
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 1.9970648288726807,
+      "learning_rate": 8.152539023908318e-06,
+      "loss": 0.1862,
+      "step": 3000
+    },
+    {
+      "epoch": 1.7783703703703704,
+      "grad_norm": 2.008411407470703,
+      "learning_rate": 8.148587235724166e-06,
+      "loss": 0.2401,
+      "step": 3001
+    },
+    {
+      "epoch": 1.778962962962963,
+      "grad_norm": 1.5912891626358032,
+      "learning_rate": 8.144635447540013e-06,
+      "loss": 0.2043,
+      "step": 3002
+    },
+    {
+      "epoch": 1.7795555555555556,
+      "grad_norm": 2.0794525146484375,
+      "learning_rate": 8.140683659355859e-06,
+      "loss": 0.2489,
+      "step": 3003
+    },
+    {
+      "epoch": 1.780148148148148,
+      "grad_norm": 2.972248077392578,
+      "learning_rate": 8.136731871171705e-06,
+      "loss": 0.2191,
+      "step": 3004
+    },
+    {
+      "epoch": 1.7807407407407407,
+      "grad_norm": 2.9418039321899414,
+      "learning_rate": 8.132780082987553e-06,
+      "loss": 0.302,
+      "step": 3005
+    },
+    {
+      "epoch": 1.7813333333333334,
+      "grad_norm": 2.9893434047698975,
+      "learning_rate": 8.1288282948034e-06,
+      "loss": 0.3523,
+      "step": 3006
+    },
+    {
+      "epoch": 1.781925925925926,
+      "grad_norm": 4.387885570526123,
+      "learning_rate": 8.124876506619245e-06,
+      "loss": 0.3947,
+      "step": 3007
+    },
+    {
+      "epoch": 1.7825185185185184,
+      "grad_norm": 2.5193471908569336,
+      "learning_rate": 8.120924718435092e-06,
+      "loss": 0.3693,
+      "step": 3008
+    },
+    {
+      "epoch": 1.783111111111111,
+      "grad_norm": 2.745957851409912,
+      "learning_rate": 8.11697293025094e-06,
+      "loss": 0.2602,
+      "step": 3009
+    },
+    {
+      "epoch": 1.7837037037037038,
+      "grad_norm": 2.189441442489624,
+      "learning_rate": 8.113021142066786e-06,
+      "loss": 0.2415,
+      "step": 3010
+    },
+    {
+      "epoch": 1.7842962962962963,
+      "grad_norm": 2.023681640625,
+      "learning_rate": 8.109069353882632e-06,
+      "loss": 0.2385,
+      "step": 3011
+    },
+    {
+      "epoch": 1.7848888888888887,
+      "grad_norm": 2.4276437759399414,
+      "learning_rate": 8.10511756569848e-06,
+      "loss": 0.4173,
+      "step": 3012
+    },
+    {
+      "epoch": 1.7854814814814814,
+      "grad_norm": 2.375272035598755,
+      "learning_rate": 8.101165777514326e-06,
+      "loss": 0.2061,
+      "step": 3013
+    },
+    {
+      "epoch": 1.7860740740740741,
+      "grad_norm": 2.6144864559173584,
+      "learning_rate": 8.097213989330173e-06,
+      "loss": 0.305,
+      "step": 3014
+    },
+    {
+      "epoch": 1.7866666666666666,
+      "grad_norm": 2.7097716331481934,
+      "learning_rate": 8.093262201146019e-06,
+      "loss": 0.2614,
+      "step": 3015
+    },
+    {
+      "epoch": 1.787259259259259,
+      "grad_norm": 2.418483018875122,
+      "learning_rate": 8.089310412961867e-06,
+      "loss": 0.2571,
+      "step": 3016
+    },
+    {
+      "epoch": 1.7878518518518518,
+      "grad_norm": 2.1495654582977295,
+      "learning_rate": 8.085358624777713e-06,
+      "loss": 0.255,
+      "step": 3017
+    },
+    {
+      "epoch": 1.7884444444444445,
+      "grad_norm": 2.1345815658569336,
+      "learning_rate": 8.08140683659356e-06,
+      "loss": 0.2782,
+      "step": 3018
+    },
+    {
+      "epoch": 1.789037037037037,
+      "grad_norm": 2.0677096843719482,
+      "learning_rate": 8.077455048409405e-06,
+      "loss": 0.3863,
+      "step": 3019
+    },
+    {
+      "epoch": 1.7896296296296297,
+      "grad_norm": 2.5957748889923096,
+      "learning_rate": 8.073503260225253e-06,
+      "loss": 0.2965,
+      "step": 3020
+    },
+    {
+      "epoch": 1.7902222222222224,
+      "grad_norm": 2.2169229984283447,
+      "learning_rate": 8.0695514720411e-06,
+      "loss": 0.2442,
+      "step": 3021
+    },
+    {
+      "epoch": 1.7908148148148149,
+      "grad_norm": 2.02750301361084,
+      "learning_rate": 8.065599683856946e-06,
+      "loss": 0.2622,
+      "step": 3022
+    },
+    {
+      "epoch": 1.7914074074074073,
+      "grad_norm": 2.5613389015197754,
+      "learning_rate": 8.061647895672792e-06,
+      "loss": 0.3815,
+      "step": 3023
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 1.6193110942840576,
+      "learning_rate": 8.05769610748864e-06,
+      "loss": 0.2474,
+      "step": 3024
+    },
+    {
+      "epoch": 1.7925925925925927,
+      "grad_norm": 2.57372784614563,
+      "learning_rate": 8.053744319304486e-06,
+      "loss": 0.3541,
+      "step": 3025
+    },
+    {
+      "epoch": 1.7931851851851852,
+      "grad_norm": 1.764376163482666,
+      "learning_rate": 8.049792531120333e-06,
+      "loss": 0.215,
+      "step": 3026
+    },
+    {
+      "epoch": 1.7937777777777777,
+      "grad_norm": 2.3850200176239014,
+      "learning_rate": 8.045840742936179e-06,
+      "loss": 0.2668,
+      "step": 3027
+    },
+    {
+      "epoch": 1.7943703703703704,
+      "grad_norm": 2.2098610401153564,
+      "learning_rate": 8.041888954752027e-06,
+      "loss": 0.2888,
+      "step": 3028
+    },
+    {
+      "epoch": 1.794962962962963,
+      "grad_norm": 2.4571659564971924,
+      "learning_rate": 8.037937166567873e-06,
+      "loss": 0.2844,
+      "step": 3029
+    },
+    {
+      "epoch": 1.7955555555555556,
+      "grad_norm": 2.3217673301696777,
+      "learning_rate": 8.03398537838372e-06,
+      "loss": 0.2982,
+      "step": 3030
+    },
+    {
+      "epoch": 1.796148148148148,
+      "grad_norm": 2.602618932723999,
+      "learning_rate": 8.030033590199565e-06,
+      "loss": 0.3377,
+      "step": 3031
+    },
+    {
+      "epoch": 1.7967407407407407,
+      "grad_norm": 1.9024748802185059,
+      "learning_rate": 8.026081802015413e-06,
+      "loss": 0.2908,
+      "step": 3032
+    },
+    {
+      "epoch": 1.7973333333333334,
+      "grad_norm": 2.1253247261047363,
+      "learning_rate": 8.02213001383126e-06,
+      "loss": 0.2445,
+      "step": 3033
+    },
+    {
+      "epoch": 1.797925925925926,
+      "grad_norm": 1.4667162895202637,
+      "learning_rate": 8.018178225647106e-06,
+      "loss": 0.227,
+      "step": 3034
+    },
+    {
+      "epoch": 1.7985185185185184,
+      "grad_norm": 1.7433141469955444,
+      "learning_rate": 8.014226437462954e-06,
+      "loss": 0.2044,
+      "step": 3035
+    },
+    {
+      "epoch": 1.799111111111111,
+      "grad_norm": 2.3057003021240234,
+      "learning_rate": 8.0102746492788e-06,
+      "loss": 0.28,
+      "step": 3036
+    },
+    {
+      "epoch": 1.7997037037037038,
+      "grad_norm": 1.4725301265716553,
+      "learning_rate": 8.006322861094646e-06,
+      "loss": 0.1615,
+      "step": 3037
+    },
+    {
+      "epoch": 1.8002962962962963,
+      "grad_norm": 2.738370656967163,
+      "learning_rate": 8.002371072910493e-06,
+      "loss": 0.3547,
+      "step": 3038
+    },
+    {
+      "epoch": 1.8008888888888888,
+      "grad_norm": 2.1829466819763184,
+      "learning_rate": 7.99841928472634e-06,
+      "loss": 0.3351,
+      "step": 3039
+    },
+    {
+      "epoch": 1.8014814814814815,
+      "grad_norm": 1.810461163520813,
+      "learning_rate": 7.994467496542187e-06,
+      "loss": 0.1943,
+      "step": 3040
+    },
+    {
+      "epoch": 1.8020740740740742,
+      "grad_norm": 2.113142251968384,
+      "learning_rate": 7.990515708358033e-06,
+      "loss": 0.2663,
+      "step": 3041
+    },
+    {
+      "epoch": 1.8026666666666666,
+      "grad_norm": 1.5465829372406006,
+      "learning_rate": 7.98656392017388e-06,
+      "loss": 0.1356,
+      "step": 3042
+    },
+    {
+      "epoch": 1.8032592592592591,
+      "grad_norm": 2.470548152923584,
+      "learning_rate": 7.982612131989727e-06,
+      "loss": 0.3243,
+      "step": 3043
+    },
+    {
+      "epoch": 1.8038518518518518,
+      "grad_norm": 2.0462684631347656,
+      "learning_rate": 7.978660343805573e-06,
+      "loss": 0.2639,
+      "step": 3044
+    },
+    {
+      "epoch": 1.8044444444444445,
+      "grad_norm": 2.009002447128296,
+      "learning_rate": 7.97470855562142e-06,
+      "loss": 0.3602,
+      "step": 3045
+    },
+    {
+      "epoch": 1.805037037037037,
+      "grad_norm": 1.875101923942566,
+      "learning_rate": 7.970756767437266e-06,
+      "loss": 0.2196,
+      "step": 3046
+    },
+    {
+      "epoch": 1.8056296296296295,
+      "grad_norm": 2.0477206707000732,
+      "learning_rate": 7.966804979253112e-06,
+      "loss": 0.2642,
+      "step": 3047
+    },
+    {
+      "epoch": 1.8062222222222222,
+      "grad_norm": 1.8290022611618042,
+      "learning_rate": 7.962853191068958e-06,
+      "loss": 0.273,
+      "step": 3048
+    },
+    {
+      "epoch": 1.8068148148148149,
+      "grad_norm": 2.068995714187622,
+      "learning_rate": 7.958901402884806e-06,
+      "loss": 0.2648,
+      "step": 3049
+    },
+    {
+      "epoch": 1.8074074074074074,
+      "grad_norm": 2.27602219581604,
+      "learning_rate": 7.954949614700653e-06,
+      "loss": 0.2636,
+      "step": 3050
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 2.1052825450897217,
+      "learning_rate": 7.950997826516499e-06,
+      "loss": 0.2714,
+      "step": 3051
+    },
+    {
+      "epoch": 1.8085925925925928,
+      "grad_norm": 2.591813087463379,
+      "learning_rate": 7.947046038332345e-06,
+      "loss": 0.3921,
+      "step": 3052
+    },
+    {
+      "epoch": 1.8091851851851852,
+      "grad_norm": 2.0273871421813965,
+      "learning_rate": 7.943094250148193e-06,
+      "loss": 0.2469,
+      "step": 3053
+    },
+    {
+      "epoch": 1.8097777777777777,
+      "grad_norm": 2.046257495880127,
+      "learning_rate": 7.93914246196404e-06,
+      "loss": 0.272,
+      "step": 3054
+    },
+    {
+      "epoch": 1.8103703703703704,
+      "grad_norm": 1.8654817342758179,
+      "learning_rate": 7.935190673779885e-06,
+      "loss": 0.2664,
+      "step": 3055
+    },
+    {
+      "epoch": 1.810962962962963,
+      "grad_norm": 2.1208317279815674,
+      "learning_rate": 7.931238885595732e-06,
+      "loss": 0.2759,
+      "step": 3056
+    },
+    {
+      "epoch": 1.8115555555555556,
+      "grad_norm": 2.218083143234253,
+      "learning_rate": 7.92728709741158e-06,
+      "loss": 0.223,
+      "step": 3057
+    },
+    {
+      "epoch": 1.812148148148148,
+      "grad_norm": 1.9066376686096191,
+      "learning_rate": 7.923335309227426e-06,
+      "loss": 0.3361,
+      "step": 3058
+    },
+    {
+      "epoch": 1.8127407407407408,
+      "grad_norm": 3.236083984375,
+      "learning_rate": 7.919383521043272e-06,
+      "loss": 0.3731,
+      "step": 3059
+    },
+    {
+      "epoch": 1.8133333333333335,
+      "grad_norm": 1.7198506593704224,
+      "learning_rate": 7.915431732859118e-06,
+      "loss": 0.2218,
+      "step": 3060
+    },
+    {
+      "epoch": 1.813925925925926,
+      "grad_norm": 2.977860689163208,
+      "learning_rate": 7.911479944674966e-06,
+      "loss": 0.428,
+      "step": 3061
+    },
+    {
+      "epoch": 1.8145185185185184,
+      "grad_norm": 1.7213350534439087,
+      "learning_rate": 7.907528156490813e-06,
+      "loss": 0.2805,
+      "step": 3062
+    },
+    {
+      "epoch": 1.8151111111111111,
+      "grad_norm": 2.021777391433716,
+      "learning_rate": 7.903576368306659e-06,
+      "loss": 0.2946,
+      "step": 3063
+    },
+    {
+      "epoch": 1.8157037037037038,
+      "grad_norm": 1.8536264896392822,
+      "learning_rate": 7.899624580122505e-06,
+      "loss": 0.3069,
+      "step": 3064
+    },
+    {
+      "epoch": 1.8162962962962963,
+      "grad_norm": 1.9288913011550903,
+      "learning_rate": 7.895672791938353e-06,
+      "loss": 0.1793,
+      "step": 3065
+    },
+    {
+      "epoch": 1.8168888888888888,
+      "grad_norm": 1.6896058320999146,
+      "learning_rate": 7.8917210037542e-06,
+      "loss": 0.2353,
+      "step": 3066
+    },
+    {
+      "epoch": 1.8174814814814815,
+      "grad_norm": 3.8151931762695312,
+      "learning_rate": 7.887769215570045e-06,
+      "loss": 0.3804,
+      "step": 3067
+    },
+    {
+      "epoch": 1.8180740740740742,
+      "grad_norm": 2.535294771194458,
+      "learning_rate": 7.883817427385892e-06,
+      "loss": 0.3888,
+      "step": 3068
+    },
+    {
+      "epoch": 1.8186666666666667,
+      "grad_norm": 2.6704866886138916,
+      "learning_rate": 7.87986563920174e-06,
+      "loss": 0.3357,
+      "step": 3069
+    },
+    {
+      "epoch": 1.8192592592592591,
+      "grad_norm": 2.422725200653076,
+      "learning_rate": 7.875913851017586e-06,
+      "loss": 0.4529,
+      "step": 3070
+    },
+    {
+      "epoch": 1.8198518518518518,
+      "grad_norm": 2.0658416748046875,
+      "learning_rate": 7.871962062833432e-06,
+      "loss": 0.3102,
+      "step": 3071
+    },
+    {
+      "epoch": 1.8204444444444445,
+      "grad_norm": 2.160285472869873,
+      "learning_rate": 7.86801027464928e-06,
+      "loss": 0.2424,
+      "step": 3072
+    },
+    {
+      "epoch": 1.821037037037037,
+      "grad_norm": 1.7413978576660156,
+      "learning_rate": 7.864058486465126e-06,
+      "loss": 0.298,
+      "step": 3073
+    },
+    {
+      "epoch": 1.8216296296296295,
+      "grad_norm": 2.5506606101989746,
+      "learning_rate": 7.860106698280973e-06,
+      "loss": 0.3352,
+      "step": 3074
+    },
+    {
+      "epoch": 1.8222222222222222,
+      "grad_norm": 1.9444527626037598,
+      "learning_rate": 7.856154910096819e-06,
+      "loss": 0.2848,
+      "step": 3075
+    },
+    {
+      "epoch": 1.8228148148148149,
+      "grad_norm": 2.5904300212860107,
+      "learning_rate": 7.852203121912667e-06,
+      "loss": 0.203,
+      "step": 3076
+    },
+    {
+      "epoch": 1.8234074074074074,
+      "grad_norm": 1.9977622032165527,
+      "learning_rate": 7.848251333728513e-06,
+      "loss": 0.2611,
+      "step": 3077
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 2.5425548553466797,
+      "learning_rate": 7.84429954554436e-06,
+      "loss": 0.2118,
+      "step": 3078
+    },
+    {
+      "epoch": 1.8245925925925925,
+      "grad_norm": 1.646548867225647,
+      "learning_rate": 7.840347757360205e-06,
+      "loss": 0.2195,
+      "step": 3079
+    },
+    {
+      "epoch": 1.8251851851851852,
+      "grad_norm": 1.882672667503357,
+      "learning_rate": 7.836395969176053e-06,
+      "loss": 0.2167,
+      "step": 3080
+    },
+    {
+      "epoch": 1.8257777777777777,
+      "grad_norm": 2.3305857181549072,
+      "learning_rate": 7.8324441809919e-06,
+      "loss": 0.3234,
+      "step": 3081
+    },
+    {
+      "epoch": 1.8263703703703704,
+      "grad_norm": 1.9190477132797241,
+      "learning_rate": 7.828492392807746e-06,
+      "loss": 0.2087,
+      "step": 3082
+    },
+    {
+      "epoch": 1.8269629629629631,
+      "grad_norm": 3.120424747467041,
+      "learning_rate": 7.824540604623592e-06,
+      "loss": 0.3501,
+      "step": 3083
+    },
+    {
+      "epoch": 1.8275555555555556,
+      "grad_norm": 2.136350154876709,
+      "learning_rate": 7.82058881643944e-06,
+      "loss": 0.2727,
+      "step": 3084
+    },
+    {
+      "epoch": 1.828148148148148,
+      "grad_norm": 1.5601966381072998,
+      "learning_rate": 7.816637028255286e-06,
+      "loss": 0.1975,
+      "step": 3085
+    },
+    {
+      "epoch": 1.8287407407407408,
+      "grad_norm": 1.7167956829071045,
+      "learning_rate": 7.812685240071133e-06,
+      "loss": 0.3177,
+      "step": 3086
+    },
+    {
+      "epoch": 1.8293333333333335,
+      "grad_norm": 2.292818784713745,
+      "learning_rate": 7.808733451886979e-06,
+      "loss": 0.3958,
+      "step": 3087
+    },
+    {
+      "epoch": 1.829925925925926,
+      "grad_norm": 2.4163646697998047,
+      "learning_rate": 7.804781663702827e-06,
+      "loss": 0.2857,
+      "step": 3088
+    },
+    {
+      "epoch": 1.8305185185185184,
+      "grad_norm": 1.9717518091201782,
+      "learning_rate": 7.800829875518673e-06,
+      "loss": 0.2332,
+      "step": 3089
+    },
+    {
+      "epoch": 1.8311111111111111,
+      "grad_norm": 2.319563627243042,
+      "learning_rate": 7.79687808733452e-06,
+      "loss": 0.2316,
+      "step": 3090
+    },
+    {
+      "epoch": 1.8317037037037038,
+      "grad_norm": 2.0608410835266113,
+      "learning_rate": 7.792926299150365e-06,
+      "loss": 0.2997,
+      "step": 3091
+    },
+    {
+      "epoch": 1.8322962962962963,
+      "grad_norm": 1.861037254333496,
+      "learning_rate": 7.788974510966213e-06,
+      "loss": 0.2207,
+      "step": 3092
+    },
+    {
+      "epoch": 1.8328888888888888,
+      "grad_norm": 2.449862241744995,
+      "learning_rate": 7.78502272278206e-06,
+      "loss": 0.4149,
+      "step": 3093
+    },
+    {
+      "epoch": 1.8334814814814815,
+      "grad_norm": 2.868720293045044,
+      "learning_rate": 7.781070934597906e-06,
+      "loss": 0.3811,
+      "step": 3094
+    },
+    {
+      "epoch": 1.8340740740740742,
+      "grad_norm": 2.0618433952331543,
+      "learning_rate": 7.777119146413754e-06,
+      "loss": 0.285,
+      "step": 3095
+    },
+    {
+      "epoch": 1.8346666666666667,
+      "grad_norm": 1.782892107963562,
+      "learning_rate": 7.7731673582296e-06,
+      "loss": 0.2338,
+      "step": 3096
+    },
+    {
+      "epoch": 1.8352592592592591,
+      "grad_norm": 2.015773296356201,
+      "learning_rate": 7.769215570045446e-06,
+      "loss": 0.3147,
+      "step": 3097
+    },
+    {
+      "epoch": 1.8358518518518518,
+      "grad_norm": 2.3677430152893066,
+      "learning_rate": 7.765263781861293e-06,
+      "loss": 0.344,
+      "step": 3098
+    },
+    {
+      "epoch": 1.8364444444444445,
+      "grad_norm": 2.463979482650757,
+      "learning_rate": 7.76131199367714e-06,
+      "loss": 0.2455,
+      "step": 3099
+    },
+    {
+      "epoch": 1.837037037037037,
+      "grad_norm": 1.666201114654541,
+      "learning_rate": 7.757360205492987e-06,
+      "loss": 0.1993,
+      "step": 3100
+    },
+    {
+      "epoch": 1.8376296296296295,
+      "grad_norm": 1.7282943725585938,
+      "learning_rate": 7.753408417308833e-06,
+      "loss": 0.1705,
+      "step": 3101
+    },
+    {
+      "epoch": 1.8382222222222222,
+      "grad_norm": 1.9877082109451294,
+      "learning_rate": 7.74945662912468e-06,
+      "loss": 0.2961,
+      "step": 3102
+    },
+    {
+      "epoch": 1.838814814814815,
+      "grad_norm": 2.8362503051757812,
+      "learning_rate": 7.745504840940527e-06,
+      "loss": 0.2764,
+      "step": 3103
+    },
+    {
+      "epoch": 1.8394074074074074,
+      "grad_norm": 2.178636312484741,
+      "learning_rate": 7.741553052756373e-06,
+      "loss": 0.2958,
+      "step": 3104
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 2.6513049602508545,
+      "learning_rate": 7.73760126457222e-06,
+      "loss": 0.2846,
+      "step": 3105
+    },
+    {
+      "epoch": 1.8405925925925926,
+      "grad_norm": 2.121211528778076,
+      "learning_rate": 7.733649476388066e-06,
+      "loss": 0.287,
+      "step": 3106
+    },
+    {
+      "epoch": 1.8411851851851853,
+      "grad_norm": 2.263031244277954,
+      "learning_rate": 7.729697688203914e-06,
+      "loss": 0.387,
+      "step": 3107
+    },
+    {
+      "epoch": 1.8417777777777777,
+      "grad_norm": 2.0790531635284424,
+      "learning_rate": 7.72574590001976e-06,
+      "loss": 0.2286,
+      "step": 3108
+    },
+    {
+      "epoch": 1.8423703703703702,
+      "grad_norm": 2.249643325805664,
+      "learning_rate": 7.721794111835606e-06,
+      "loss": 0.2898,
+      "step": 3109
+    },
+    {
+      "epoch": 1.842962962962963,
+      "grad_norm": 2.1978414058685303,
+      "learning_rate": 7.717842323651453e-06,
+      "loss": 0.2767,
+      "step": 3110
+    },
+    {
+      "epoch": 1.8435555555555556,
+      "grad_norm": 1.7232747077941895,
+      "learning_rate": 7.7138905354673e-06,
+      "loss": 0.2394,
+      "step": 3111
+    },
+    {
+      "epoch": 1.844148148148148,
+      "grad_norm": 2.1804685592651367,
+      "learning_rate": 7.709938747283147e-06,
+      "loss": 0.2708,
+      "step": 3112
+    },
+    {
+      "epoch": 1.8447407407407408,
+      "grad_norm": 1.4302172660827637,
+      "learning_rate": 7.705986959098993e-06,
+      "loss": 0.1632,
+      "step": 3113
+    },
+    {
+      "epoch": 1.8453333333333335,
+      "grad_norm": 2.496873140335083,
+      "learning_rate": 7.70203517091484e-06,
+      "loss": 0.2413,
+      "step": 3114
+    },
+    {
+      "epoch": 1.845925925925926,
+      "grad_norm": 1.651871681213379,
+      "learning_rate": 7.698083382730687e-06,
+      "loss": 0.2748,
+      "step": 3115
+    },
+    {
+      "epoch": 1.8465185185185184,
+      "grad_norm": 1.5791614055633545,
+      "learning_rate": 7.694131594546533e-06,
+      "loss": 0.2038,
+      "step": 3116
+    },
+    {
+      "epoch": 1.8471111111111111,
+      "grad_norm": 1.7287096977233887,
+      "learning_rate": 7.69017980636238e-06,
+      "loss": 0.1866,
+      "step": 3117
+    },
+    {
+      "epoch": 1.8477037037037038,
+      "grad_norm": 1.878746747970581,
+      "learning_rate": 7.686228018178226e-06,
+      "loss": 0.2368,
+      "step": 3118
+    },
+    {
+      "epoch": 1.8482962962962963,
+      "grad_norm": 1.690338373184204,
+      "learning_rate": 7.682276229994074e-06,
+      "loss": 0.1544,
+      "step": 3119
+    },
+    {
+      "epoch": 1.8488888888888888,
+      "grad_norm": 1.5626217126846313,
+      "learning_rate": 7.67832444180992e-06,
+      "loss": 0.2428,
+      "step": 3120
+    },
+    {
+      "epoch": 1.8494814814814815,
+      "grad_norm": 2.7668023109436035,
+      "learning_rate": 7.674372653625766e-06,
+      "loss": 0.264,
+      "step": 3121
+    },
+    {
+      "epoch": 1.8500740740740742,
+      "grad_norm": 2.2737486362457275,
+      "learning_rate": 7.670420865441614e-06,
+      "loss": 0.3073,
+      "step": 3122
+    },
+    {
+      "epoch": 1.8506666666666667,
+      "grad_norm": 2.7799124717712402,
+      "learning_rate": 7.66646907725746e-06,
+      "loss": 0.2039,
+      "step": 3123
+    },
+    {
+      "epoch": 1.8512592592592592,
+      "grad_norm": 2.010618209838867,
+      "learning_rate": 7.662517289073307e-06,
+      "loss": 0.2835,
+      "step": 3124
+    },
+    {
+      "epoch": 1.8518518518518519,
+      "grad_norm": 1.5797110795974731,
+      "learning_rate": 7.658565500889153e-06,
+      "loss": 0.1846,
+      "step": 3125
+    },
+    {
+      "epoch": 1.8524444444444446,
+      "grad_norm": 2.606889486312866,
+      "learning_rate": 7.654613712705e-06,
+      "loss": 0.3013,
+      "step": 3126
+    },
+    {
+      "epoch": 1.853037037037037,
+      "grad_norm": 2.273500919342041,
+      "learning_rate": 7.650661924520845e-06,
+      "loss": 0.2743,
+      "step": 3127
+    },
+    {
+      "epoch": 1.8536296296296295,
+      "grad_norm": 1.9333134889602661,
+      "learning_rate": 7.646710136336692e-06,
+      "loss": 0.2081,
+      "step": 3128
+    },
+    {
+      "epoch": 1.8542222222222222,
+      "grad_norm": 2.9338603019714355,
+      "learning_rate": 7.64275834815254e-06,
+      "loss": 0.4575,
+      "step": 3129
+    },
+    {
+      "epoch": 1.854814814814815,
+      "grad_norm": 1.692299723625183,
+      "learning_rate": 7.638806559968386e-06,
+      "loss": 0.2106,
+      "step": 3130
+    },
+    {
+      "epoch": 1.8554074074074074,
+      "grad_norm": 1.556281328201294,
+      "learning_rate": 7.634854771784232e-06,
+      "loss": 0.1877,
+      "step": 3131
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 2.124207019805908,
+      "learning_rate": 7.630902983600078e-06,
+      "loss": 0.2719,
+      "step": 3132
+    },
+    {
+      "epoch": 1.8565925925925926,
+      "grad_norm": 1.4728546142578125,
+      "learning_rate": 7.626951195415926e-06,
+      "loss": 0.1989,
+      "step": 3133
+    },
+    {
+      "epoch": 1.8571851851851853,
+      "grad_norm": 1.6482350826263428,
+      "learning_rate": 7.6229994072317725e-06,
+      "loss": 0.2436,
+      "step": 3134
+    },
+    {
+      "epoch": 1.8577777777777778,
+      "grad_norm": 1.602943778038025,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 0.273,
+      "step": 3135
+    },
+    {
+      "epoch": 1.8583703703703702,
+      "grad_norm": 6.8546462059021,
+      "learning_rate": 7.615095830863466e-06,
+      "loss": 0.3784,
+      "step": 3136
+    },
+    {
+      "epoch": 1.858962962962963,
+      "grad_norm": 1.98646879196167,
+      "learning_rate": 7.611144042679313e-06,
+      "loss": 0.3319,
+      "step": 3137
+    },
+    {
+      "epoch": 1.8595555555555556,
+      "grad_norm": 1.993868112564087,
+      "learning_rate": 7.607192254495159e-06,
+      "loss": 0.3541,
+      "step": 3138
+    },
+    {
+      "epoch": 1.860148148148148,
+      "grad_norm": 2.430436849594116,
+      "learning_rate": 7.603240466311006e-06,
+      "loss": 0.2427,
+      "step": 3139
+    },
+    {
+      "epoch": 1.8607407407407406,
+      "grad_norm": 2.663569211959839,
+      "learning_rate": 7.5992886781268525e-06,
+      "loss": 0.2044,
+      "step": 3140
+    },
+    {
+      "epoch": 1.8613333333333333,
+      "grad_norm": 2.0700650215148926,
+      "learning_rate": 7.5953368899427e-06,
+      "loss": 0.2925,
+      "step": 3141
+    },
+    {
+      "epoch": 1.861925925925926,
+      "grad_norm": 1.8168436288833618,
+      "learning_rate": 7.591385101758546e-06,
+      "loss": 0.2169,
+      "step": 3142
+    },
+    {
+      "epoch": 1.8625185185185185,
+      "grad_norm": 1.7529103755950928,
+      "learning_rate": 7.587433313574393e-06,
+      "loss": 0.2196,
+      "step": 3143
+    },
+    {
+      "epoch": 1.8631111111111112,
+      "grad_norm": 1.8091566562652588,
+      "learning_rate": 7.583481525390239e-06,
+      "loss": 0.2195,
+      "step": 3144
+    },
+    {
+      "epoch": 1.8637037037037039,
+      "grad_norm": 1.2311540842056274,
+      "learning_rate": 7.579529737206086e-06,
+      "loss": 0.1538,
+      "step": 3145
+    },
+    {
+      "epoch": 1.8642962962962963,
+      "grad_norm": 1.730385422706604,
+      "learning_rate": 7.5755779490219325e-06,
+      "loss": 0.2151,
+      "step": 3146
+    },
+    {
+      "epoch": 1.8648888888888888,
+      "grad_norm": 1.6770031452178955,
+      "learning_rate": 7.57162616083778e-06,
+      "loss": 0.2144,
+      "step": 3147
+    },
+    {
+      "epoch": 1.8654814814814815,
+      "grad_norm": 2.8262670040130615,
+      "learning_rate": 7.567674372653626e-06,
+      "loss": 0.2391,
+      "step": 3148
+    },
+    {
+      "epoch": 1.8660740740740742,
+      "grad_norm": 2.0176186561584473,
+      "learning_rate": 7.563722584469473e-06,
+      "loss": 0.2449,
+      "step": 3149
+    },
+    {
+      "epoch": 1.8666666666666667,
+      "grad_norm": 1.906973123550415,
+      "learning_rate": 7.559770796285319e-06,
+      "loss": 0.3139,
+      "step": 3150
+    },
+    {
+      "epoch": 1.8672592592592592,
+      "grad_norm": 3.1402013301849365,
+      "learning_rate": 7.555819008101166e-06,
+      "loss": 0.3625,
+      "step": 3151
+    },
+    {
+      "epoch": 1.8678518518518519,
+      "grad_norm": 2.0794479846954346,
+      "learning_rate": 7.5518672199170125e-06,
+      "loss": 0.2999,
+      "step": 3152
+    },
+    {
+      "epoch": 1.8684444444444446,
+      "grad_norm": 2.1395864486694336,
+      "learning_rate": 7.54791543173286e-06,
+      "loss": 0.2701,
+      "step": 3153
+    },
+    {
+      "epoch": 1.869037037037037,
+      "grad_norm": 1.8182647228240967,
+      "learning_rate": 7.543963643548706e-06,
+      "loss": 0.2428,
+      "step": 3154
+    },
+    {
+      "epoch": 1.8696296296296295,
+      "grad_norm": 1.6160415410995483,
+      "learning_rate": 7.540011855364553e-06,
+      "loss": 0.1922,
+      "step": 3155
+    },
+    {
+      "epoch": 1.8702222222222222,
+      "grad_norm": 2.080928087234497,
+      "learning_rate": 7.536060067180399e-06,
+      "loss": 0.1925,
+      "step": 3156
+    },
+    {
+      "epoch": 1.870814814814815,
+      "grad_norm": 2.1450204849243164,
+      "learning_rate": 7.532108278996246e-06,
+      "loss": 0.2336,
+      "step": 3157
+    },
+    {
+      "epoch": 1.8714074074074074,
+      "grad_norm": 2.0599563121795654,
+      "learning_rate": 7.528156490812093e-06,
+      "loss": 0.2959,
+      "step": 3158
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 1.9121206998825073,
+      "learning_rate": 7.52420470262794e-06,
+      "loss": 0.2093,
+      "step": 3159
+    },
+    {
+      "epoch": 1.8725925925925926,
+      "grad_norm": 2.1517021656036377,
+      "learning_rate": 7.520252914443787e-06,
+      "loss": 0.1909,
+      "step": 3160
+    },
+    {
+      "epoch": 1.8731851851851853,
+      "grad_norm": 1.8783502578735352,
+      "learning_rate": 7.516301126259633e-06,
+      "loss": 0.2566,
+      "step": 3161
+    },
+    {
+      "epoch": 1.8737777777777778,
+      "grad_norm": 2.6619553565979004,
+      "learning_rate": 7.51234933807548e-06,
+      "loss": 0.1869,
+      "step": 3162
+    },
+    {
+      "epoch": 1.8743703703703702,
+      "grad_norm": 1.8074400424957275,
+      "learning_rate": 7.508397549891326e-06,
+      "loss": 0.2105,
+      "step": 3163
+    },
+    {
+      "epoch": 1.874962962962963,
+      "grad_norm": 1.56841242313385,
+      "learning_rate": 7.504445761707173e-06,
+      "loss": 0.2308,
+      "step": 3164
+    },
+    {
+      "epoch": 1.8755555555555556,
+      "grad_norm": 1.9699499607086182,
+      "learning_rate": 7.50049397352302e-06,
+      "loss": 0.2029,
+      "step": 3165
+    },
+    {
+      "epoch": 1.8761481481481481,
+      "grad_norm": 2.7146174907684326,
+      "learning_rate": 7.496542185338867e-06,
+      "loss": 0.2103,
+      "step": 3166
+    },
+    {
+      "epoch": 1.8767407407407406,
+      "grad_norm": 2.252864360809326,
+      "learning_rate": 7.492590397154713e-06,
+      "loss": 0.2665,
+      "step": 3167
+    },
+    {
+      "epoch": 1.8773333333333333,
+      "grad_norm": 2.275169610977173,
+      "learning_rate": 7.48863860897056e-06,
+      "loss": 0.2134,
+      "step": 3168
+    },
+    {
+      "epoch": 1.877925925925926,
+      "grad_norm": 1.988403081893921,
+      "learning_rate": 7.484686820786406e-06,
+      "loss": 0.2305,
+      "step": 3169
+    },
+    {
+      "epoch": 1.8785185185185185,
+      "grad_norm": 2.308802366256714,
+      "learning_rate": 7.480735032602253e-06,
+      "loss": 0.2692,
+      "step": 3170
+    },
+    {
+      "epoch": 1.879111111111111,
+      "grad_norm": 2.0073328018188477,
+      "learning_rate": 7.4767832444181e-06,
+      "loss": 0.2551,
+      "step": 3171
+    },
+    {
+      "epoch": 1.8797037037037037,
+      "grad_norm": 3.1714236736297607,
+      "learning_rate": 7.472831456233947e-06,
+      "loss": 0.4193,
+      "step": 3172
+    },
+    {
+      "epoch": 1.8802962962962964,
+      "grad_norm": 2.277350664138794,
+      "learning_rate": 7.468879668049793e-06,
+      "loss": 0.2907,
+      "step": 3173
+    },
+    {
+      "epoch": 1.8808888888888888,
+      "grad_norm": 1.9094727039337158,
+      "learning_rate": 7.46492787986564e-06,
+      "loss": 0.2113,
+      "step": 3174
+    },
+    {
+      "epoch": 1.8814814814814815,
+      "grad_norm": 2.2297523021698,
+      "learning_rate": 7.460976091681486e-06,
+      "loss": 0.207,
+      "step": 3175
+    },
+    {
+      "epoch": 1.8820740740740742,
+      "grad_norm": 2.254145860671997,
+      "learning_rate": 7.457024303497333e-06,
+      "loss": 0.3136,
+      "step": 3176
+    },
+    {
+      "epoch": 1.8826666666666667,
+      "grad_norm": 2.007532835006714,
+      "learning_rate": 7.45307251531318e-06,
+      "loss": 0.3019,
+      "step": 3177
+    },
+    {
+      "epoch": 1.8832592592592592,
+      "grad_norm": 2.180910348892212,
+      "learning_rate": 7.449120727129027e-06,
+      "loss": 0.2259,
+      "step": 3178
+    },
+    {
+      "epoch": 1.8838518518518519,
+      "grad_norm": 1.6106408834457397,
+      "learning_rate": 7.445168938944873e-06,
+      "loss": 0.2677,
+      "step": 3179
+    },
+    {
+      "epoch": 1.8844444444444446,
+      "grad_norm": 2.197611093521118,
+      "learning_rate": 7.44121715076072e-06,
+      "loss": 0.3735,
+      "step": 3180
+    },
+    {
+      "epoch": 1.885037037037037,
+      "grad_norm": 1.5770924091339111,
+      "learning_rate": 7.437265362576567e-06,
+      "loss": 0.2025,
+      "step": 3181
+    },
+    {
+      "epoch": 1.8856296296296295,
+      "grad_norm": 1.9159226417541504,
+      "learning_rate": 7.433313574392413e-06,
+      "loss": 0.2485,
+      "step": 3182
+    },
+    {
+      "epoch": 1.8862222222222222,
+      "grad_norm": 1.9739538431167603,
+      "learning_rate": 7.4293617862082605e-06,
+      "loss": 0.2313,
+      "step": 3183
+    },
+    {
+      "epoch": 1.886814814814815,
+      "grad_norm": 2.495237350463867,
+      "learning_rate": 7.425409998024107e-06,
+      "loss": 0.2883,
+      "step": 3184
+    },
+    {
+      "epoch": 1.8874074074074074,
+      "grad_norm": 2.4452502727508545,
+      "learning_rate": 7.421458209839954e-06,
+      "loss": 0.2941,
+      "step": 3185
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 1.783265233039856,
+      "learning_rate": 7.4175064216558e-06,
+      "loss": 0.3264,
+      "step": 3186
+    },
+    {
+      "epoch": 1.8885925925925926,
+      "grad_norm": 1.7662361860275269,
+      "learning_rate": 7.413554633471647e-06,
+      "loss": 0.2329,
+      "step": 3187
+    },
+    {
+      "epoch": 1.8891851851851853,
+      "grad_norm": 1.621607780456543,
+      "learning_rate": 7.409602845287493e-06,
+      "loss": 0.175,
+      "step": 3188
+    },
+    {
+      "epoch": 1.8897777777777778,
+      "grad_norm": 2.1540908813476562,
+      "learning_rate": 7.4056510571033405e-06,
+      "loss": 0.2868,
+      "step": 3189
+    },
+    {
+      "epoch": 1.8903703703703703,
+      "grad_norm": 2.549497604370117,
+      "learning_rate": 7.401699268919187e-06,
+      "loss": 0.2992,
+      "step": 3190
+    },
+    {
+      "epoch": 1.890962962962963,
+      "grad_norm": 1.7673581838607788,
+      "learning_rate": 7.397747480735034e-06,
+      "loss": 0.2332,
+      "step": 3191
+    },
+    {
+      "epoch": 1.8915555555555557,
+      "grad_norm": 2.001246929168701,
+      "learning_rate": 7.39379569255088e-06,
+      "loss": 0.3436,
+      "step": 3192
+    },
+    {
+      "epoch": 1.8921481481481481,
+      "grad_norm": 1.7078129053115845,
+      "learning_rate": 7.389843904366727e-06,
+      "loss": 0.2066,
+      "step": 3193
+    },
+    {
+      "epoch": 1.8927407407407406,
+      "grad_norm": 2.023287057876587,
+      "learning_rate": 7.385892116182573e-06,
+      "loss": 0.2984,
+      "step": 3194
+    },
+    {
+      "epoch": 1.8933333333333333,
+      "grad_norm": 2.131103515625,
+      "learning_rate": 7.3819403279984205e-06,
+      "loss": 0.3051,
+      "step": 3195
+    },
+    {
+      "epoch": 1.893925925925926,
+      "grad_norm": 2.5095415115356445,
+      "learning_rate": 7.377988539814267e-06,
+      "loss": 0.2375,
+      "step": 3196
+    },
+    {
+      "epoch": 1.8945185185185185,
+      "grad_norm": 2.5038113594055176,
+      "learning_rate": 7.374036751630114e-06,
+      "loss": 0.2714,
+      "step": 3197
+    },
+    {
+      "epoch": 1.895111111111111,
+      "grad_norm": 1.494916319847107,
+      "learning_rate": 7.37008496344596e-06,
+      "loss": 0.2033,
+      "step": 3198
+    },
+    {
+      "epoch": 1.8957037037037037,
+      "grad_norm": 1.963660478591919,
+      "learning_rate": 7.366133175261807e-06,
+      "loss": 0.2171,
+      "step": 3199
+    },
+    {
+      "epoch": 1.8962962962962964,
+      "grad_norm": 2.034064292907715,
+      "learning_rate": 7.362181387077653e-06,
+      "loss": 0.2058,
+      "step": 3200
+    },
+    {
+      "epoch": 1.8968888888888888,
+      "grad_norm": 1.7918635606765747,
+      "learning_rate": 7.3582295988935005e-06,
+      "loss": 0.2241,
+      "step": 3201
+    },
+    {
+      "epoch": 1.8974814814814813,
+      "grad_norm": 2.3345847129821777,
+      "learning_rate": 7.354277810709347e-06,
+      "loss": 0.3142,
+      "step": 3202
+    },
+    {
+      "epoch": 1.898074074074074,
+      "grad_norm": 3.9282302856445312,
+      "learning_rate": 7.350326022525194e-06,
+      "loss": 0.3898,
+      "step": 3203
+    },
+    {
+      "epoch": 1.8986666666666667,
+      "grad_norm": 2.1323158740997314,
+      "learning_rate": 7.346374234341041e-06,
+      "loss": 0.1638,
+      "step": 3204
+    },
+    {
+      "epoch": 1.8992592592592592,
+      "grad_norm": 1.9272449016571045,
+      "learning_rate": 7.342422446156886e-06,
+      "loss": 0.1743,
+      "step": 3205
+    },
+    {
+      "epoch": 1.899851851851852,
+      "grad_norm": 3.492011785507202,
+      "learning_rate": 7.3384706579727325e-06,
+      "loss": 0.3913,
+      "step": 3206
+    },
+    {
+      "epoch": 1.9004444444444446,
+      "grad_norm": 1.9495962858200073,
+      "learning_rate": 7.33451886978858e-06,
+      "loss": 0.215,
+      "step": 3207
+    },
+    {
+      "epoch": 1.901037037037037,
+      "grad_norm": 1.273066759109497,
+      "learning_rate": 7.330567081604426e-06,
+      "loss": 0.1477,
+      "step": 3208
+    },
+    {
+      "epoch": 1.9016296296296296,
+      "grad_norm": 2.314990520477295,
+      "learning_rate": 7.326615293420273e-06,
+      "loss": 0.2245,
+      "step": 3209
+    },
+    {
+      "epoch": 1.9022222222222223,
+      "grad_norm": 3.7168281078338623,
+      "learning_rate": 7.322663505236119e-06,
+      "loss": 0.3633,
+      "step": 3210
+    },
+    {
+      "epoch": 1.902814814814815,
+      "grad_norm": 2.4523229598999023,
+      "learning_rate": 7.318711717051966e-06,
+      "loss": 0.2895,
+      "step": 3211
+    },
+    {
+      "epoch": 1.9034074074074074,
+      "grad_norm": 1.851425290107727,
+      "learning_rate": 7.3147599288678125e-06,
+      "loss": 0.2662,
+      "step": 3212
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 2.2717738151550293,
+      "learning_rate": 7.31080814068366e-06,
+      "loss": 0.277,
+      "step": 3213
+    },
+    {
+      "epoch": 1.9045925925925926,
+      "grad_norm": 1.7864848375320435,
+      "learning_rate": 7.306856352499506e-06,
+      "loss": 0.2004,
+      "step": 3214
+    },
+    {
+      "epoch": 1.9051851851851853,
+      "grad_norm": 1.9599438905715942,
+      "learning_rate": 7.302904564315353e-06,
+      "loss": 0.2237,
+      "step": 3215
+    },
+    {
+      "epoch": 1.9057777777777778,
+      "grad_norm": 2.356327533721924,
+      "learning_rate": 7.298952776131199e-06,
+      "loss": 0.2447,
+      "step": 3216
+    },
+    {
+      "epoch": 1.9063703703703703,
+      "grad_norm": 1.4188939332962036,
+      "learning_rate": 7.295000987947046e-06,
+      "loss": 0.2062,
+      "step": 3217
+    },
+    {
+      "epoch": 1.906962962962963,
+      "grad_norm": 2.880932569503784,
+      "learning_rate": 7.2910491997628925e-06,
+      "loss": 0.4569,
+      "step": 3218
+    },
+    {
+      "epoch": 1.9075555555555557,
+      "grad_norm": 2.1558680534362793,
+      "learning_rate": 7.28709741157874e-06,
+      "loss": 0.2359,
+      "step": 3219
+    },
+    {
+      "epoch": 1.9081481481481481,
+      "grad_norm": 1.6568809747695923,
+      "learning_rate": 7.283145623394587e-06,
+      "loss": 0.2409,
+      "step": 3220
+    },
+    {
+      "epoch": 1.9087407407407406,
+      "grad_norm": 1.8688892126083374,
+      "learning_rate": 7.279193835210433e-06,
+      "loss": 0.1948,
+      "step": 3221
+    },
+    {
+      "epoch": 1.9093333333333333,
+      "grad_norm": 1.7872833013534546,
+      "learning_rate": 7.27524204702628e-06,
+      "loss": 0.2457,
+      "step": 3222
+    },
+    {
+      "epoch": 1.909925925925926,
+      "grad_norm": 2.129321813583374,
+      "learning_rate": 7.271290258842126e-06,
+      "loss": 0.2559,
+      "step": 3223
+    },
+    {
+      "epoch": 1.9105185185185185,
+      "grad_norm": 1.8255162239074707,
+      "learning_rate": 7.267338470657973e-06,
+      "loss": 0.2074,
+      "step": 3224
+    },
+    {
+      "epoch": 1.911111111111111,
+      "grad_norm": 1.73768150806427,
+      "learning_rate": 7.26338668247382e-06,
+      "loss": 0.2118,
+      "step": 3225
+    },
+    {
+      "epoch": 1.9117037037037037,
+      "grad_norm": 2.7301838397979736,
+      "learning_rate": 7.259434894289667e-06,
+      "loss": 0.3224,
+      "step": 3226
+    },
+    {
+      "epoch": 1.9122962962962964,
+      "grad_norm": 1.82205331325531,
+      "learning_rate": 7.255483106105513e-06,
+      "loss": 0.2519,
+      "step": 3227
+    },
+    {
+      "epoch": 1.9128888888888889,
+      "grad_norm": 2.192042350769043,
+      "learning_rate": 7.25153131792136e-06,
+      "loss": 0.2865,
+      "step": 3228
+    },
+    {
+      "epoch": 1.9134814814814813,
+      "grad_norm": 1.5194973945617676,
+      "learning_rate": 7.247579529737206e-06,
+      "loss": 0.1734,
+      "step": 3229
+    },
+    {
+      "epoch": 1.914074074074074,
+      "grad_norm": 4.335606575012207,
+      "learning_rate": 7.243627741553053e-06,
+      "loss": 0.5072,
+      "step": 3230
+    },
+    {
+      "epoch": 1.9146666666666667,
+      "grad_norm": 2.747256278991699,
+      "learning_rate": 7.2396759533689e-06,
+      "loss": 0.3095,
+      "step": 3231
+    },
+    {
+      "epoch": 1.9152592592592592,
+      "grad_norm": 2.2774157524108887,
+      "learning_rate": 7.235724165184747e-06,
+      "loss": 0.2552,
+      "step": 3232
+    },
+    {
+      "epoch": 1.9158518518518517,
+      "grad_norm": 3.083685874938965,
+      "learning_rate": 7.231772377000593e-06,
+      "loss": 0.2684,
+      "step": 3233
+    },
+    {
+      "epoch": 1.9164444444444444,
+      "grad_norm": 2.081791639328003,
+      "learning_rate": 7.22782058881644e-06,
+      "loss": 0.3594,
+      "step": 3234
+    },
+    {
+      "epoch": 1.917037037037037,
+      "grad_norm": 3.0360615253448486,
+      "learning_rate": 7.223868800632286e-06,
+      "loss": 0.3441,
+      "step": 3235
+    },
+    {
+      "epoch": 1.9176296296296296,
+      "grad_norm": 2.5803136825561523,
+      "learning_rate": 7.219917012448133e-06,
+      "loss": 0.342,
+      "step": 3236
+    },
+    {
+      "epoch": 1.9182222222222223,
+      "grad_norm": 2.043827772140503,
+      "learning_rate": 7.21596522426398e-06,
+      "loss": 0.2169,
+      "step": 3237
+    },
+    {
+      "epoch": 1.918814814814815,
+      "grad_norm": 2.393888235092163,
+      "learning_rate": 7.212013436079827e-06,
+      "loss": 0.2348,
+      "step": 3238
+    },
+    {
+      "epoch": 1.9194074074074075,
+      "grad_norm": 2.3425092697143555,
+      "learning_rate": 7.208061647895673e-06,
+      "loss": 0.2119,
+      "step": 3239
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.124314546585083,
+      "learning_rate": 7.20410985971152e-06,
+      "loss": 0.343,
+      "step": 3240
+    },
+    {
+      "epoch": 1.9205925925925926,
+      "grad_norm": 2.527470588684082,
+      "learning_rate": 7.200158071527366e-06,
+      "loss": 0.3305,
+      "step": 3241
+    },
+    {
+      "epoch": 1.9211851851851853,
+      "grad_norm": 2.0635364055633545,
+      "learning_rate": 7.196206283343213e-06,
+      "loss": 0.3711,
+      "step": 3242
+    },
+    {
+      "epoch": 1.9217777777777778,
+      "grad_norm": 1.6901216506958008,
+      "learning_rate": 7.1922544951590604e-06,
+      "loss": 0.205,
+      "step": 3243
+    },
+    {
+      "epoch": 1.9223703703703703,
+      "grad_norm": 1.889601469039917,
+      "learning_rate": 7.188302706974907e-06,
+      "loss": 0.2049,
+      "step": 3244
+    },
+    {
+      "epoch": 1.922962962962963,
+      "grad_norm": 2.1268372535705566,
+      "learning_rate": 7.184350918790754e-06,
+      "loss": 0.2204,
+      "step": 3245
+    },
+    {
+      "epoch": 1.9235555555555557,
+      "grad_norm": 2.1181466579437256,
+      "learning_rate": 7.1803991306066e-06,
+      "loss": 0.3012,
+      "step": 3246
+    },
+    {
+      "epoch": 1.9241481481481482,
+      "grad_norm": 2.1675190925598145,
+      "learning_rate": 7.176447342422447e-06,
+      "loss": 0.2255,
+      "step": 3247
+    },
+    {
+      "epoch": 1.9247407407407406,
+      "grad_norm": 2.195160388946533,
+      "learning_rate": 7.172495554238293e-06,
+      "loss": 0.2311,
+      "step": 3248
+    },
+    {
+      "epoch": 1.9253333333333333,
+      "grad_norm": 1.3711282014846802,
+      "learning_rate": 7.1685437660541404e-06,
+      "loss": 0.1386,
+      "step": 3249
+    },
+    {
+      "epoch": 1.925925925925926,
+      "grad_norm": 2.30734920501709,
+      "learning_rate": 7.164591977869987e-06,
+      "loss": 0.2146,
+      "step": 3250
+    },
+    {
+      "epoch": 1.9265185185185185,
+      "grad_norm": 1.6893199682235718,
+      "learning_rate": 7.160640189685834e-06,
+      "loss": 0.2193,
+      "step": 3251
+    },
+    {
+      "epoch": 1.927111111111111,
+      "grad_norm": 1.9613697528839111,
+      "learning_rate": 7.15668840150168e-06,
+      "loss": 0.2158,
+      "step": 3252
+    },
+    {
+      "epoch": 1.9277037037037037,
+      "grad_norm": 1.562820553779602,
+      "learning_rate": 7.152736613317527e-06,
+      "loss": 0.2066,
+      "step": 3253
+    },
+    {
+      "epoch": 1.9282962962962964,
+      "grad_norm": 1.9168422222137451,
+      "learning_rate": 7.148784825133373e-06,
+      "loss": 0.2759,
+      "step": 3254
+    },
+    {
+      "epoch": 1.9288888888888889,
+      "grad_norm": 2.5176687240600586,
+      "learning_rate": 7.1448330369492204e-06,
+      "loss": 0.2624,
+      "step": 3255
+    },
+    {
+      "epoch": 1.9294814814814814,
+      "grad_norm": 1.7413691282272339,
+      "learning_rate": 7.140881248765067e-06,
+      "loss": 0.2022,
+      "step": 3256
+    },
+    {
+      "epoch": 1.930074074074074,
+      "grad_norm": 1.8298150300979614,
+      "learning_rate": 7.136929460580914e-06,
+      "loss": 0.2436,
+      "step": 3257
+    },
+    {
+      "epoch": 1.9306666666666668,
+      "grad_norm": 1.8024886846542358,
+      "learning_rate": 7.13297767239676e-06,
+      "loss": 0.2219,
+      "step": 3258
+    },
+    {
+      "epoch": 1.9312592592592592,
+      "grad_norm": 1.5105679035186768,
+      "learning_rate": 7.129025884212607e-06,
+      "loss": 0.2572,
+      "step": 3259
+    },
+    {
+      "epoch": 1.9318518518518517,
+      "grad_norm": 3.3013293743133545,
+      "learning_rate": 7.125074096028453e-06,
+      "loss": 0.5589,
+      "step": 3260
+    },
+    {
+      "epoch": 1.9324444444444444,
+      "grad_norm": 1.7074873447418213,
+      "learning_rate": 7.1211223078443004e-06,
+      "loss": 0.1914,
+      "step": 3261
+    },
+    {
+      "epoch": 1.933037037037037,
+      "grad_norm": 2.2846224308013916,
+      "learning_rate": 7.117170519660147e-06,
+      "loss": 0.2314,
+      "step": 3262
+    },
+    {
+      "epoch": 1.9336296296296296,
+      "grad_norm": 3.144522190093994,
+      "learning_rate": 7.113218731475994e-06,
+      "loss": 0.3474,
+      "step": 3263
+    },
+    {
+      "epoch": 1.934222222222222,
+      "grad_norm": 2.5165317058563232,
+      "learning_rate": 7.10926694329184e-06,
+      "loss": 0.3037,
+      "step": 3264
+    },
+    {
+      "epoch": 1.9348148148148148,
+      "grad_norm": 2.579094886779785,
+      "learning_rate": 7.105315155107687e-06,
+      "loss": 0.3411,
+      "step": 3265
+    },
+    {
+      "epoch": 1.9354074074074075,
+      "grad_norm": 2.0858352184295654,
+      "learning_rate": 7.101363366923534e-06,
+      "loss": 0.3083,
+      "step": 3266
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 2.8584647178649902,
+      "learning_rate": 7.0974115787393804e-06,
+      "loss": 0.3357,
+      "step": 3267
+    },
+    {
+      "epoch": 1.9365925925925926,
+      "grad_norm": 1.7452939748764038,
+      "learning_rate": 7.0934597905552275e-06,
+      "loss": 0.1934,
+      "step": 3268
+    },
+    {
+      "epoch": 1.9371851851851853,
+      "grad_norm": 3.208458662033081,
+      "learning_rate": 7.089508002371074e-06,
+      "loss": 0.2375,
+      "step": 3269
+    },
+    {
+      "epoch": 1.9377777777777778,
+      "grad_norm": 2.5783629417419434,
+      "learning_rate": 7.085556214186921e-06,
+      "loss": 0.4083,
+      "step": 3270
+    },
+    {
+      "epoch": 1.9383703703703703,
+      "grad_norm": 1.6727075576782227,
+      "learning_rate": 7.081604426002767e-06,
+      "loss": 0.2773,
+      "step": 3271
+    },
+    {
+      "epoch": 1.938962962962963,
+      "grad_norm": 1.9146356582641602,
+      "learning_rate": 7.077652637818614e-06,
+      "loss": 0.2027,
+      "step": 3272
+    },
+    {
+      "epoch": 1.9395555555555557,
+      "grad_norm": 2.5416975021362305,
+      "learning_rate": 7.0737008496344604e-06,
+      "loss": 0.3133,
+      "step": 3273
+    },
+    {
+      "epoch": 1.9401481481481482,
+      "grad_norm": 2.209669351577759,
+      "learning_rate": 7.0697490614503075e-06,
+      "loss": 0.305,
+      "step": 3274
+    },
+    {
+      "epoch": 1.9407407407407407,
+      "grad_norm": 2.0305111408233643,
+      "learning_rate": 7.065797273266154e-06,
+      "loss": 0.2273,
+      "step": 3275
+    },
+    {
+      "epoch": 1.9413333333333334,
+      "grad_norm": 2.065609931945801,
+      "learning_rate": 7.061845485082001e-06,
+      "loss": 0.3042,
+      "step": 3276
+    },
+    {
+      "epoch": 1.941925925925926,
+      "grad_norm": 1.9937634468078613,
+      "learning_rate": 7.057893696897847e-06,
+      "loss": 0.2213,
+      "step": 3277
+    },
+    {
+      "epoch": 1.9425185185185185,
+      "grad_norm": 2.0854732990264893,
+      "learning_rate": 7.053941908713694e-06,
+      "loss": 0.3141,
+      "step": 3278
+    },
+    {
+      "epoch": 1.943111111111111,
+      "grad_norm": 1.752682089805603,
+      "learning_rate": 7.0499901205295404e-06,
+      "loss": 0.2607,
+      "step": 3279
+    },
+    {
+      "epoch": 1.9437037037037037,
+      "grad_norm": 1.705073356628418,
+      "learning_rate": 7.0460383323453875e-06,
+      "loss": 0.2294,
+      "step": 3280
+    },
+    {
+      "epoch": 1.9442962962962964,
+      "grad_norm": 2.3236474990844727,
+      "learning_rate": 7.042086544161234e-06,
+      "loss": 0.2869,
+      "step": 3281
+    },
+    {
+      "epoch": 1.944888888888889,
+      "grad_norm": 3.3840057849884033,
+      "learning_rate": 7.038134755977081e-06,
+      "loss": 0.3866,
+      "step": 3282
+    },
+    {
+      "epoch": 1.9454814814814814,
+      "grad_norm": 2.348243236541748,
+      "learning_rate": 7.034182967792927e-06,
+      "loss": 0.2746,
+      "step": 3283
+    },
+    {
+      "epoch": 1.946074074074074,
+      "grad_norm": 2.397885799407959,
+      "learning_rate": 7.030231179608773e-06,
+      "loss": 0.2967,
+      "step": 3284
+    },
+    {
+      "epoch": 1.9466666666666668,
+      "grad_norm": 1.8983908891677856,
+      "learning_rate": 7.02627939142462e-06,
+      "loss": 0.1863,
+      "step": 3285
+    },
+    {
+      "epoch": 1.9472592592592592,
+      "grad_norm": 2.181011199951172,
+      "learning_rate": 7.022327603240467e-06,
+      "loss": 0.2805,
+      "step": 3286
+    },
+    {
+      "epoch": 1.9478518518518517,
+      "grad_norm": 1.9888044595718384,
+      "learning_rate": 7.018375815056313e-06,
+      "loss": 0.2661,
+      "step": 3287
+    },
+    {
+      "epoch": 1.9484444444444444,
+      "grad_norm": 1.8151829242706299,
+      "learning_rate": 7.01442402687216e-06,
+      "loss": 0.2529,
+      "step": 3288
+    },
+    {
+      "epoch": 1.9490370370370371,
+      "grad_norm": 2.842963933944702,
+      "learning_rate": 7.010472238688006e-06,
+      "loss": 0.2623,
+      "step": 3289
+    },
+    {
+      "epoch": 1.9496296296296296,
+      "grad_norm": 2.9511022567749023,
+      "learning_rate": 7.006520450503853e-06,
+      "loss": 0.2878,
+      "step": 3290
+    },
+    {
+      "epoch": 1.950222222222222,
+      "grad_norm": 2.8291542530059814,
+      "learning_rate": 7.0025686623197e-06,
+      "loss": 0.2297,
+      "step": 3291
+    },
+    {
+      "epoch": 1.9508148148148148,
+      "grad_norm": 1.9667491912841797,
+      "learning_rate": 6.998616874135547e-06,
+      "loss": 0.1982,
+      "step": 3292
+    },
+    {
+      "epoch": 1.9514074074074075,
+      "grad_norm": 2.065978527069092,
+      "learning_rate": 6.994665085951393e-06,
+      "loss": 0.3272,
+      "step": 3293
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 1.720990538597107,
+      "learning_rate": 6.99071329776724e-06,
+      "loss": 0.2888,
+      "step": 3294
+    },
+    {
+      "epoch": 1.9525925925925924,
+      "grad_norm": 2.2570486068725586,
+      "learning_rate": 6.986761509583086e-06,
+      "loss": 0.3179,
+      "step": 3295
+    },
+    {
+      "epoch": 1.9531851851851851,
+      "grad_norm": 2.3585023880004883,
+      "learning_rate": 6.982809721398933e-06,
+      "loss": 0.3484,
+      "step": 3296
+    },
+    {
+      "epoch": 1.9537777777777778,
+      "grad_norm": 1.8689119815826416,
+      "learning_rate": 6.97885793321478e-06,
+      "loss": 0.2432,
+      "step": 3297
+    },
+    {
+      "epoch": 1.9543703703703703,
+      "grad_norm": 1.891492247581482,
+      "learning_rate": 6.974906145030627e-06,
+      "loss": 0.2022,
+      "step": 3298
+    },
+    {
+      "epoch": 1.954962962962963,
+      "grad_norm": 1.9671682119369507,
+      "learning_rate": 6.970954356846473e-06,
+      "loss": 0.2477,
+      "step": 3299
+    },
+    {
+      "epoch": 1.9555555555555557,
+      "grad_norm": 1.5787416696548462,
+      "learning_rate": 6.96700256866232e-06,
+      "loss": 0.3005,
+      "step": 3300
+    },
+    {
+      "epoch": 1.9561481481481482,
+      "grad_norm": 2.3343746662139893,
+      "learning_rate": 6.963050780478166e-06,
+      "loss": 0.2798,
+      "step": 3301
+    },
+    {
+      "epoch": 1.9567407407407407,
+      "grad_norm": 1.6978679895401,
+      "learning_rate": 6.959098992294013e-06,
+      "loss": 0.2751,
+      "step": 3302
+    },
+    {
+      "epoch": 1.9573333333333334,
+      "grad_norm": 2.062009334564209,
+      "learning_rate": 6.95514720410986e-06,
+      "loss": 0.3022,
+      "step": 3303
+    },
+    {
+      "epoch": 1.957925925925926,
+      "grad_norm": 1.6305882930755615,
+      "learning_rate": 6.951195415925707e-06,
+      "loss": 0.2355,
+      "step": 3304
+    },
+    {
+      "epoch": 1.9585185185185185,
+      "grad_norm": 1.4648418426513672,
+      "learning_rate": 6.947243627741554e-06,
+      "loss": 0.1886,
+      "step": 3305
+    },
+    {
+      "epoch": 1.959111111111111,
+      "grad_norm": 2.4117002487182617,
+      "learning_rate": 6.9432918395574e-06,
+      "loss": 0.2668,
+      "step": 3306
+    },
+    {
+      "epoch": 1.9597037037037037,
+      "grad_norm": 2.1643741130828857,
+      "learning_rate": 6.939340051373247e-06,
+      "loss": 0.2489,
+      "step": 3307
+    },
+    {
+      "epoch": 1.9602962962962964,
+      "grad_norm": 1.5768386125564575,
+      "learning_rate": 6.935388263189093e-06,
+      "loss": 0.2826,
+      "step": 3308
+    },
+    {
+      "epoch": 1.960888888888889,
+      "grad_norm": 1.3777393102645874,
+      "learning_rate": 6.9314364750049404e-06,
+      "loss": 0.2093,
+      "step": 3309
+    },
+    {
+      "epoch": 1.9614814814814814,
+      "grad_norm": 2.020051956176758,
+      "learning_rate": 6.927484686820787e-06,
+      "loss": 0.2642,
+      "step": 3310
+    },
+    {
+      "epoch": 1.962074074074074,
+      "grad_norm": 1.6546986103057861,
+      "learning_rate": 6.923532898636634e-06,
+      "loss": 0.275,
+      "step": 3311
+    },
+    {
+      "epoch": 1.9626666666666668,
+      "grad_norm": 1.928969144821167,
+      "learning_rate": 6.91958111045248e-06,
+      "loss": 0.2528,
+      "step": 3312
+    },
+    {
+      "epoch": 1.9632592592592593,
+      "grad_norm": 2.4591543674468994,
+      "learning_rate": 6.915629322268327e-06,
+      "loss": 0.2814,
+      "step": 3313
+    },
+    {
+      "epoch": 1.9638518518518517,
+      "grad_norm": 2.430290937423706,
+      "learning_rate": 6.911677534084173e-06,
+      "loss": 0.2408,
+      "step": 3314
+    },
+    {
+      "epoch": 1.9644444444444444,
+      "grad_norm": 2.3361639976501465,
+      "learning_rate": 6.90772574590002e-06,
+      "loss": 0.3644,
+      "step": 3315
+    },
+    {
+      "epoch": 1.9650370370370371,
+      "grad_norm": 2.4558303356170654,
+      "learning_rate": 6.903773957715867e-06,
+      "loss": 0.2551,
+      "step": 3316
+    },
+    {
+      "epoch": 1.9656296296296296,
+      "grad_norm": 2.1182024478912354,
+      "learning_rate": 6.899822169531714e-06,
+      "loss": 0.2797,
+      "step": 3317
+    },
+    {
+      "epoch": 1.966222222222222,
+      "grad_norm": 2.830535411834717,
+      "learning_rate": 6.89587038134756e-06,
+      "loss": 0.3895,
+      "step": 3318
+    },
+    {
+      "epoch": 1.9668148148148148,
+      "grad_norm": 2.6123626232147217,
+      "learning_rate": 6.891918593163407e-06,
+      "loss": 0.3314,
+      "step": 3319
+    },
+    {
+      "epoch": 1.9674074074074075,
+      "grad_norm": 2.4812047481536865,
+      "learning_rate": 6.887966804979253e-06,
+      "loss": 0.2279,
+      "step": 3320
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 1.8006707429885864,
+      "learning_rate": 6.8840150167951e-06,
+      "loss": 0.1674,
+      "step": 3321
+    },
+    {
+      "epoch": 1.9685925925925924,
+      "grad_norm": 1.8659894466400146,
+      "learning_rate": 6.880063228610947e-06,
+      "loss": 0.2362,
+      "step": 3322
+    },
+    {
+      "epoch": 1.9691851851851851,
+      "grad_norm": 3.3986592292785645,
+      "learning_rate": 6.876111440426794e-06,
+      "loss": 0.2372,
+      "step": 3323
+    },
+    {
+      "epoch": 1.9697777777777778,
+      "grad_norm": 2.287212371826172,
+      "learning_rate": 6.87215965224264e-06,
+      "loss": 0.2061,
+      "step": 3324
+    },
+    {
+      "epoch": 1.9703703703703703,
+      "grad_norm": 2.33723521232605,
+      "learning_rate": 6.868207864058487e-06,
+      "loss": 0.3019,
+      "step": 3325
+    },
+    {
+      "epoch": 1.9709629629629628,
+      "grad_norm": 3.2473840713500977,
+      "learning_rate": 6.864256075874333e-06,
+      "loss": 0.3385,
+      "step": 3326
+    },
+    {
+      "epoch": 1.9715555555555555,
+      "grad_norm": 1.774461269378662,
+      "learning_rate": 6.86030428769018e-06,
+      "loss": 0.2024,
+      "step": 3327
+    },
+    {
+      "epoch": 1.9721481481481482,
+      "grad_norm": 2.534574270248413,
+      "learning_rate": 6.8563524995060275e-06,
+      "loss": 0.2617,
+      "step": 3328
+    },
+    {
+      "epoch": 1.9727407407407407,
+      "grad_norm": 2.1812868118286133,
+      "learning_rate": 6.852400711321874e-06,
+      "loss": 0.2179,
+      "step": 3329
+    },
+    {
+      "epoch": 1.9733333333333334,
+      "grad_norm": 3.954679012298584,
+      "learning_rate": 6.848448923137721e-06,
+      "loss": 0.4758,
+      "step": 3330
+    },
+    {
+      "epoch": 1.973925925925926,
+      "grad_norm": 1.9108985662460327,
+      "learning_rate": 6.844497134953567e-06,
+      "loss": 0.1918,
+      "step": 3331
+    },
+    {
+      "epoch": 1.9745185185185186,
+      "grad_norm": 2.650965929031372,
+      "learning_rate": 6.840545346769414e-06,
+      "loss": 0.3304,
+      "step": 3332
+    },
+    {
+      "epoch": 1.975111111111111,
+      "grad_norm": 2.171229362487793,
+      "learning_rate": 6.83659355858526e-06,
+      "loss": 0.1741,
+      "step": 3333
+    },
+    {
+      "epoch": 1.9757037037037037,
+      "grad_norm": 2.637939214706421,
+      "learning_rate": 6.8326417704011075e-06,
+      "loss": 0.3289,
+      "step": 3334
+    },
+    {
+      "epoch": 1.9762962962962964,
+      "grad_norm": 2.15305495262146,
+      "learning_rate": 6.828689982216954e-06,
+      "loss": 0.256,
+      "step": 3335
+    },
+    {
+      "epoch": 1.976888888888889,
+      "grad_norm": 2.5916049480438232,
+      "learning_rate": 6.824738194032801e-06,
+      "loss": 0.2056,
+      "step": 3336
+    },
+    {
+      "epoch": 1.9774814814814814,
+      "grad_norm": 1.517339825630188,
+      "learning_rate": 6.820786405848647e-06,
+      "loss": 0.1989,
+      "step": 3337
+    },
+    {
+      "epoch": 1.978074074074074,
+      "grad_norm": 1.5904617309570312,
+      "learning_rate": 6.816834617664494e-06,
+      "loss": 0.2339,
+      "step": 3338
+    },
+    {
+      "epoch": 1.9786666666666668,
+      "grad_norm": 2.3190970420837402,
+      "learning_rate": 6.81288282948034e-06,
+      "loss": 0.2701,
+      "step": 3339
+    },
+    {
+      "epoch": 1.9792592592592593,
+      "grad_norm": 1.6169663667678833,
+      "learning_rate": 6.8089310412961875e-06,
+      "loss": 0.2529,
+      "step": 3340
+    },
+    {
+      "epoch": 1.9798518518518518,
+      "grad_norm": 2.060600996017456,
+      "learning_rate": 6.804979253112034e-06,
+      "loss": 0.2995,
+      "step": 3341
+    },
+    {
+      "epoch": 1.9804444444444445,
+      "grad_norm": 1.7233482599258423,
+      "learning_rate": 6.801027464927881e-06,
+      "loss": 0.2415,
+      "step": 3342
+    },
+    {
+      "epoch": 1.9810370370370372,
+      "grad_norm": 2.4129388332366943,
+      "learning_rate": 6.797075676743727e-06,
+      "loss": 0.2957,
+      "step": 3343
+    },
+    {
+      "epoch": 1.9816296296296296,
+      "grad_norm": 1.8140047788619995,
+      "learning_rate": 6.793123888559574e-06,
+      "loss": 0.2311,
+      "step": 3344
+    },
+    {
+      "epoch": 1.982222222222222,
+      "grad_norm": 3.349799394607544,
+      "learning_rate": 6.78917210037542e-06,
+      "loss": 0.2947,
+      "step": 3345
+    },
+    {
+      "epoch": 1.9828148148148148,
+      "grad_norm": 5.987142562866211,
+      "learning_rate": 6.7852203121912675e-06,
+      "loss": 0.4823,
+      "step": 3346
+    },
+    {
+      "epoch": 1.9834074074074075,
+      "grad_norm": 1.7862939834594727,
+      "learning_rate": 6.781268524007114e-06,
+      "loss": 0.2628,
+      "step": 3347
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 2.2196598052978516,
+      "learning_rate": 6.777316735822961e-06,
+      "loss": 0.297,
+      "step": 3348
+    },
+    {
+      "epoch": 1.9845925925925925,
+      "grad_norm": 1.8778001070022583,
+      "learning_rate": 6.773364947638807e-06,
+      "loss": 0.2579,
+      "step": 3349
+    },
+    {
+      "epoch": 1.9851851851851852,
+      "grad_norm": 2.286541700363159,
+      "learning_rate": 6.769413159454654e-06,
+      "loss": 0.2876,
+      "step": 3350
+    },
+    {
+      "epoch": 1.9857777777777779,
+      "grad_norm": 2.20914363861084,
+      "learning_rate": 6.765461371270501e-06,
+      "loss": 0.183,
+      "step": 3351
+    },
+    {
+      "epoch": 1.9863703703703703,
+      "grad_norm": 2.4322643280029297,
+      "learning_rate": 6.7615095830863475e-06,
+      "loss": 0.3183,
+      "step": 3352
+    },
+    {
+      "epoch": 1.9869629629629628,
+      "grad_norm": 1.6307332515716553,
+      "learning_rate": 6.757557794902195e-06,
+      "loss": 0.1968,
+      "step": 3353
+    },
+    {
+      "epoch": 1.9875555555555555,
+      "grad_norm": 2.2455692291259766,
+      "learning_rate": 6.753606006718041e-06,
+      "loss": 0.2777,
+      "step": 3354
+    },
+    {
+      "epoch": 1.9881481481481482,
+      "grad_norm": 3.1981136798858643,
+      "learning_rate": 6.749654218533888e-06,
+      "loss": 0.3463,
+      "step": 3355
+    },
+    {
+      "epoch": 1.9887407407407407,
+      "grad_norm": 3.5636017322540283,
+      "learning_rate": 6.745702430349734e-06,
+      "loss": 0.1991,
+      "step": 3356
+    },
+    {
+      "epoch": 1.9893333333333332,
+      "grad_norm": 2.428732395172119,
+      "learning_rate": 6.741750642165581e-06,
+      "loss": 0.2718,
+      "step": 3357
+    },
+    {
+      "epoch": 1.9899259259259259,
+      "grad_norm": 1.9227361679077148,
+      "learning_rate": 6.7377988539814275e-06,
+      "loss": 0.2209,
+      "step": 3358
+    },
+    {
+      "epoch": 1.9905185185185186,
+      "grad_norm": 1.8174234628677368,
+      "learning_rate": 6.733847065797275e-06,
+      "loss": 0.2592,
+      "step": 3359
+    },
+    {
+      "epoch": 1.991111111111111,
+      "grad_norm": 1.7343745231628418,
+      "learning_rate": 6.729895277613121e-06,
+      "loss": 0.2801,
+      "step": 3360
+    },
+    {
+      "epoch": 1.9917037037037038,
+      "grad_norm": 1.8778501749038696,
+      "learning_rate": 6.725943489428968e-06,
+      "loss": 0.2384,
+      "step": 3361
+    },
+    {
+      "epoch": 1.9922962962962965,
+      "grad_norm": 2.2471189498901367,
+      "learning_rate": 6.721991701244814e-06,
+      "loss": 0.3111,
+      "step": 3362
+    },
+    {
+      "epoch": 1.992888888888889,
+      "grad_norm": 1.9849950075149536,
+      "learning_rate": 6.7180399130606596e-06,
+      "loss": 0.1897,
+      "step": 3363
+    },
+    {
+      "epoch": 1.9934814814814814,
+      "grad_norm": 2.4588513374328613,
+      "learning_rate": 6.714088124876507e-06,
+      "loss": 0.3397,
+      "step": 3364
+    },
+    {
+      "epoch": 1.994074074074074,
+      "grad_norm": 2.2077219486236572,
+      "learning_rate": 6.710136336692353e-06,
+      "loss": 0.3953,
+      "step": 3365
+    },
+    {
+      "epoch": 1.9946666666666668,
+      "grad_norm": 2.0891880989074707,
+      "learning_rate": 6.7061845485082e-06,
+      "loss": 0.2883,
+      "step": 3366
+    },
+    {
+      "epoch": 1.9952592592592593,
+      "grad_norm": 2.0964105129241943,
+      "learning_rate": 6.702232760324047e-06,
+      "loss": 0.2809,
+      "step": 3367
+    },
+    {
+      "epoch": 1.9958518518518518,
+      "grad_norm": 1.763536810874939,
+      "learning_rate": 6.698280972139893e-06,
+      "loss": 0.2152,
+      "step": 3368
+    },
+    {
+      "epoch": 1.9964444444444445,
+      "grad_norm": 2.7687268257141113,
+      "learning_rate": 6.69432918395574e-06,
+      "loss": 0.288,
+      "step": 3369
+    },
+    {
+      "epoch": 1.9970370370370372,
+      "grad_norm": 2.004242420196533,
+      "learning_rate": 6.690377395771587e-06,
+      "loss": 0.277,
+      "step": 3370
+    },
+    {
+      "epoch": 1.9976296296296296,
+      "grad_norm": 1.9711761474609375,
+      "learning_rate": 6.686425607587434e-06,
+      "loss": 0.2176,
+      "step": 3371
+    },
+    {
+      "epoch": 1.9982222222222221,
+      "grad_norm": 2.1297621726989746,
+      "learning_rate": 6.68247381940328e-06,
+      "loss": 0.2887,
+      "step": 3372
+    },
+    {
+      "epoch": 1.9988148148148148,
+      "grad_norm": 2.825065851211548,
+      "learning_rate": 6.678522031219127e-06,
+      "loss": 0.3165,
+      "step": 3373
+    },
+    {
+      "epoch": 1.9994074074074075,
+      "grad_norm": 1.8903781175613403,
+      "learning_rate": 6.674570243034973e-06,
+      "loss": 0.2498,
+      "step": 3374
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4763081073760986,
+      "learning_rate": 6.67061845485082e-06,
+      "loss": 0.202,
+      "step": 3375
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.284396767616272,
+      "eval_runtime": 169.3962,
+      "eval_samples_per_second": 35.42,
+      "eval_steps_per_second": 35.42,
+      "step": 3375
+    },
+    {
+      "epoch": 2.0005925925925925,
+      "grad_norm": 1.69547700881958,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.2434,
+      "step": 3376
+    },
+    {
+      "epoch": 2.0011851851851854,
+      "grad_norm": 2.152818441390991,
+      "learning_rate": 6.662714878482514e-06,
+      "loss": 0.2183,
+      "step": 3377
+    },
+    {
+      "epoch": 2.001777777777778,
+      "grad_norm": 2.1777024269104004,
+      "learning_rate": 6.65876309029836e-06,
+      "loss": 0.1544,
+      "step": 3378
+    },
+    {
+      "epoch": 2.0023703703703704,
+      "grad_norm": 2.1785624027252197,
+      "learning_rate": 6.654811302114207e-06,
+      "loss": 0.1928,
+      "step": 3379
+    },
+    {
+      "epoch": 2.002962962962963,
+      "grad_norm": 2.1276209354400635,
+      "learning_rate": 6.650859513930053e-06,
+      "loss": 0.1532,
+      "step": 3380
+    },
+    {
+      "epoch": 2.0035555555555558,
+      "grad_norm": 1.5840874910354614,
+      "learning_rate": 6.6469077257459e-06,
+      "loss": 0.1546,
+      "step": 3381
+    },
+    {
+      "epoch": 2.0041481481481482,
+      "grad_norm": 1.743159532546997,
+      "learning_rate": 6.642955937561747e-06,
+      "loss": 0.1779,
+      "step": 3382
+    },
+    {
+      "epoch": 2.0047407407407407,
+      "grad_norm": 1.7124453783035278,
+      "learning_rate": 6.639004149377594e-06,
+      "loss": 0.129,
+      "step": 3383
+    },
+    {
+      "epoch": 2.005333333333333,
+      "grad_norm": 2.4132299423217773,
+      "learning_rate": 6.63505236119344e-06,
+      "loss": 0.1738,
+      "step": 3384
+    },
+    {
+      "epoch": 2.005925925925926,
+      "grad_norm": 2.444908380508423,
+      "learning_rate": 6.631100573009287e-06,
+      "loss": 0.1917,
+      "step": 3385
+    },
+    {
+      "epoch": 2.0065185185185186,
+      "grad_norm": 3.968837022781372,
+      "learning_rate": 6.627148784825133e-06,
+      "loss": 0.1796,
+      "step": 3386
+    },
+    {
+      "epoch": 2.007111111111111,
+      "grad_norm": 1.6904175281524658,
+      "learning_rate": 6.62319699664098e-06,
+      "loss": 0.2896,
+      "step": 3387
+    },
+    {
+      "epoch": 2.0077037037037035,
+      "grad_norm": 2.2473864555358887,
+      "learning_rate": 6.619245208456827e-06,
+      "loss": 0.2095,
+      "step": 3388
+    },
+    {
+      "epoch": 2.0082962962962965,
+      "grad_norm": 2.698801040649414,
+      "learning_rate": 6.615293420272674e-06,
+      "loss": 0.1924,
+      "step": 3389
+    },
+    {
+      "epoch": 2.008888888888889,
+      "grad_norm": 2.153038501739502,
+      "learning_rate": 6.611341632088521e-06,
+      "loss": 0.2322,
+      "step": 3390
+    },
+    {
+      "epoch": 2.0094814814814814,
+      "grad_norm": 2.1097731590270996,
+      "learning_rate": 6.607389843904367e-06,
+      "loss": 0.1177,
+      "step": 3391
+    },
+    {
+      "epoch": 2.010074074074074,
+      "grad_norm": 2.9107837677001953,
+      "learning_rate": 6.603438055720214e-06,
+      "loss": 0.2343,
+      "step": 3392
+    },
+    {
+      "epoch": 2.010666666666667,
+      "grad_norm": 1.8394381999969482,
+      "learning_rate": 6.59948626753606e-06,
+      "loss": 0.2186,
+      "step": 3393
+    },
+    {
+      "epoch": 2.0112592592592593,
+      "grad_norm": 1.8307627439498901,
+      "learning_rate": 6.5955344793519075e-06,
+      "loss": 0.1913,
+      "step": 3394
+    },
+    {
+      "epoch": 2.011851851851852,
+      "grad_norm": 2.378425121307373,
+      "learning_rate": 6.591582691167754e-06,
+      "loss": 0.2532,
+      "step": 3395
+    },
+    {
+      "epoch": 2.0124444444444443,
+      "grad_norm": 1.768801212310791,
+      "learning_rate": 6.587630902983601e-06,
+      "loss": 0.1903,
+      "step": 3396
+    },
+    {
+      "epoch": 2.013037037037037,
+      "grad_norm": 2.165327548980713,
+      "learning_rate": 6.583679114799447e-06,
+      "loss": 0.1657,
+      "step": 3397
+    },
+    {
+      "epoch": 2.0136296296296297,
+      "grad_norm": 1.9814223051071167,
+      "learning_rate": 6.579727326615294e-06,
+      "loss": 0.1784,
+      "step": 3398
+    },
+    {
+      "epoch": 2.014222222222222,
+      "grad_norm": 2.5745491981506348,
+      "learning_rate": 6.57577553843114e-06,
+      "loss": 0.2544,
+      "step": 3399
+    },
+    {
+      "epoch": 2.0148148148148146,
+      "grad_norm": 2.005289316177368,
+      "learning_rate": 6.5718237502469875e-06,
+      "loss": 0.1732,
+      "step": 3400
+    },
+    {
+      "epoch": 2.0154074074074075,
+      "grad_norm": 1.8920613527297974,
+      "learning_rate": 6.567871962062834e-06,
+      "loss": 0.1882,
+      "step": 3401
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 2.333807945251465,
+      "learning_rate": 6.563920173878681e-06,
+      "loss": 0.1735,
+      "step": 3402
+    },
+    {
+      "epoch": 2.0165925925925925,
+      "grad_norm": 1.652557134628296,
+      "learning_rate": 6.559968385694527e-06,
+      "loss": 0.1764,
+      "step": 3403
+    },
+    {
+      "epoch": 2.017185185185185,
+      "grad_norm": 1.87971031665802,
+      "learning_rate": 6.556016597510374e-06,
+      "loss": 0.2081,
+      "step": 3404
+    },
+    {
+      "epoch": 2.017777777777778,
+      "grad_norm": 2.506694793701172,
+      "learning_rate": 6.55206480932622e-06,
+      "loss": 0.2132,
+      "step": 3405
+    },
+    {
+      "epoch": 2.0183703703703704,
+      "grad_norm": 2.490078926086426,
+      "learning_rate": 6.5481130211420675e-06,
+      "loss": 0.3371,
+      "step": 3406
+    },
+    {
+      "epoch": 2.018962962962963,
+      "grad_norm": 2.162909746170044,
+      "learning_rate": 6.544161232957914e-06,
+      "loss": 0.1981,
+      "step": 3407
+    },
+    {
+      "epoch": 2.0195555555555558,
+      "grad_norm": 2.1677353382110596,
+      "learning_rate": 6.540209444773761e-06,
+      "loss": 0.2079,
+      "step": 3408
+    },
+    {
+      "epoch": 2.0201481481481482,
+      "grad_norm": 2.394774913787842,
+      "learning_rate": 6.536257656589607e-06,
+      "loss": 0.2144,
+      "step": 3409
+    },
+    {
+      "epoch": 2.0207407407407407,
+      "grad_norm": 2.080420970916748,
+      "learning_rate": 6.532305868405454e-06,
+      "loss": 0.2174,
+      "step": 3410
+    },
+    {
+      "epoch": 2.021333333333333,
+      "grad_norm": 1.5731886625289917,
+      "learning_rate": 6.5283540802213e-06,
+      "loss": 0.177,
+      "step": 3411
+    },
+    {
+      "epoch": 2.021925925925926,
+      "grad_norm": 2.357145071029663,
+      "learning_rate": 6.5244022920371475e-06,
+      "loss": 0.205,
+      "step": 3412
+    },
+    {
+      "epoch": 2.0225185185185186,
+      "grad_norm": 1.5392361879348755,
+      "learning_rate": 6.5204505038529946e-06,
+      "loss": 0.1846,
+      "step": 3413
+    },
+    {
+      "epoch": 2.023111111111111,
+      "grad_norm": 1.7238672971725464,
+      "learning_rate": 6.516498715668841e-06,
+      "loss": 0.2067,
+      "step": 3414
+    },
+    {
+      "epoch": 2.0237037037037036,
+      "grad_norm": 1.923115849494934,
+      "learning_rate": 6.512546927484688e-06,
+      "loss": 0.2057,
+      "step": 3415
+    },
+    {
+      "epoch": 2.0242962962962965,
+      "grad_norm": 1.7704333066940308,
+      "learning_rate": 6.508595139300534e-06,
+      "loss": 0.1414,
+      "step": 3416
+    },
+    {
+      "epoch": 2.024888888888889,
+      "grad_norm": 1.6886004209518433,
+      "learning_rate": 6.504643351116381e-06,
+      "loss": 0.1952,
+      "step": 3417
+    },
+    {
+      "epoch": 2.0254814814814814,
+      "grad_norm": 2.2009501457214355,
+      "learning_rate": 6.5006915629322275e-06,
+      "loss": 0.1592,
+      "step": 3418
+    },
+    {
+      "epoch": 2.026074074074074,
+      "grad_norm": 2.303438425064087,
+      "learning_rate": 6.4967397747480746e-06,
+      "loss": 0.243,
+      "step": 3419
+    },
+    {
+      "epoch": 2.026666666666667,
+      "grad_norm": 2.022925853729248,
+      "learning_rate": 6.492787986563921e-06,
+      "loss": 0.1672,
+      "step": 3420
+    },
+    {
+      "epoch": 2.0272592592592593,
+      "grad_norm": 1.5532667636871338,
+      "learning_rate": 6.488836198379768e-06,
+      "loss": 0.1741,
+      "step": 3421
+    },
+    {
+      "epoch": 2.027851851851852,
+      "grad_norm": 1.843652606010437,
+      "learning_rate": 6.484884410195614e-06,
+      "loss": 0.2088,
+      "step": 3422
+    },
+    {
+      "epoch": 2.0284444444444443,
+      "grad_norm": 2.0094826221466064,
+      "learning_rate": 6.480932622011461e-06,
+      "loss": 0.2148,
+      "step": 3423
+    },
+    {
+      "epoch": 2.029037037037037,
+      "grad_norm": 1.7647525072097778,
+      "learning_rate": 6.4769808338273075e-06,
+      "loss": 0.115,
+      "step": 3424
+    },
+    {
+      "epoch": 2.0296296296296297,
+      "grad_norm": 2.2348034381866455,
+      "learning_rate": 6.4730290456431546e-06,
+      "loss": 0.1864,
+      "step": 3425
+    },
+    {
+      "epoch": 2.030222222222222,
+      "grad_norm": 1.3531190156936646,
+      "learning_rate": 6.469077257459001e-06,
+      "loss": 0.0981,
+      "step": 3426
+    },
+    {
+      "epoch": 2.0308148148148146,
+      "grad_norm": 1.8812849521636963,
+      "learning_rate": 6.465125469274848e-06,
+      "loss": 0.1528,
+      "step": 3427
+    },
+    {
+      "epoch": 2.0314074074074076,
+      "grad_norm": 1.8431155681610107,
+      "learning_rate": 6.461173681090694e-06,
+      "loss": 0.1726,
+      "step": 3428
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 2.2347006797790527,
+      "learning_rate": 6.457221892906541e-06,
+      "loss": 0.2532,
+      "step": 3429
+    },
+    {
+      "epoch": 2.0325925925925925,
+      "grad_norm": 2.1870853900909424,
+      "learning_rate": 6.4532701047223875e-06,
+      "loss": 0.203,
+      "step": 3430
+    },
+    {
+      "epoch": 2.033185185185185,
+      "grad_norm": 2.374680519104004,
+      "learning_rate": 6.4493183165382346e-06,
+      "loss": 0.1902,
+      "step": 3431
+    },
+    {
+      "epoch": 2.033777777777778,
+      "grad_norm": 1.717908501625061,
+      "learning_rate": 6.445366528354081e-06,
+      "loss": 0.2267,
+      "step": 3432
+    },
+    {
+      "epoch": 2.0343703703703704,
+      "grad_norm": 2.951277494430542,
+      "learning_rate": 6.441414740169928e-06,
+      "loss": 0.379,
+      "step": 3433
+    },
+    {
+      "epoch": 2.034962962962963,
+      "grad_norm": 1.8061403036117554,
+      "learning_rate": 6.437462951985774e-06,
+      "loss": 0.1637,
+      "step": 3434
+    },
+    {
+      "epoch": 2.0355555555555553,
+      "grad_norm": 1.427119255065918,
+      "learning_rate": 6.433511163801621e-06,
+      "loss": 0.2286,
+      "step": 3435
+    },
+    {
+      "epoch": 2.0361481481481483,
+      "grad_norm": 2.0395805835723877,
+      "learning_rate": 6.429559375617468e-06,
+      "loss": 0.1824,
+      "step": 3436
+    },
+    {
+      "epoch": 2.0367407407407407,
+      "grad_norm": 1.8193371295928955,
+      "learning_rate": 6.4256075874333146e-06,
+      "loss": 0.2237,
+      "step": 3437
+    },
+    {
+      "epoch": 2.037333333333333,
+      "grad_norm": 1.70955491065979,
+      "learning_rate": 6.421655799249162e-06,
+      "loss": 0.2047,
+      "step": 3438
+    },
+    {
+      "epoch": 2.037925925925926,
+      "grad_norm": 1.7725789546966553,
+      "learning_rate": 6.417704011065008e-06,
+      "loss": 0.2004,
+      "step": 3439
+    },
+    {
+      "epoch": 2.0385185185185186,
+      "grad_norm": 1.964396595954895,
+      "learning_rate": 6.413752222880855e-06,
+      "loss": 0.2399,
+      "step": 3440
+    },
+    {
+      "epoch": 2.039111111111111,
+      "grad_norm": 2.045922040939331,
+      "learning_rate": 6.409800434696701e-06,
+      "loss": 0.1764,
+      "step": 3441
+    },
+    {
+      "epoch": 2.0397037037037036,
+      "grad_norm": 3.472876787185669,
+      "learning_rate": 6.405848646512547e-06,
+      "loss": 0.251,
+      "step": 3442
+    },
+    {
+      "epoch": 2.0402962962962965,
+      "grad_norm": 3.0538463592529297,
+      "learning_rate": 6.401896858328394e-06,
+      "loss": 0.2633,
+      "step": 3443
+    },
+    {
+      "epoch": 2.040888888888889,
+      "grad_norm": 2.376619815826416,
+      "learning_rate": 6.39794507014424e-06,
+      "loss": 0.2029,
+      "step": 3444
+    },
+    {
+      "epoch": 2.0414814814814815,
+      "grad_norm": 2.2123444080352783,
+      "learning_rate": 6.393993281960087e-06,
+      "loss": 0.2205,
+      "step": 3445
+    },
+    {
+      "epoch": 2.042074074074074,
+      "grad_norm": 1.5449782609939575,
+      "learning_rate": 6.390041493775933e-06,
+      "loss": 0.2214,
+      "step": 3446
+    },
+    {
+      "epoch": 2.042666666666667,
+      "grad_norm": 2.091628074645996,
+      "learning_rate": 6.38608970559178e-06,
+      "loss": 0.2504,
+      "step": 3447
+    },
+    {
+      "epoch": 2.0432592592592593,
+      "grad_norm": 2.7278592586517334,
+      "learning_rate": 6.382137917407627e-06,
+      "loss": 0.2941,
+      "step": 3448
+    },
+    {
+      "epoch": 2.043851851851852,
+      "grad_norm": 2.6500473022460938,
+      "learning_rate": 6.378186129223474e-06,
+      "loss": 0.2349,
+      "step": 3449
+    },
+    {
+      "epoch": 2.0444444444444443,
+      "grad_norm": 3.322272539138794,
+      "learning_rate": 6.37423434103932e-06,
+      "loss": 0.2317,
+      "step": 3450
+    },
+    {
+      "epoch": 2.045037037037037,
+      "grad_norm": 2.693176746368408,
+      "learning_rate": 6.370282552855167e-06,
+      "loss": 0.2341,
+      "step": 3451
+    },
+    {
+      "epoch": 2.0456296296296297,
+      "grad_norm": 2.010737180709839,
+      "learning_rate": 6.366330764671014e-06,
+      "loss": 0.2591,
+      "step": 3452
+    },
+    {
+      "epoch": 2.046222222222222,
+      "grad_norm": 2.1280605792999268,
+      "learning_rate": 6.36237897648686e-06,
+      "loss": 0.2339,
+      "step": 3453
+    },
+    {
+      "epoch": 2.0468148148148146,
+      "grad_norm": 2.0335004329681396,
+      "learning_rate": 6.3584271883027075e-06,
+      "loss": 0.1893,
+      "step": 3454
+    },
+    {
+      "epoch": 2.0474074074074076,
+      "grad_norm": 2.539130449295044,
+      "learning_rate": 6.354475400118554e-06,
+      "loss": 0.2383,
+      "step": 3455
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 2.552173614501953,
+      "learning_rate": 6.350523611934401e-06,
+      "loss": 0.2652,
+      "step": 3456
+    },
+    {
+      "epoch": 2.0485925925925925,
+      "grad_norm": 2.314741849899292,
+      "learning_rate": 6.346571823750247e-06,
+      "loss": 0.1401,
+      "step": 3457
+    },
+    {
+      "epoch": 2.049185185185185,
+      "grad_norm": 2.059671401977539,
+      "learning_rate": 6.342620035566094e-06,
+      "loss": 0.2235,
+      "step": 3458
+    },
+    {
+      "epoch": 2.049777777777778,
+      "grad_norm": 2.422234058380127,
+      "learning_rate": 6.33866824738194e-06,
+      "loss": 0.2255,
+      "step": 3459
+    },
+    {
+      "epoch": 2.0503703703703704,
+      "grad_norm": 2.1150460243225098,
+      "learning_rate": 6.3347164591977875e-06,
+      "loss": 0.1957,
+      "step": 3460
+    },
+    {
+      "epoch": 2.050962962962963,
+      "grad_norm": 1.8067865371704102,
+      "learning_rate": 6.330764671013634e-06,
+      "loss": 0.2073,
+      "step": 3461
+    },
+    {
+      "epoch": 2.0515555555555554,
+      "grad_norm": 2.6041641235351562,
+      "learning_rate": 6.326812882829481e-06,
+      "loss": 0.2592,
+      "step": 3462
+    },
+    {
+      "epoch": 2.0521481481481483,
+      "grad_norm": 2.2044756412506104,
+      "learning_rate": 6.322861094645327e-06,
+      "loss": 0.2284,
+      "step": 3463
+    },
+    {
+      "epoch": 2.0527407407407408,
+      "grad_norm": 2.100971221923828,
+      "learning_rate": 6.318909306461174e-06,
+      "loss": 0.2553,
+      "step": 3464
+    },
+    {
+      "epoch": 2.0533333333333332,
+      "grad_norm": 2.7908852100372314,
+      "learning_rate": 6.31495751827702e-06,
+      "loss": 0.146,
+      "step": 3465
+    },
+    {
+      "epoch": 2.0539259259259257,
+      "grad_norm": 2.4862561225891113,
+      "learning_rate": 6.3110057300928675e-06,
+      "loss": 0.2875,
+      "step": 3466
+    },
+    {
+      "epoch": 2.0545185185185186,
+      "grad_norm": 1.7847248315811157,
+      "learning_rate": 6.307053941908714e-06,
+      "loss": 0.2297,
+      "step": 3467
+    },
+    {
+      "epoch": 2.055111111111111,
+      "grad_norm": 2.059884786605835,
+      "learning_rate": 6.303102153724561e-06,
+      "loss": 0.1258,
+      "step": 3468
+    },
+    {
+      "epoch": 2.0557037037037036,
+      "grad_norm": 1.4407224655151367,
+      "learning_rate": 6.299150365540407e-06,
+      "loss": 0.1433,
+      "step": 3469
+    },
+    {
+      "epoch": 2.0562962962962965,
+      "grad_norm": 2.5582833290100098,
+      "learning_rate": 6.295198577356254e-06,
+      "loss": 0.2576,
+      "step": 3470
+    },
+    {
+      "epoch": 2.056888888888889,
+      "grad_norm": 3.1923375129699707,
+      "learning_rate": 6.2912467891721e-06,
+      "loss": 0.3368,
+      "step": 3471
+    },
+    {
+      "epoch": 2.0574814814814815,
+      "grad_norm": 1.9265085458755493,
+      "learning_rate": 6.2872950009879475e-06,
+      "loss": 0.2156,
+      "step": 3472
+    },
+    {
+      "epoch": 2.058074074074074,
+      "grad_norm": 1.889402985572815,
+      "learning_rate": 6.283343212803794e-06,
+      "loss": 0.2603,
+      "step": 3473
+    },
+    {
+      "epoch": 2.058666666666667,
+      "grad_norm": 1.812591791152954,
+      "learning_rate": 6.279391424619641e-06,
+      "loss": 0.1983,
+      "step": 3474
+    },
+    {
+      "epoch": 2.0592592592592593,
+      "grad_norm": 1.8562099933624268,
+      "learning_rate": 6.275439636435488e-06,
+      "loss": 0.2459,
+      "step": 3475
+    },
+    {
+      "epoch": 2.059851851851852,
+      "grad_norm": 2.075989007949829,
+      "learning_rate": 6.271487848251334e-06,
+      "loss": 0.1635,
+      "step": 3476
+    },
+    {
+      "epoch": 2.0604444444444443,
+      "grad_norm": 2.168527841567993,
+      "learning_rate": 6.267536060067181e-06,
+      "loss": 0.1528,
+      "step": 3477
+    },
+    {
+      "epoch": 2.0610370370370372,
+      "grad_norm": 1.854296326637268,
+      "learning_rate": 6.2635842718830275e-06,
+      "loss": 0.1348,
+      "step": 3478
+    },
+    {
+      "epoch": 2.0616296296296297,
+      "grad_norm": 1.6448304653167725,
+      "learning_rate": 6.2596324836988746e-06,
+      "loss": 0.1324,
+      "step": 3479
+    },
+    {
+      "epoch": 2.062222222222222,
+      "grad_norm": 1.760864019393921,
+      "learning_rate": 6.255680695514721e-06,
+      "loss": 0.1715,
+      "step": 3480
+    },
+    {
+      "epoch": 2.0628148148148147,
+      "grad_norm": 2.132528305053711,
+      "learning_rate": 6.251728907330568e-06,
+      "loss": 0.1994,
+      "step": 3481
+    },
+    {
+      "epoch": 2.0634074074074076,
+      "grad_norm": 2.4838080406188965,
+      "learning_rate": 6.247777119146414e-06,
+      "loss": 0.1442,
+      "step": 3482
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 2.4231603145599365,
+      "learning_rate": 6.243825330962261e-06,
+      "loss": 0.1896,
+      "step": 3483
+    },
+    {
+      "epoch": 2.0645925925925925,
+      "grad_norm": 2.7686562538146973,
+      "learning_rate": 6.2398735427781075e-06,
+      "loss": 0.1931,
+      "step": 3484
+    },
+    {
+      "epoch": 2.065185185185185,
+      "grad_norm": 1.8741475343704224,
+      "learning_rate": 6.2359217545939546e-06,
+      "loss": 0.1608,
+      "step": 3485
+    },
+    {
+      "epoch": 2.065777777777778,
+      "grad_norm": 2.247204303741455,
+      "learning_rate": 6.231969966409801e-06,
+      "loss": 0.1574,
+      "step": 3486
+    },
+    {
+      "epoch": 2.0663703703703704,
+      "grad_norm": 3.3516194820404053,
+      "learning_rate": 6.228018178225648e-06,
+      "loss": 0.2075,
+      "step": 3487
+    },
+    {
+      "epoch": 2.066962962962963,
+      "grad_norm": 2.5519919395446777,
+      "learning_rate": 6.224066390041494e-06,
+      "loss": 0.2582,
+      "step": 3488
+    },
+    {
+      "epoch": 2.0675555555555554,
+      "grad_norm": 2.7084174156188965,
+      "learning_rate": 6.220114601857341e-06,
+      "loss": 0.3083,
+      "step": 3489
+    },
+    {
+      "epoch": 2.0681481481481483,
+      "grad_norm": 2.4946017265319824,
+      "learning_rate": 6.2161628136731875e-06,
+      "loss": 0.1901,
+      "step": 3490
+    },
+    {
+      "epoch": 2.0687407407407408,
+      "grad_norm": 2.3361940383911133,
+      "learning_rate": 6.2122110254890345e-06,
+      "loss": 0.2657,
+      "step": 3491
+    },
+    {
+      "epoch": 2.0693333333333332,
+      "grad_norm": 2.4847466945648193,
+      "learning_rate": 6.208259237304881e-06,
+      "loss": 0.2831,
+      "step": 3492
+    },
+    {
+      "epoch": 2.0699259259259257,
+      "grad_norm": 2.185384511947632,
+      "learning_rate": 6.204307449120728e-06,
+      "loss": 0.2034,
+      "step": 3493
+    },
+    {
+      "epoch": 2.0705185185185186,
+      "grad_norm": 3.6335928440093994,
+      "learning_rate": 6.200355660936574e-06,
+      "loss": 0.3795,
+      "step": 3494
+    },
+    {
+      "epoch": 2.071111111111111,
+      "grad_norm": 1.775154948234558,
+      "learning_rate": 6.196403872752421e-06,
+      "loss": 0.1351,
+      "step": 3495
+    },
+    {
+      "epoch": 2.0717037037037036,
+      "grad_norm": 2.4240286350250244,
+      "learning_rate": 6.1924520845682675e-06,
+      "loss": 0.1908,
+      "step": 3496
+    },
+    {
+      "epoch": 2.0722962962962965,
+      "grad_norm": 1.621910572052002,
+      "learning_rate": 6.1885002963841145e-06,
+      "loss": 0.248,
+      "step": 3497
+    },
+    {
+      "epoch": 2.072888888888889,
+      "grad_norm": 2.105346202850342,
+      "learning_rate": 6.184548508199962e-06,
+      "loss": 0.2197,
+      "step": 3498
+    },
+    {
+      "epoch": 2.0734814814814815,
+      "grad_norm": 2.193406820297241,
+      "learning_rate": 6.180596720015808e-06,
+      "loss": 0.2772,
+      "step": 3499
+    },
+    {
+      "epoch": 2.074074074074074,
+      "grad_norm": 1.7381172180175781,
+      "learning_rate": 6.176644931831655e-06,
+      "loss": 0.25,
+      "step": 3500
+    },
+    {
+      "epoch": 2.074666666666667,
+      "grad_norm": 3.70731520652771,
+      "learning_rate": 6.172693143647501e-06,
+      "loss": 0.2652,
+      "step": 3501
+    },
+    {
+      "epoch": 2.0752592592592594,
+      "grad_norm": 2.5361835956573486,
+      "learning_rate": 6.168741355463348e-06,
+      "loss": 0.31,
+      "step": 3502
+    },
+    {
+      "epoch": 2.075851851851852,
+      "grad_norm": 2.160841226577759,
+      "learning_rate": 6.1647895672791945e-06,
+      "loss": 0.1738,
+      "step": 3503
+    },
+    {
+      "epoch": 2.0764444444444443,
+      "grad_norm": 2.9112420082092285,
+      "learning_rate": 6.160837779095042e-06,
+      "loss": 0.1757,
+      "step": 3504
+    },
+    {
+      "epoch": 2.0770370370370372,
+      "grad_norm": 2.5223746299743652,
+      "learning_rate": 6.156885990910888e-06,
+      "loss": 0.1558,
+      "step": 3505
+    },
+    {
+      "epoch": 2.0776296296296297,
+      "grad_norm": 2.345869302749634,
+      "learning_rate": 6.152934202726735e-06,
+      "loss": 0.2169,
+      "step": 3506
+    },
+    {
+      "epoch": 2.078222222222222,
+      "grad_norm": 1.9248409271240234,
+      "learning_rate": 6.148982414542581e-06,
+      "loss": 0.2075,
+      "step": 3507
+    },
+    {
+      "epoch": 2.0788148148148147,
+      "grad_norm": 3.324411153793335,
+      "learning_rate": 6.145030626358428e-06,
+      "loss": 0.1454,
+      "step": 3508
+    },
+    {
+      "epoch": 2.0794074074074076,
+      "grad_norm": 2.4760513305664062,
+      "learning_rate": 6.1410788381742745e-06,
+      "loss": 0.2137,
+      "step": 3509
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.6407538652420044,
+      "learning_rate": 6.137127049990122e-06,
+      "loss": 0.1577,
+      "step": 3510
+    },
+    {
+      "epoch": 2.0805925925925925,
+      "grad_norm": 1.6548291444778442,
+      "learning_rate": 6.133175261805968e-06,
+      "loss": 0.2102,
+      "step": 3511
+    },
+    {
+      "epoch": 2.081185185185185,
+      "grad_norm": 2.838209867477417,
+      "learning_rate": 6.129223473621815e-06,
+      "loss": 0.186,
+      "step": 3512
+    },
+    {
+      "epoch": 2.081777777777778,
+      "grad_norm": 1.7626652717590332,
+      "learning_rate": 6.125271685437661e-06,
+      "loss": 0.1961,
+      "step": 3513
+    },
+    {
+      "epoch": 2.0823703703703704,
+      "grad_norm": 1.6741745471954346,
+      "learning_rate": 6.121319897253508e-06,
+      "loss": 0.1718,
+      "step": 3514
+    },
+    {
+      "epoch": 2.082962962962963,
+      "grad_norm": 1.6288682222366333,
+      "learning_rate": 6.1173681090693545e-06,
+      "loss": 0.1721,
+      "step": 3515
+    },
+    {
+      "epoch": 2.0835555555555554,
+      "grad_norm": 1.8796281814575195,
+      "learning_rate": 6.113416320885202e-06,
+      "loss": 0.223,
+      "step": 3516
+    },
+    {
+      "epoch": 2.0841481481481483,
+      "grad_norm": 1.4125055074691772,
+      "learning_rate": 6.109464532701048e-06,
+      "loss": 0.1172,
+      "step": 3517
+    },
+    {
+      "epoch": 2.084740740740741,
+      "grad_norm": 3.1681878566741943,
+      "learning_rate": 6.105512744516895e-06,
+      "loss": 0.2724,
+      "step": 3518
+    },
+    {
+      "epoch": 2.0853333333333333,
+      "grad_norm": 1.8557376861572266,
+      "learning_rate": 6.101560956332741e-06,
+      "loss": 0.1808,
+      "step": 3519
+    },
+    {
+      "epoch": 2.0859259259259257,
+      "grad_norm": 2.314699649810791,
+      "learning_rate": 6.097609168148588e-06,
+      "loss": 0.1695,
+      "step": 3520
+    },
+    {
+      "epoch": 2.0865185185185187,
+      "grad_norm": 1.9144665002822876,
+      "learning_rate": 6.093657379964434e-06,
+      "loss": 0.1818,
+      "step": 3521
+    },
+    {
+      "epoch": 2.087111111111111,
+      "grad_norm": 1.6567996740341187,
+      "learning_rate": 6.089705591780281e-06,
+      "loss": 0.1584,
+      "step": 3522
+    },
+    {
+      "epoch": 2.0877037037037036,
+      "grad_norm": 1.6701128482818604,
+      "learning_rate": 6.085753803596127e-06,
+      "loss": 0.1394,
+      "step": 3523
+    },
+    {
+      "epoch": 2.088296296296296,
+      "grad_norm": 2.2134745121002197,
+      "learning_rate": 6.081802015411974e-06,
+      "loss": 0.217,
+      "step": 3524
+    },
+    {
+      "epoch": 2.088888888888889,
+      "grad_norm": 2.666135787963867,
+      "learning_rate": 6.07785022722782e-06,
+      "loss": 0.1726,
+      "step": 3525
+    },
+    {
+      "epoch": 2.0894814814814815,
+      "grad_norm": 3.332193613052368,
+      "learning_rate": 6.0738984390436674e-06,
+      "loss": 0.2367,
+      "step": 3526
+    },
+    {
+      "epoch": 2.090074074074074,
+      "grad_norm": 1.4514575004577637,
+      "learning_rate": 6.069946650859514e-06,
+      "loss": 0.1288,
+      "step": 3527
+    },
+    {
+      "epoch": 2.0906666666666665,
+      "grad_norm": 2.0706870555877686,
+      "learning_rate": 6.065994862675361e-06,
+      "loss": 0.2707,
+      "step": 3528
+    },
+    {
+      "epoch": 2.0912592592592594,
+      "grad_norm": 1.9376089572906494,
+      "learning_rate": 6.062043074491207e-06,
+      "loss": 0.2386,
+      "step": 3529
+    },
+    {
+      "epoch": 2.091851851851852,
+      "grad_norm": 1.4169648885726929,
+      "learning_rate": 6.058091286307054e-06,
+      "loss": 0.1386,
+      "step": 3530
+    },
+    {
+      "epoch": 2.0924444444444443,
+      "grad_norm": 2.264846086502075,
+      "learning_rate": 6.0541394981229e-06,
+      "loss": 0.1757,
+      "step": 3531
+    },
+    {
+      "epoch": 2.0930370370370373,
+      "grad_norm": 2.69500732421875,
+      "learning_rate": 6.0501877099387474e-06,
+      "loss": 0.2658,
+      "step": 3532
+    },
+    {
+      "epoch": 2.0936296296296297,
+      "grad_norm": 2.2064096927642822,
+      "learning_rate": 6.046235921754594e-06,
+      "loss": 0.3322,
+      "step": 3533
+    },
+    {
+      "epoch": 2.094222222222222,
+      "grad_norm": 2.017637252807617,
+      "learning_rate": 6.042284133570441e-06,
+      "loss": 0.2502,
+      "step": 3534
+    },
+    {
+      "epoch": 2.0948148148148147,
+      "grad_norm": 3.1731486320495605,
+      "learning_rate": 6.038332345386287e-06,
+      "loss": 0.2386,
+      "step": 3535
+    },
+    {
+      "epoch": 2.0954074074074076,
+      "grad_norm": 2.426884412765503,
+      "learning_rate": 6.034380557202134e-06,
+      "loss": 0.1862,
+      "step": 3536
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 3.0460236072540283,
+      "learning_rate": 6.030428769017981e-06,
+      "loss": 0.174,
+      "step": 3537
+    },
+    {
+      "epoch": 2.0965925925925926,
+      "grad_norm": 2.4789881706237793,
+      "learning_rate": 6.0264769808338274e-06,
+      "loss": 0.201,
+      "step": 3538
+    },
+    {
+      "epoch": 2.097185185185185,
+      "grad_norm": 1.5616146326065063,
+      "learning_rate": 6.0225251926496745e-06,
+      "loss": 0.1485,
+      "step": 3539
+    },
+    {
+      "epoch": 2.097777777777778,
+      "grad_norm": 2.1728272438049316,
+      "learning_rate": 6.018573404465521e-06,
+      "loss": 0.2198,
+      "step": 3540
+    },
+    {
+      "epoch": 2.0983703703703704,
+      "grad_norm": 1.7116999626159668,
+      "learning_rate": 6.014621616281368e-06,
+      "loss": 0.1829,
+      "step": 3541
+    },
+    {
+      "epoch": 2.098962962962963,
+      "grad_norm": 3.607231616973877,
+      "learning_rate": 6.010669828097214e-06,
+      "loss": 0.1779,
+      "step": 3542
+    },
+    {
+      "epoch": 2.0995555555555554,
+      "grad_norm": 2.0600852966308594,
+      "learning_rate": 6.006718039913061e-06,
+      "loss": 0.2609,
+      "step": 3543
+    },
+    {
+      "epoch": 2.1001481481481483,
+      "grad_norm": 2.3569209575653076,
+      "learning_rate": 6.0027662517289074e-06,
+      "loss": 0.2486,
+      "step": 3544
+    },
+    {
+      "epoch": 2.100740740740741,
+      "grad_norm": 1.548992395401001,
+      "learning_rate": 5.9988144635447545e-06,
+      "loss": 0.1453,
+      "step": 3545
+    },
+    {
+      "epoch": 2.1013333333333333,
+      "grad_norm": 1.9826208353042603,
+      "learning_rate": 5.994862675360601e-06,
+      "loss": 0.1835,
+      "step": 3546
+    },
+    {
+      "epoch": 2.1019259259259258,
+      "grad_norm": 2.0032095909118652,
+      "learning_rate": 5.990910887176448e-06,
+      "loss": 0.1603,
+      "step": 3547
+    },
+    {
+      "epoch": 2.1025185185185187,
+      "grad_norm": 2.223158359527588,
+      "learning_rate": 5.986959098992294e-06,
+      "loss": 0.1602,
+      "step": 3548
+    },
+    {
+      "epoch": 2.103111111111111,
+      "grad_norm": 1.989181637763977,
+      "learning_rate": 5.983007310808141e-06,
+      "loss": 0.1828,
+      "step": 3549
+    },
+    {
+      "epoch": 2.1037037037037036,
+      "grad_norm": 1.8343863487243652,
+      "learning_rate": 5.9790555226239874e-06,
+      "loss": 0.1287,
+      "step": 3550
+    },
+    {
+      "epoch": 2.104296296296296,
+      "grad_norm": 2.1965222358703613,
+      "learning_rate": 5.9751037344398345e-06,
+      "loss": 0.2289,
+      "step": 3551
+    },
+    {
+      "epoch": 2.104888888888889,
+      "grad_norm": 2.216294765472412,
+      "learning_rate": 5.971151946255681e-06,
+      "loss": 0.2918,
+      "step": 3552
+    },
+    {
+      "epoch": 2.1054814814814815,
+      "grad_norm": 1.7567098140716553,
+      "learning_rate": 5.967200158071528e-06,
+      "loss": 0.1588,
+      "step": 3553
+    },
+    {
+      "epoch": 2.106074074074074,
+      "grad_norm": 2.8666038513183594,
+      "learning_rate": 5.963248369887374e-06,
+      "loss": 0.1585,
+      "step": 3554
+    },
+    {
+      "epoch": 2.1066666666666665,
+      "grad_norm": 2.095102548599243,
+      "learning_rate": 5.959296581703221e-06,
+      "loss": 0.2536,
+      "step": 3555
+    },
+    {
+      "epoch": 2.1072592592592594,
+      "grad_norm": 2.0505802631378174,
+      "learning_rate": 5.9553447935190674e-06,
+      "loss": 0.1783,
+      "step": 3556
+    },
+    {
+      "epoch": 2.107851851851852,
+      "grad_norm": 2.542098045349121,
+      "learning_rate": 5.9513930053349145e-06,
+      "loss": 0.259,
+      "step": 3557
+    },
+    {
+      "epoch": 2.1084444444444443,
+      "grad_norm": 1.5500980615615845,
+      "learning_rate": 5.947441217150761e-06,
+      "loss": 0.1699,
+      "step": 3558
+    },
+    {
+      "epoch": 2.1090370370370373,
+      "grad_norm": 1.6616374254226685,
+      "learning_rate": 5.943489428966608e-06,
+      "loss": 0.1469,
+      "step": 3559
+    },
+    {
+      "epoch": 2.1096296296296297,
+      "grad_norm": 2.081533670425415,
+      "learning_rate": 5.939537640782455e-06,
+      "loss": 0.2148,
+      "step": 3560
+    },
+    {
+      "epoch": 2.110222222222222,
+      "grad_norm": 4.179418563842773,
+      "learning_rate": 5.935585852598301e-06,
+      "loss": 0.3569,
+      "step": 3561
+    },
+    {
+      "epoch": 2.1108148148148147,
+      "grad_norm": 2.8508424758911133,
+      "learning_rate": 5.931634064414148e-06,
+      "loss": 0.2033,
+      "step": 3562
+    },
+    {
+      "epoch": 2.1114074074074076,
+      "grad_norm": 1.7162904739379883,
+      "learning_rate": 5.9276822762299945e-06,
+      "loss": 0.2499,
+      "step": 3563
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 1.9955880641937256,
+      "learning_rate": 5.923730488045842e-06,
+      "loss": 0.177,
+      "step": 3564
+    },
+    {
+      "epoch": 2.1125925925925926,
+      "grad_norm": 2.577460289001465,
+      "learning_rate": 5.919778699861688e-06,
+      "loss": 0.2879,
+      "step": 3565
+    },
+    {
+      "epoch": 2.113185185185185,
+      "grad_norm": 2.396188974380493,
+      "learning_rate": 5.915826911677535e-06,
+      "loss": 0.2236,
+      "step": 3566
+    },
+    {
+      "epoch": 2.113777777777778,
+      "grad_norm": 2.77396821975708,
+      "learning_rate": 5.911875123493381e-06,
+      "loss": 0.2394,
+      "step": 3567
+    },
+    {
+      "epoch": 2.1143703703703705,
+      "grad_norm": 1.481104850769043,
+      "learning_rate": 5.907923335309228e-06,
+      "loss": 0.1632,
+      "step": 3568
+    },
+    {
+      "epoch": 2.114962962962963,
+      "grad_norm": 1.9730231761932373,
+      "learning_rate": 5.9039715471250745e-06,
+      "loss": 0.1841,
+      "step": 3569
+    },
+    {
+      "epoch": 2.1155555555555554,
+      "grad_norm": 2.6336159706115723,
+      "learning_rate": 5.900019758940922e-06,
+      "loss": 0.2859,
+      "step": 3570
+    },
+    {
+      "epoch": 2.1161481481481483,
+      "grad_norm": 1.6799319982528687,
+      "learning_rate": 5.896067970756768e-06,
+      "loss": 0.1549,
+      "step": 3571
+    },
+    {
+      "epoch": 2.116740740740741,
+      "grad_norm": 1.97369384765625,
+      "learning_rate": 5.892116182572615e-06,
+      "loss": 0.1929,
+      "step": 3572
+    },
+    {
+      "epoch": 2.1173333333333333,
+      "grad_norm": 1.4923983812332153,
+      "learning_rate": 5.888164394388461e-06,
+      "loss": 0.1243,
+      "step": 3573
+    },
+    {
+      "epoch": 2.1179259259259258,
+      "grad_norm": 2.442094326019287,
+      "learning_rate": 5.884212606204308e-06,
+      "loss": 0.2084,
+      "step": 3574
+    },
+    {
+      "epoch": 2.1185185185185187,
+      "grad_norm": 1.2563084363937378,
+      "learning_rate": 5.8802608180201545e-06,
+      "loss": 0.0951,
+      "step": 3575
+    },
+    {
+      "epoch": 2.119111111111111,
+      "grad_norm": 2.2298202514648438,
+      "learning_rate": 5.876309029836002e-06,
+      "loss": 0.2076,
+      "step": 3576
+    },
+    {
+      "epoch": 2.1197037037037036,
+      "grad_norm": 2.887589693069458,
+      "learning_rate": 5.872357241651848e-06,
+      "loss": 0.2289,
+      "step": 3577
+    },
+    {
+      "epoch": 2.120296296296296,
+      "grad_norm": 2.019664764404297,
+      "learning_rate": 5.868405453467695e-06,
+      "loss": 0.1596,
+      "step": 3578
+    },
+    {
+      "epoch": 2.120888888888889,
+      "grad_norm": 2.3449130058288574,
+      "learning_rate": 5.864453665283541e-06,
+      "loss": 0.1142,
+      "step": 3579
+    },
+    {
+      "epoch": 2.1214814814814815,
+      "grad_norm": 1.805521845817566,
+      "learning_rate": 5.860501877099388e-06,
+      "loss": 0.1794,
+      "step": 3580
+    },
+    {
+      "epoch": 2.122074074074074,
+      "grad_norm": 2.8023102283477783,
+      "learning_rate": 5.8565500889152345e-06,
+      "loss": 0.1905,
+      "step": 3581
+    },
+    {
+      "epoch": 2.1226666666666665,
+      "grad_norm": 2.84012508392334,
+      "learning_rate": 5.852598300731082e-06,
+      "loss": 0.1701,
+      "step": 3582
+    },
+    {
+      "epoch": 2.1232592592592594,
+      "grad_norm": 1.6897846460342407,
+      "learning_rate": 5.848646512546929e-06,
+      "loss": 0.1609,
+      "step": 3583
+    },
+    {
+      "epoch": 2.123851851851852,
+      "grad_norm": 2.0900845527648926,
+      "learning_rate": 5.844694724362775e-06,
+      "loss": 0.1999,
+      "step": 3584
+    },
+    {
+      "epoch": 2.1244444444444444,
+      "grad_norm": 1.7350748777389526,
+      "learning_rate": 5.840742936178622e-06,
+      "loss": 0.1714,
+      "step": 3585
+    },
+    {
+      "epoch": 2.125037037037037,
+      "grad_norm": 2.0732452869415283,
+      "learning_rate": 5.836791147994468e-06,
+      "loss": 0.2145,
+      "step": 3586
+    },
+    {
+      "epoch": 2.1256296296296298,
+      "grad_norm": 3.1085779666900635,
+      "learning_rate": 5.832839359810315e-06,
+      "loss": 0.368,
+      "step": 3587
+    },
+    {
+      "epoch": 2.1262222222222222,
+      "grad_norm": 1.7460179328918457,
+      "learning_rate": 5.828887571626162e-06,
+      "loss": 0.2293,
+      "step": 3588
+    },
+    {
+      "epoch": 2.1268148148148147,
+      "grad_norm": 3.332646369934082,
+      "learning_rate": 5.824935783442009e-06,
+      "loss": 0.1994,
+      "step": 3589
+    },
+    {
+      "epoch": 2.127407407407407,
+      "grad_norm": 2.5412373542785645,
+      "learning_rate": 5.820983995257855e-06,
+      "loss": 0.284,
+      "step": 3590
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 2.5741305351257324,
+      "learning_rate": 5.817032207073702e-06,
+      "loss": 0.2237,
+      "step": 3591
+    },
+    {
+      "epoch": 2.1285925925925926,
+      "grad_norm": 2.158519983291626,
+      "learning_rate": 5.813080418889548e-06,
+      "loss": 0.2323,
+      "step": 3592
+    },
+    {
+      "epoch": 2.129185185185185,
+      "grad_norm": 2.6968846321105957,
+      "learning_rate": 5.809128630705395e-06,
+      "loss": 0.1682,
+      "step": 3593
+    },
+    {
+      "epoch": 2.129777777777778,
+      "grad_norm": 2.625566244125366,
+      "learning_rate": 5.805176842521242e-06,
+      "loss": 0.1965,
+      "step": 3594
+    },
+    {
+      "epoch": 2.1303703703703705,
+      "grad_norm": 1.7087246179580688,
+      "learning_rate": 5.801225054337089e-06,
+      "loss": 0.1761,
+      "step": 3595
+    },
+    {
+      "epoch": 2.130962962962963,
+      "grad_norm": 1.8509886264801025,
+      "learning_rate": 5.797273266152935e-06,
+      "loss": 0.1911,
+      "step": 3596
+    },
+    {
+      "epoch": 2.1315555555555554,
+      "grad_norm": 1.674166202545166,
+      "learning_rate": 5.793321477968782e-06,
+      "loss": 0.1807,
+      "step": 3597
+    },
+    {
+      "epoch": 2.1321481481481483,
+      "grad_norm": 2.000932216644287,
+      "learning_rate": 5.789369689784628e-06,
+      "loss": 0.2124,
+      "step": 3598
+    },
+    {
+      "epoch": 2.132740740740741,
+      "grad_norm": 1.7453861236572266,
+      "learning_rate": 5.785417901600475e-06,
+      "loss": 0.1355,
+      "step": 3599
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 2.663353681564331,
+      "learning_rate": 5.781466113416322e-06,
+      "loss": 0.2877,
+      "step": 3600
+    },
+    {
+      "epoch": 2.133925925925926,
+      "grad_norm": 1.9936517477035522,
+      "learning_rate": 5.777514325232168e-06,
+      "loss": 0.152,
+      "step": 3601
+    },
+    {
+      "epoch": 2.1345185185185187,
+      "grad_norm": 1.971954345703125,
+      "learning_rate": 5.773562537048014e-06,
+      "loss": 0.2819,
+      "step": 3602
+    },
+    {
+      "epoch": 2.135111111111111,
+      "grad_norm": 2.297029495239258,
+      "learning_rate": 5.769610748863861e-06,
+      "loss": 0.164,
+      "step": 3603
+    },
+    {
+      "epoch": 2.1357037037037037,
+      "grad_norm": 2.2827231884002686,
+      "learning_rate": 5.7656589606797074e-06,
+      "loss": 0.1771,
+      "step": 3604
+    },
+    {
+      "epoch": 2.136296296296296,
+      "grad_norm": 2.8015236854553223,
+      "learning_rate": 5.7617071724955545e-06,
+      "loss": 0.2604,
+      "step": 3605
+    },
+    {
+      "epoch": 2.136888888888889,
+      "grad_norm": 2.3313536643981934,
+      "learning_rate": 5.757755384311401e-06,
+      "loss": 0.2944,
+      "step": 3606
+    },
+    {
+      "epoch": 2.1374814814814815,
+      "grad_norm": 2.0413718223571777,
+      "learning_rate": 5.753803596127248e-06,
+      "loss": 0.1863,
+      "step": 3607
+    },
+    {
+      "epoch": 2.138074074074074,
+      "grad_norm": 2.2039005756378174,
+      "learning_rate": 5.749851807943094e-06,
+      "loss": 0.2036,
+      "step": 3608
+    },
+    {
+      "epoch": 2.1386666666666665,
+      "grad_norm": 2.2354085445404053,
+      "learning_rate": 5.745900019758941e-06,
+      "loss": 0.2203,
+      "step": 3609
+    },
+    {
+      "epoch": 2.1392592592592594,
+      "grad_norm": 3.091771364212036,
+      "learning_rate": 5.741948231574787e-06,
+      "loss": 0.2854,
+      "step": 3610
+    },
+    {
+      "epoch": 2.139851851851852,
+      "grad_norm": 1.980134129524231,
+      "learning_rate": 5.7379964433906345e-06,
+      "loss": 0.221,
+      "step": 3611
+    },
+    {
+      "epoch": 2.1404444444444444,
+      "grad_norm": 2.8522863388061523,
+      "learning_rate": 5.734044655206481e-06,
+      "loss": 0.195,
+      "step": 3612
+    },
+    {
+      "epoch": 2.141037037037037,
+      "grad_norm": 1.5446805953979492,
+      "learning_rate": 5.730092867022328e-06,
+      "loss": 0.1786,
+      "step": 3613
+    },
+    {
+      "epoch": 2.1416296296296298,
+      "grad_norm": 2.2287440299987793,
+      "learning_rate": 5.726141078838174e-06,
+      "loss": 0.205,
+      "step": 3614
+    },
+    {
+      "epoch": 2.1422222222222222,
+      "grad_norm": 1.5924639701843262,
+      "learning_rate": 5.722189290654021e-06,
+      "loss": 0.1615,
+      "step": 3615
+    },
+    {
+      "epoch": 2.1428148148148147,
+      "grad_norm": 2.562150716781616,
+      "learning_rate": 5.718237502469867e-06,
+      "loss": 0.2958,
+      "step": 3616
+    },
+    {
+      "epoch": 2.143407407407407,
+      "grad_norm": 2.5498197078704834,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.1975,
+      "step": 3617
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 1.535134196281433,
+      "learning_rate": 5.710333926101561e-06,
+      "loss": 0.1179,
+      "step": 3618
+    },
+    {
+      "epoch": 2.1445925925925926,
+      "grad_norm": 1.9937914609909058,
+      "learning_rate": 5.706382137917408e-06,
+      "loss": 0.2054,
+      "step": 3619
+    },
+    {
+      "epoch": 2.145185185185185,
+      "grad_norm": 1.9459736347198486,
+      "learning_rate": 5.702430349733254e-06,
+      "loss": 0.1597,
+      "step": 3620
+    },
+    {
+      "epoch": 2.145777777777778,
+      "grad_norm": 2.098567485809326,
+      "learning_rate": 5.698478561549101e-06,
+      "loss": 0.2295,
+      "step": 3621
+    },
+    {
+      "epoch": 2.1463703703703705,
+      "grad_norm": 2.345392942428589,
+      "learning_rate": 5.694526773364948e-06,
+      "loss": 0.1766,
+      "step": 3622
+    },
+    {
+      "epoch": 2.146962962962963,
+      "grad_norm": 3.5429983139038086,
+      "learning_rate": 5.6905749851807945e-06,
+      "loss": 0.2966,
+      "step": 3623
+    },
+    {
+      "epoch": 2.1475555555555554,
+      "grad_norm": 1.9270485639572144,
+      "learning_rate": 5.686623196996642e-06,
+      "loss": 0.2071,
+      "step": 3624
+    },
+    {
+      "epoch": 2.148148148148148,
+      "grad_norm": 3.1090288162231445,
+      "learning_rate": 5.682671408812488e-06,
+      "loss": 0.2569,
+      "step": 3625
+    },
+    {
+      "epoch": 2.148740740740741,
+      "grad_norm": 3.0199601650238037,
+      "learning_rate": 5.678719620628335e-06,
+      "loss": 0.2563,
+      "step": 3626
+    },
+    {
+      "epoch": 2.1493333333333333,
+      "grad_norm": 2.313119649887085,
+      "learning_rate": 5.674767832444181e-06,
+      "loss": 0.1815,
+      "step": 3627
+    },
+    {
+      "epoch": 2.149925925925926,
+      "grad_norm": 2.0529825687408447,
+      "learning_rate": 5.670816044260028e-06,
+      "loss": 0.3155,
+      "step": 3628
+    },
+    {
+      "epoch": 2.1505185185185187,
+      "grad_norm": 2.1089437007904053,
+      "learning_rate": 5.6668642560758745e-06,
+      "loss": 0.2052,
+      "step": 3629
+    },
+    {
+      "epoch": 2.151111111111111,
+      "grad_norm": 2.4338772296905518,
+      "learning_rate": 5.662912467891722e-06,
+      "loss": 0.3737,
+      "step": 3630
+    },
+    {
+      "epoch": 2.1517037037037037,
+      "grad_norm": 1.6722900867462158,
+      "learning_rate": 5.658960679707568e-06,
+      "loss": 0.2269,
+      "step": 3631
+    },
+    {
+      "epoch": 2.152296296296296,
+      "grad_norm": 1.832259178161621,
+      "learning_rate": 5.655008891523415e-06,
+      "loss": 0.2102,
+      "step": 3632
+    },
+    {
+      "epoch": 2.152888888888889,
+      "grad_norm": 3.069863796234131,
+      "learning_rate": 5.651057103339261e-06,
+      "loss": 0.1859,
+      "step": 3633
+    },
+    {
+      "epoch": 2.1534814814814816,
+      "grad_norm": 2.5206854343414307,
+      "learning_rate": 5.647105315155108e-06,
+      "loss": 0.1691,
+      "step": 3634
+    },
+    {
+      "epoch": 2.154074074074074,
+      "grad_norm": 2.0881881713867188,
+      "learning_rate": 5.6431535269709545e-06,
+      "loss": 0.2285,
+      "step": 3635
+    },
+    {
+      "epoch": 2.1546666666666665,
+      "grad_norm": 1.8375319242477417,
+      "learning_rate": 5.639201738786802e-06,
+      "loss": 0.1561,
+      "step": 3636
+    },
+    {
+      "epoch": 2.1552592592592594,
+      "grad_norm": 2.7607040405273438,
+      "learning_rate": 5.635249950602648e-06,
+      "loss": 0.1574,
+      "step": 3637
+    },
+    {
+      "epoch": 2.155851851851852,
+      "grad_norm": 2.5031027793884277,
+      "learning_rate": 5.631298162418495e-06,
+      "loss": 0.2448,
+      "step": 3638
+    },
+    {
+      "epoch": 2.1564444444444444,
+      "grad_norm": 1.7392041683197021,
+      "learning_rate": 5.627346374234341e-06,
+      "loss": 0.2157,
+      "step": 3639
+    },
+    {
+      "epoch": 2.157037037037037,
+      "grad_norm": 1.586578369140625,
+      "learning_rate": 5.623394586050188e-06,
+      "loss": 0.2524,
+      "step": 3640
+    },
+    {
+      "epoch": 2.15762962962963,
+      "grad_norm": 2.294314384460449,
+      "learning_rate": 5.6194427978660345e-06,
+      "loss": 0.1988,
+      "step": 3641
+    },
+    {
+      "epoch": 2.1582222222222223,
+      "grad_norm": 1.6922017335891724,
+      "learning_rate": 5.615491009681882e-06,
+      "loss": 0.138,
+      "step": 3642
+    },
+    {
+      "epoch": 2.1588148148148147,
+      "grad_norm": 2.5840394496917725,
+      "learning_rate": 5.611539221497728e-06,
+      "loss": 0.2234,
+      "step": 3643
+    },
+    {
+      "epoch": 2.159407407407407,
+      "grad_norm": 1.604444980621338,
+      "learning_rate": 5.607587433313575e-06,
+      "loss": 0.1627,
+      "step": 3644
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.552633285522461,
+      "learning_rate": 5.603635645129422e-06,
+      "loss": 0.1487,
+      "step": 3645
+    },
+    {
+      "epoch": 2.1605925925925926,
+      "grad_norm": 1.7188726663589478,
+      "learning_rate": 5.599683856945268e-06,
+      "loss": 0.2052,
+      "step": 3646
+    },
+    {
+      "epoch": 2.161185185185185,
+      "grad_norm": 1.7280807495117188,
+      "learning_rate": 5.595732068761115e-06,
+      "loss": 0.2194,
+      "step": 3647
+    },
+    {
+      "epoch": 2.1617777777777776,
+      "grad_norm": 2.2208287715911865,
+      "learning_rate": 5.591780280576962e-06,
+      "loss": 0.2359,
+      "step": 3648
+    },
+    {
+      "epoch": 2.1623703703703705,
+      "grad_norm": 3.1480612754821777,
+      "learning_rate": 5.587828492392809e-06,
+      "loss": 0.271,
+      "step": 3649
+    },
+    {
+      "epoch": 2.162962962962963,
+      "grad_norm": 2.108999252319336,
+      "learning_rate": 5.583876704208655e-06,
+      "loss": 0.1887,
+      "step": 3650
+    },
+    {
+      "epoch": 2.1635555555555555,
+      "grad_norm": 2.159137487411499,
+      "learning_rate": 5.579924916024502e-06,
+      "loss": 0.204,
+      "step": 3651
+    },
+    {
+      "epoch": 2.164148148148148,
+      "grad_norm": 1.745863437652588,
+      "learning_rate": 5.575973127840348e-06,
+      "loss": 0.1381,
+      "step": 3652
+    },
+    {
+      "epoch": 2.164740740740741,
+      "grad_norm": 2.4931187629699707,
+      "learning_rate": 5.572021339656195e-06,
+      "loss": 0.2483,
+      "step": 3653
+    },
+    {
+      "epoch": 2.1653333333333333,
+      "grad_norm": 1.8734885454177856,
+      "learning_rate": 5.568069551472042e-06,
+      "loss": 0.1693,
+      "step": 3654
+    },
+    {
+      "epoch": 2.165925925925926,
+      "grad_norm": 1.9604368209838867,
+      "learning_rate": 5.564117763287889e-06,
+      "loss": 0.1815,
+      "step": 3655
+    },
+    {
+      "epoch": 2.1665185185185187,
+      "grad_norm": 2.018303632736206,
+      "learning_rate": 5.560165975103735e-06,
+      "loss": 0.2584,
+      "step": 3656
+    },
+    {
+      "epoch": 2.167111111111111,
+      "grad_norm": 2.0094544887542725,
+      "learning_rate": 5.556214186919582e-06,
+      "loss": 0.2566,
+      "step": 3657
+    },
+    {
+      "epoch": 2.1677037037037037,
+      "grad_norm": 1.9057528972625732,
+      "learning_rate": 5.552262398735428e-06,
+      "loss": 0.1832,
+      "step": 3658
+    },
+    {
+      "epoch": 2.168296296296296,
+      "grad_norm": 2.388122797012329,
+      "learning_rate": 5.548310610551275e-06,
+      "loss": 0.2213,
+      "step": 3659
+    },
+    {
+      "epoch": 2.168888888888889,
+      "grad_norm": 1.7708079814910889,
+      "learning_rate": 5.544358822367122e-06,
+      "loss": 0.218,
+      "step": 3660
+    },
+    {
+      "epoch": 2.1694814814814816,
+      "grad_norm": 1.559786319732666,
+      "learning_rate": 5.540407034182969e-06,
+      "loss": 0.1579,
+      "step": 3661
+    },
+    {
+      "epoch": 2.170074074074074,
+      "grad_norm": 1.9712425470352173,
+      "learning_rate": 5.536455245998815e-06,
+      "loss": 0.1847,
+      "step": 3662
+    },
+    {
+      "epoch": 2.1706666666666665,
+      "grad_norm": 2.314115285873413,
+      "learning_rate": 5.532503457814662e-06,
+      "loss": 0.2562,
+      "step": 3663
+    },
+    {
+      "epoch": 2.1712592592592594,
+      "grad_norm": 2.2202086448669434,
+      "learning_rate": 5.528551669630508e-06,
+      "loss": 0.2408,
+      "step": 3664
+    },
+    {
+      "epoch": 2.171851851851852,
+      "grad_norm": 2.019721269607544,
+      "learning_rate": 5.524599881446355e-06,
+      "loss": 0.189,
+      "step": 3665
+    },
+    {
+      "epoch": 2.1724444444444444,
+      "grad_norm": 2.6216037273406982,
+      "learning_rate": 5.520648093262202e-06,
+      "loss": 0.2181,
+      "step": 3666
+    },
+    {
+      "epoch": 2.173037037037037,
+      "grad_norm": 2.3595664501190186,
+      "learning_rate": 5.516696305078049e-06,
+      "loss": 0.1958,
+      "step": 3667
+    },
+    {
+      "epoch": 2.17362962962963,
+      "grad_norm": 2.2733919620513916,
+      "learning_rate": 5.512744516893895e-06,
+      "loss": 0.2118,
+      "step": 3668
+    },
+    {
+      "epoch": 2.1742222222222223,
+      "grad_norm": 2.3581435680389404,
+      "learning_rate": 5.508792728709742e-06,
+      "loss": 0.193,
+      "step": 3669
+    },
+    {
+      "epoch": 2.1748148148148148,
+      "grad_norm": 1.5595569610595703,
+      "learning_rate": 5.504840940525589e-06,
+      "loss": 0.1684,
+      "step": 3670
+    },
+    {
+      "epoch": 2.1754074074074072,
+      "grad_norm": 1.5482864379882812,
+      "learning_rate": 5.500889152341435e-06,
+      "loss": 0.1921,
+      "step": 3671
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 2.5249083042144775,
+      "learning_rate": 5.4969373641572824e-06,
+      "loss": 0.2391,
+      "step": 3672
+    },
+    {
+      "epoch": 2.1765925925925926,
+      "grad_norm": 1.7725437879562378,
+      "learning_rate": 5.492985575973129e-06,
+      "loss": 0.2179,
+      "step": 3673
+    },
+    {
+      "epoch": 2.177185185185185,
+      "grad_norm": 3.2483086585998535,
+      "learning_rate": 5.489033787788976e-06,
+      "loss": 0.2884,
+      "step": 3674
+    },
+    {
+      "epoch": 2.1777777777777776,
+      "grad_norm": 2.135423421859741,
+      "learning_rate": 5.485081999604822e-06,
+      "loss": 0.1822,
+      "step": 3675
+    },
+    {
+      "epoch": 2.1783703703703705,
+      "grad_norm": 1.673815131187439,
+      "learning_rate": 5.481130211420669e-06,
+      "loss": 0.1692,
+      "step": 3676
+    },
+    {
+      "epoch": 2.178962962962963,
+      "grad_norm": 1.7130045890808105,
+      "learning_rate": 5.477178423236515e-06,
+      "loss": 0.2323,
+      "step": 3677
+    },
+    {
+      "epoch": 2.1795555555555555,
+      "grad_norm": 1.5098919868469238,
+      "learning_rate": 5.4732266350523624e-06,
+      "loss": 0.1618,
+      "step": 3678
+    },
+    {
+      "epoch": 2.180148148148148,
+      "grad_norm": 2.2684531211853027,
+      "learning_rate": 5.469274846868209e-06,
+      "loss": 0.3046,
+      "step": 3679
+    },
+    {
+      "epoch": 2.180740740740741,
+      "grad_norm": 2.479144334793091,
+      "learning_rate": 5.465323058684054e-06,
+      "loss": 0.1293,
+      "step": 3680
+    },
+    {
+      "epoch": 2.1813333333333333,
+      "grad_norm": 1.9417438507080078,
+      "learning_rate": 5.461371270499901e-06,
+      "loss": 0.1523,
+      "step": 3681
+    },
+    {
+      "epoch": 2.181925925925926,
+      "grad_norm": 2.2057013511657715,
+      "learning_rate": 5.457419482315747e-06,
+      "loss": 0.2191,
+      "step": 3682
+    },
+    {
+      "epoch": 2.1825185185185187,
+      "grad_norm": 2.289529323577881,
+      "learning_rate": 5.4534676941315945e-06,
+      "loss": 0.2122,
+      "step": 3683
+    },
+    {
+      "epoch": 2.1831111111111112,
+      "grad_norm": 1.874332070350647,
+      "learning_rate": 5.449515905947442e-06,
+      "loss": 0.2164,
+      "step": 3684
+    },
+    {
+      "epoch": 2.1837037037037037,
+      "grad_norm": 2.1769156455993652,
+      "learning_rate": 5.445564117763288e-06,
+      "loss": 0.2131,
+      "step": 3685
+    },
+    {
+      "epoch": 2.184296296296296,
+      "grad_norm": 1.9271364212036133,
+      "learning_rate": 5.441612329579135e-06,
+      "loss": 0.2201,
+      "step": 3686
+    },
+    {
+      "epoch": 2.1848888888888887,
+      "grad_norm": 1.7279820442199707,
+      "learning_rate": 5.437660541394981e-06,
+      "loss": 0.2018,
+      "step": 3687
+    },
+    {
+      "epoch": 2.1854814814814816,
+      "grad_norm": 2.2885468006134033,
+      "learning_rate": 5.433708753210828e-06,
+      "loss": 0.2533,
+      "step": 3688
+    },
+    {
+      "epoch": 2.186074074074074,
+      "grad_norm": 2.1100013256073,
+      "learning_rate": 5.4297569650266745e-06,
+      "loss": 0.1896,
+      "step": 3689
+    },
+    {
+      "epoch": 2.1866666666666665,
+      "grad_norm": 2.3105123043060303,
+      "learning_rate": 5.425805176842522e-06,
+      "loss": 0.2106,
+      "step": 3690
+    },
+    {
+      "epoch": 2.1872592592592595,
+      "grad_norm": 2.73861026763916,
+      "learning_rate": 5.421853388658368e-06,
+      "loss": 0.2788,
+      "step": 3691
+    },
+    {
+      "epoch": 2.187851851851852,
+      "grad_norm": 1.6925899982452393,
+      "learning_rate": 5.417901600474215e-06,
+      "loss": 0.1574,
+      "step": 3692
+    },
+    {
+      "epoch": 2.1884444444444444,
+      "grad_norm": 2.443852424621582,
+      "learning_rate": 5.413949812290061e-06,
+      "loss": 0.1925,
+      "step": 3693
+    },
+    {
+      "epoch": 2.189037037037037,
+      "grad_norm": 2.393282175064087,
+      "learning_rate": 5.409998024105908e-06,
+      "loss": 0.1969,
+      "step": 3694
+    },
+    {
+      "epoch": 2.18962962962963,
+      "grad_norm": 2.2108981609344482,
+      "learning_rate": 5.4060462359217545e-06,
+      "loss": 0.154,
+      "step": 3695
+    },
+    {
+      "epoch": 2.1902222222222223,
+      "grad_norm": 2.280555009841919,
+      "learning_rate": 5.402094447737602e-06,
+      "loss": 0.1825,
+      "step": 3696
+    },
+    {
+      "epoch": 2.1908148148148148,
+      "grad_norm": 2.8542134761810303,
+      "learning_rate": 5.398142659553448e-06,
+      "loss": 0.197,
+      "step": 3697
+    },
+    {
+      "epoch": 2.1914074074074072,
+      "grad_norm": 1.6386120319366455,
+      "learning_rate": 5.394190871369295e-06,
+      "loss": 0.1225,
+      "step": 3698
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 2.2768707275390625,
+      "learning_rate": 5.390239083185141e-06,
+      "loss": 0.2233,
+      "step": 3699
+    },
+    {
+      "epoch": 2.1925925925925926,
+      "grad_norm": 2.0767412185668945,
+      "learning_rate": 5.386287295000988e-06,
+      "loss": 0.1862,
+      "step": 3700
+    },
+    {
+      "epoch": 2.193185185185185,
+      "grad_norm": 2.8260438442230225,
+      "learning_rate": 5.3823355068168345e-06,
+      "loss": 0.2669,
+      "step": 3701
+    },
+    {
+      "epoch": 2.1937777777777776,
+      "grad_norm": 2.232069969177246,
+      "learning_rate": 5.3783837186326816e-06,
+      "loss": 0.1947,
+      "step": 3702
+    },
+    {
+      "epoch": 2.1943703703703705,
+      "grad_norm": 2.132333278656006,
+      "learning_rate": 5.374431930448528e-06,
+      "loss": 0.1953,
+      "step": 3703
+    },
+    {
+      "epoch": 2.194962962962963,
+      "grad_norm": 2.3552422523498535,
+      "learning_rate": 5.370480142264375e-06,
+      "loss": 0.1967,
+      "step": 3704
+    },
+    {
+      "epoch": 2.1955555555555555,
+      "grad_norm": 2.1524276733398438,
+      "learning_rate": 5.366528354080221e-06,
+      "loss": 0.1602,
+      "step": 3705
+    },
+    {
+      "epoch": 2.196148148148148,
+      "grad_norm": 1.9778637886047363,
+      "learning_rate": 5.362576565896068e-06,
+      "loss": 0.2217,
+      "step": 3706
+    },
+    {
+      "epoch": 2.196740740740741,
+      "grad_norm": 1.8622218370437622,
+      "learning_rate": 5.358624777711915e-06,
+      "loss": 0.2024,
+      "step": 3707
+    },
+    {
+      "epoch": 2.1973333333333334,
+      "grad_norm": 2.263998031616211,
+      "learning_rate": 5.3546729895277616e-06,
+      "loss": 0.2651,
+      "step": 3708
+    },
+    {
+      "epoch": 2.197925925925926,
+      "grad_norm": 2.2669010162353516,
+      "learning_rate": 5.350721201343609e-06,
+      "loss": 0.2068,
+      "step": 3709
+    },
+    {
+      "epoch": 2.1985185185185183,
+      "grad_norm": 2.3853280544281006,
+      "learning_rate": 5.346769413159455e-06,
+      "loss": 0.1826,
+      "step": 3710
+    },
+    {
+      "epoch": 2.1991111111111112,
+      "grad_norm": 2.5295395851135254,
+      "learning_rate": 5.342817624975302e-06,
+      "loss": 0.2269,
+      "step": 3711
+    },
+    {
+      "epoch": 2.1997037037037037,
+      "grad_norm": 2.047919273376465,
+      "learning_rate": 5.338865836791148e-06,
+      "loss": 0.1837,
+      "step": 3712
+    },
+    {
+      "epoch": 2.200296296296296,
+      "grad_norm": 2.7371275424957275,
+      "learning_rate": 5.334914048606995e-06,
+      "loss": 0.2185,
+      "step": 3713
+    },
+    {
+      "epoch": 2.2008888888888887,
+      "grad_norm": 1.9614245891571045,
+      "learning_rate": 5.3309622604228416e-06,
+      "loss": 0.1679,
+      "step": 3714
+    },
+    {
+      "epoch": 2.2014814814814816,
+      "grad_norm": 2.6006357669830322,
+      "learning_rate": 5.327010472238689e-06,
+      "loss": 0.2454,
+      "step": 3715
+    },
+    {
+      "epoch": 2.202074074074074,
+      "grad_norm": 1.7764636278152466,
+      "learning_rate": 5.323058684054535e-06,
+      "loss": 0.1743,
+      "step": 3716
+    },
+    {
+      "epoch": 2.2026666666666666,
+      "grad_norm": 1.6766822338104248,
+      "learning_rate": 5.319106895870382e-06,
+      "loss": 0.1798,
+      "step": 3717
+    },
+    {
+      "epoch": 2.2032592592592595,
+      "grad_norm": 2.0477707386016846,
+      "learning_rate": 5.315155107686228e-06,
+      "loss": 0.2177,
+      "step": 3718
+    },
+    {
+      "epoch": 2.203851851851852,
+      "grad_norm": 1.8706510066986084,
+      "learning_rate": 5.311203319502075e-06,
+      "loss": 0.2551,
+      "step": 3719
+    },
+    {
+      "epoch": 2.2044444444444444,
+      "grad_norm": 2.4376509189605713,
+      "learning_rate": 5.3072515313179216e-06,
+      "loss": 0.184,
+      "step": 3720
+    },
+    {
+      "epoch": 2.205037037037037,
+      "grad_norm": 1.7228279113769531,
+      "learning_rate": 5.303299743133769e-06,
+      "loss": 0.192,
+      "step": 3721
+    },
+    {
+      "epoch": 2.20562962962963,
+      "grad_norm": 2.6055047512054443,
+      "learning_rate": 5.299347954949615e-06,
+      "loss": 0.2668,
+      "step": 3722
+    },
+    {
+      "epoch": 2.2062222222222223,
+      "grad_norm": 1.982017993927002,
+      "learning_rate": 5.295396166765462e-06,
+      "loss": 0.2011,
+      "step": 3723
+    },
+    {
+      "epoch": 2.206814814814815,
+      "grad_norm": 2.205735206604004,
+      "learning_rate": 5.291444378581308e-06,
+      "loss": 0.2342,
+      "step": 3724
+    },
+    {
+      "epoch": 2.2074074074074073,
+      "grad_norm": 3.001981496810913,
+      "learning_rate": 5.287492590397155e-06,
+      "loss": 0.3109,
+      "step": 3725
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 2.7915308475494385,
+      "learning_rate": 5.2835408022130016e-06,
+      "loss": 0.2093,
+      "step": 3726
+    },
+    {
+      "epoch": 2.2085925925925927,
+      "grad_norm": 2.300628423690796,
+      "learning_rate": 5.279589014028849e-06,
+      "loss": 0.2575,
+      "step": 3727
+    },
+    {
+      "epoch": 2.209185185185185,
+      "grad_norm": 1.800301432609558,
+      "learning_rate": 5.275637225844695e-06,
+      "loss": 0.179,
+      "step": 3728
+    },
+    {
+      "epoch": 2.2097777777777776,
+      "grad_norm": 2.522925853729248,
+      "learning_rate": 5.271685437660542e-06,
+      "loss": 0.1578,
+      "step": 3729
+    },
+    {
+      "epoch": 2.2103703703703705,
+      "grad_norm": 2.3060336112976074,
+      "learning_rate": 5.267733649476388e-06,
+      "loss": 0.1393,
+      "step": 3730
+    },
+    {
+      "epoch": 2.210962962962963,
+      "grad_norm": 1.8616286516189575,
+      "learning_rate": 5.263781861292235e-06,
+      "loss": 0.1925,
+      "step": 3731
+    },
+    {
+      "epoch": 2.2115555555555555,
+      "grad_norm": 1.8687642812728882,
+      "learning_rate": 5.259830073108082e-06,
+      "loss": 0.2483,
+      "step": 3732
+    },
+    {
+      "epoch": 2.212148148148148,
+      "grad_norm": 1.527543306350708,
+      "learning_rate": 5.255878284923929e-06,
+      "loss": 0.1721,
+      "step": 3733
+    },
+    {
+      "epoch": 2.212740740740741,
+      "grad_norm": 2.0280439853668213,
+      "learning_rate": 5.251926496739776e-06,
+      "loss": 0.1494,
+      "step": 3734
+    },
+    {
+      "epoch": 2.2133333333333334,
+      "grad_norm": 2.386495351791382,
+      "learning_rate": 5.247974708555622e-06,
+      "loss": 0.2326,
+      "step": 3735
+    },
+    {
+      "epoch": 2.213925925925926,
+      "grad_norm": 2.189265489578247,
+      "learning_rate": 5.244022920371469e-06,
+      "loss": 0.2008,
+      "step": 3736
+    },
+    {
+      "epoch": 2.2145185185185183,
+      "grad_norm": 2.2230775356292725,
+      "learning_rate": 5.240071132187315e-06,
+      "loss": 0.1153,
+      "step": 3737
+    },
+    {
+      "epoch": 2.2151111111111113,
+      "grad_norm": 1.6599135398864746,
+      "learning_rate": 5.236119344003162e-06,
+      "loss": 0.1788,
+      "step": 3738
+    },
+    {
+      "epoch": 2.2157037037037037,
+      "grad_norm": 1.939400315284729,
+      "learning_rate": 5.232167555819009e-06,
+      "loss": 0.1739,
+      "step": 3739
+    },
+    {
+      "epoch": 2.216296296296296,
+      "grad_norm": 1.9926320314407349,
+      "learning_rate": 5.228215767634856e-06,
+      "loss": 0.2464,
+      "step": 3740
+    },
+    {
+      "epoch": 2.2168888888888887,
+      "grad_norm": 1.802361011505127,
+      "learning_rate": 5.224263979450702e-06,
+      "loss": 0.1824,
+      "step": 3741
+    },
+    {
+      "epoch": 2.2174814814814816,
+      "grad_norm": 2.036846876144409,
+      "learning_rate": 5.220312191266549e-06,
+      "loss": 0.1697,
+      "step": 3742
+    },
+    {
+      "epoch": 2.218074074074074,
+      "grad_norm": 2.6075432300567627,
+      "learning_rate": 5.216360403082395e-06,
+      "loss": 0.2109,
+      "step": 3743
+    },
+    {
+      "epoch": 2.2186666666666666,
+      "grad_norm": 2.0259382724761963,
+      "learning_rate": 5.212408614898242e-06,
+      "loss": 0.232,
+      "step": 3744
+    },
+    {
+      "epoch": 2.2192592592592595,
+      "grad_norm": 2.125795364379883,
+      "learning_rate": 5.208456826714089e-06,
+      "loss": 0.2354,
+      "step": 3745
+    },
+    {
+      "epoch": 2.219851851851852,
+      "grad_norm": 2.2485196590423584,
+      "learning_rate": 5.204505038529936e-06,
+      "loss": 0.1388,
+      "step": 3746
+    },
+    {
+      "epoch": 2.2204444444444444,
+      "grad_norm": 2.414804458618164,
+      "learning_rate": 5.200553250345782e-06,
+      "loss": 0.3221,
+      "step": 3747
+    },
+    {
+      "epoch": 2.221037037037037,
+      "grad_norm": 2.767051935195923,
+      "learning_rate": 5.196601462161629e-06,
+      "loss": 0.2599,
+      "step": 3748
+    },
+    {
+      "epoch": 2.2216296296296294,
+      "grad_norm": 2.365813970565796,
+      "learning_rate": 5.192649673977475e-06,
+      "loss": 0.2102,
+      "step": 3749
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 1.8578526973724365,
+      "learning_rate": 5.188697885793322e-06,
+      "loss": 0.1526,
+      "step": 3750
+    },
+    {
+      "epoch": 2.222814814814815,
+      "grad_norm": 2.253699541091919,
+      "learning_rate": 5.184746097609169e-06,
+      "loss": 0.1745,
+      "step": 3751
+    },
+    {
+      "epoch": 2.2234074074074073,
+      "grad_norm": 2.2167935371398926,
+      "learning_rate": 5.180794309425016e-06,
+      "loss": 0.2635,
+      "step": 3752
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 2.291055202484131,
+      "learning_rate": 5.176842521240862e-06,
+      "loss": 0.178,
+      "step": 3753
+    },
+    {
+      "epoch": 2.2245925925925927,
+      "grad_norm": 2.014711380004883,
+      "learning_rate": 5.172890733056709e-06,
+      "loss": 0.1413,
+      "step": 3754
+    },
+    {
+      "epoch": 2.225185185185185,
+      "grad_norm": 1.9063568115234375,
+      "learning_rate": 5.168938944872556e-06,
+      "loss": 0.1698,
+      "step": 3755
+    },
+    {
+      "epoch": 2.2257777777777776,
+      "grad_norm": 1.8492164611816406,
+      "learning_rate": 5.164987156688402e-06,
+      "loss": 0.1775,
+      "step": 3756
+    },
+    {
+      "epoch": 2.2263703703703706,
+      "grad_norm": 2.6380412578582764,
+      "learning_rate": 5.1610353685042495e-06,
+      "loss": 0.257,
+      "step": 3757
+    },
+    {
+      "epoch": 2.226962962962963,
+      "grad_norm": 2.786858320236206,
+      "learning_rate": 5.157083580320096e-06,
+      "loss": 0.178,
+      "step": 3758
+    },
+    {
+      "epoch": 2.2275555555555555,
+      "grad_norm": 1.9536372423171997,
+      "learning_rate": 5.153131792135941e-06,
+      "loss": 0.1955,
+      "step": 3759
+    },
+    {
+      "epoch": 2.228148148148148,
+      "grad_norm": 1.3475526571273804,
+      "learning_rate": 5.149180003951788e-06,
+      "loss": 0.109,
+      "step": 3760
+    },
+    {
+      "epoch": 2.228740740740741,
+      "grad_norm": 1.7687084674835205,
+      "learning_rate": 5.1452282157676345e-06,
+      "loss": 0.1777,
+      "step": 3761
+    },
+    {
+      "epoch": 2.2293333333333334,
+      "grad_norm": 1.7319815158843994,
+      "learning_rate": 5.1412764275834816e-06,
+      "loss": 0.1523,
+      "step": 3762
+    },
+    {
+      "epoch": 2.229925925925926,
+      "grad_norm": 1.8801072835922241,
+      "learning_rate": 5.137324639399328e-06,
+      "loss": 0.2028,
+      "step": 3763
+    },
+    {
+      "epoch": 2.2305185185185183,
+      "grad_norm": 1.6354634761810303,
+      "learning_rate": 5.133372851215175e-06,
+      "loss": 0.2068,
+      "step": 3764
+    },
+    {
+      "epoch": 2.2311111111111113,
+      "grad_norm": 2.764622449874878,
+      "learning_rate": 5.129421063031021e-06,
+      "loss": 0.2714,
+      "step": 3765
+    },
+    {
+      "epoch": 2.2317037037037037,
+      "grad_norm": 2.2763636112213135,
+      "learning_rate": 5.125469274846868e-06,
+      "loss": 0.2578,
+      "step": 3766
+    },
+    {
+      "epoch": 2.2322962962962962,
+      "grad_norm": 2.11308217048645,
+      "learning_rate": 5.1215174866627145e-06,
+      "loss": 0.1795,
+      "step": 3767
+    },
+    {
+      "epoch": 2.2328888888888887,
+      "grad_norm": 2.3310868740081787,
+      "learning_rate": 5.1175656984785616e-06,
+      "loss": 0.2099,
+      "step": 3768
+    },
+    {
+      "epoch": 2.2334814814814816,
+      "grad_norm": 1.627609372138977,
+      "learning_rate": 5.113613910294409e-06,
+      "loss": 0.191,
+      "step": 3769
+    },
+    {
+      "epoch": 2.234074074074074,
+      "grad_norm": 2.1197214126586914,
+      "learning_rate": 5.109662122110255e-06,
+      "loss": 0.2233,
+      "step": 3770
+    },
+    {
+      "epoch": 2.2346666666666666,
+      "grad_norm": 1.7707817554473877,
+      "learning_rate": 5.105710333926102e-06,
+      "loss": 0.1437,
+      "step": 3771
+    },
+    {
+      "epoch": 2.235259259259259,
+      "grad_norm": 2.189764976501465,
+      "learning_rate": 5.101758545741948e-06,
+      "loss": 0.1805,
+      "step": 3772
+    },
+    {
+      "epoch": 2.235851851851852,
+      "grad_norm": 4.174278736114502,
+      "learning_rate": 5.097806757557795e-06,
+      "loss": 0.259,
+      "step": 3773
+    },
+    {
+      "epoch": 2.2364444444444445,
+      "grad_norm": 1.5983021259307861,
+      "learning_rate": 5.0938549693736416e-06,
+      "loss": 0.1245,
+      "step": 3774
+    },
+    {
+      "epoch": 2.237037037037037,
+      "grad_norm": 1.9467685222625732,
+      "learning_rate": 5.089903181189489e-06,
+      "loss": 0.1473,
+      "step": 3775
+    },
+    {
+      "epoch": 2.2376296296296294,
+      "grad_norm": 2.284013509750366,
+      "learning_rate": 5.085951393005335e-06,
+      "loss": 0.1829,
+      "step": 3776
+    },
+    {
+      "epoch": 2.2382222222222223,
+      "grad_norm": 2.2971913814544678,
+      "learning_rate": 5.081999604821182e-06,
+      "loss": 0.1807,
+      "step": 3777
+    },
+    {
+      "epoch": 2.238814814814815,
+      "grad_norm": 1.8395458459854126,
+      "learning_rate": 5.078047816637028e-06,
+      "loss": 0.1541,
+      "step": 3778
+    },
+    {
+      "epoch": 2.2394074074074073,
+      "grad_norm": 2.8145358562469482,
+      "learning_rate": 5.074096028452875e-06,
+      "loss": 0.2192,
+      "step": 3779
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.04508900642395,
+      "learning_rate": 5.0701442402687216e-06,
+      "loss": 0.2616,
+      "step": 3780
+    },
+    {
+      "epoch": 2.2405925925925927,
+      "grad_norm": 2.0508222579956055,
+      "learning_rate": 5.066192452084569e-06,
+      "loss": 0.1914,
+      "step": 3781
+    },
+    {
+      "epoch": 2.241185185185185,
+      "grad_norm": 1.9596792459487915,
+      "learning_rate": 5.062240663900415e-06,
+      "loss": 0.2067,
+      "step": 3782
+    },
+    {
+      "epoch": 2.2417777777777776,
+      "grad_norm": 2.186427593231201,
+      "learning_rate": 5.058288875716262e-06,
+      "loss": 0.2089,
+      "step": 3783
+    },
+    {
+      "epoch": 2.2423703703703706,
+      "grad_norm": 1.8678196668624878,
+      "learning_rate": 5.054337087532108e-06,
+      "loss": 0.1903,
+      "step": 3784
+    },
+    {
+      "epoch": 2.242962962962963,
+      "grad_norm": 2.258607864379883,
+      "learning_rate": 5.050385299347955e-06,
+      "loss": 0.2022,
+      "step": 3785
+    },
+    {
+      "epoch": 2.2435555555555555,
+      "grad_norm": 2.1284902095794678,
+      "learning_rate": 5.0464335111638015e-06,
+      "loss": 0.2131,
+      "step": 3786
+    },
+    {
+      "epoch": 2.244148148148148,
+      "grad_norm": 2.5619163513183594,
+      "learning_rate": 5.042481722979649e-06,
+      "loss": 0.2126,
+      "step": 3787
+    },
+    {
+      "epoch": 2.244740740740741,
+      "grad_norm": 1.954166293144226,
+      "learning_rate": 5.038529934795495e-06,
+      "loss": 0.1548,
+      "step": 3788
+    },
+    {
+      "epoch": 2.2453333333333334,
+      "grad_norm": 2.4128129482269287,
+      "learning_rate": 5.034578146611342e-06,
+      "loss": 0.2339,
+      "step": 3789
+    },
+    {
+      "epoch": 2.245925925925926,
+      "grad_norm": 1.4658982753753662,
+      "learning_rate": 5.030626358427188e-06,
+      "loss": 0.1812,
+      "step": 3790
+    },
+    {
+      "epoch": 2.2465185185185184,
+      "grad_norm": 2.022852897644043,
+      "learning_rate": 5.026674570243035e-06,
+      "loss": 0.1817,
+      "step": 3791
+    },
+    {
+      "epoch": 2.2471111111111113,
+      "grad_norm": 2.4327402114868164,
+      "learning_rate": 5.022722782058882e-06,
+      "loss": 0.195,
+      "step": 3792
+    },
+    {
+      "epoch": 2.2477037037037038,
+      "grad_norm": 3.1769678592681885,
+      "learning_rate": 5.018770993874729e-06,
+      "loss": 0.2105,
+      "step": 3793
+    },
+    {
+      "epoch": 2.2482962962962962,
+      "grad_norm": 1.845606803894043,
+      "learning_rate": 5.014819205690576e-06,
+      "loss": 0.2111,
+      "step": 3794
+    },
+    {
+      "epoch": 2.2488888888888887,
+      "grad_norm": 1.9784932136535645,
+      "learning_rate": 5.010867417506422e-06,
+      "loss": 0.1623,
+      "step": 3795
+    },
+    {
+      "epoch": 2.2494814814814816,
+      "grad_norm": 2.251276731491089,
+      "learning_rate": 5.006915629322269e-06,
+      "loss": 0.1878,
+      "step": 3796
+    },
+    {
+      "epoch": 2.250074074074074,
+      "grad_norm": 2.34645938873291,
+      "learning_rate": 5.002963841138115e-06,
+      "loss": 0.1309,
+      "step": 3797
+    },
+    {
+      "epoch": 2.2506666666666666,
+      "grad_norm": 2.33489990234375,
+      "learning_rate": 4.999012052953962e-06,
+      "loss": 0.2871,
+      "step": 3798
+    },
+    {
+      "epoch": 2.251259259259259,
+      "grad_norm": 2.4078500270843506,
+      "learning_rate": 4.995060264769809e-06,
+      "loss": 0.2279,
+      "step": 3799
+    },
+    {
+      "epoch": 2.251851851851852,
+      "grad_norm": 2.1309449672698975,
+      "learning_rate": 4.991108476585656e-06,
+      "loss": 0.2043,
+      "step": 3800
+    },
+    {
+      "epoch": 2.2524444444444445,
+      "grad_norm": 1.9703147411346436,
+      "learning_rate": 4.987156688401502e-06,
+      "loss": 0.2055,
+      "step": 3801
+    },
+    {
+      "epoch": 2.253037037037037,
+      "grad_norm": 1.6915254592895508,
+      "learning_rate": 4.983204900217349e-06,
+      "loss": 0.1506,
+      "step": 3802
+    },
+    {
+      "epoch": 2.2536296296296294,
+      "grad_norm": 1.995257019996643,
+      "learning_rate": 4.979253112033195e-06,
+      "loss": 0.2075,
+      "step": 3803
+    },
+    {
+      "epoch": 2.2542222222222223,
+      "grad_norm": 2.411566972732544,
+      "learning_rate": 4.975301323849042e-06,
+      "loss": 0.2014,
+      "step": 3804
+    },
+    {
+      "epoch": 2.254814814814815,
+      "grad_norm": 3.598682165145874,
+      "learning_rate": 4.971349535664889e-06,
+      "loss": 0.2389,
+      "step": 3805
+    },
+    {
+      "epoch": 2.2554074074074073,
+      "grad_norm": 2.6143569946289062,
+      "learning_rate": 4.967397747480736e-06,
+      "loss": 0.2185,
+      "step": 3806
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 1.638625144958496,
+      "learning_rate": 4.963445959296582e-06,
+      "loss": 0.1598,
+      "step": 3807
+    },
+    {
+      "epoch": 2.2565925925925927,
+      "grad_norm": 2.138460874557495,
+      "learning_rate": 4.959494171112429e-06,
+      "loss": 0.1815,
+      "step": 3808
+    },
+    {
+      "epoch": 2.257185185185185,
+      "grad_norm": 1.9625184535980225,
+      "learning_rate": 4.955542382928275e-06,
+      "loss": 0.1777,
+      "step": 3809
+    },
+    {
+      "epoch": 2.2577777777777777,
+      "grad_norm": 1.947710633277893,
+      "learning_rate": 4.951590594744122e-06,
+      "loss": 0.1873,
+      "step": 3810
+    },
+    {
+      "epoch": 2.25837037037037,
+      "grad_norm": 2.7207846641540527,
+      "learning_rate": 4.947638806559969e-06,
+      "loss": 0.193,
+      "step": 3811
+    },
+    {
+      "epoch": 2.258962962962963,
+      "grad_norm": 2.1969916820526123,
+      "learning_rate": 4.943687018375816e-06,
+      "loss": 0.3726,
+      "step": 3812
+    },
+    {
+      "epoch": 2.2595555555555555,
+      "grad_norm": 2.7828145027160645,
+      "learning_rate": 4.939735230191662e-06,
+      "loss": 0.1624,
+      "step": 3813
+    },
+    {
+      "epoch": 2.260148148148148,
+      "grad_norm": 1.9634194374084473,
+      "learning_rate": 4.935783442007509e-06,
+      "loss": 0.1799,
+      "step": 3814
+    },
+    {
+      "epoch": 2.260740740740741,
+      "grad_norm": 2.228282928466797,
+      "learning_rate": 4.931831653823355e-06,
+      "loss": 0.182,
+      "step": 3815
+    },
+    {
+      "epoch": 2.2613333333333334,
+      "grad_norm": 1.9828801155090332,
+      "learning_rate": 4.927879865639202e-06,
+      "loss": 0.1624,
+      "step": 3816
+    },
+    {
+      "epoch": 2.261925925925926,
+      "grad_norm": 1.9889986515045166,
+      "learning_rate": 4.9239280774550495e-06,
+      "loss": 0.2616,
+      "step": 3817
+    },
+    {
+      "epoch": 2.2625185185185184,
+      "grad_norm": 1.7596163749694824,
+      "learning_rate": 4.919976289270895e-06,
+      "loss": 0.174,
+      "step": 3818
+    },
+    {
+      "epoch": 2.2631111111111113,
+      "grad_norm": 1.853547215461731,
+      "learning_rate": 4.916024501086742e-06,
+      "loss": 0.1725,
+      "step": 3819
+    },
+    {
+      "epoch": 2.2637037037037038,
+      "grad_norm": 1.828109622001648,
+      "learning_rate": 4.912072712902588e-06,
+      "loss": 0.1406,
+      "step": 3820
+    },
+    {
+      "epoch": 2.2642962962962963,
+      "grad_norm": 2.0931057929992676,
+      "learning_rate": 4.908120924718435e-06,
+      "loss": 0.1787,
+      "step": 3821
+    },
+    {
+      "epoch": 2.2648888888888887,
+      "grad_norm": 2.1513986587524414,
+      "learning_rate": 4.9041691365342815e-06,
+      "loss": 0.1913,
+      "step": 3822
+    },
+    {
+      "epoch": 2.2654814814814817,
+      "grad_norm": 1.852495789527893,
+      "learning_rate": 4.900217348350129e-06,
+      "loss": 0.1882,
+      "step": 3823
+    },
+    {
+      "epoch": 2.266074074074074,
+      "grad_norm": 2.3051204681396484,
+      "learning_rate": 4.896265560165976e-06,
+      "loss": 0.1375,
+      "step": 3824
+    },
+    {
+      "epoch": 2.2666666666666666,
+      "grad_norm": 2.0297904014587402,
+      "learning_rate": 4.892313771981822e-06,
+      "loss": 0.2148,
+      "step": 3825
+    },
+    {
+      "epoch": 2.267259259259259,
+      "grad_norm": 2.991475820541382,
+      "learning_rate": 4.888361983797669e-06,
+      "loss": 0.2185,
+      "step": 3826
+    },
+    {
+      "epoch": 2.267851851851852,
+      "grad_norm": 2.4037935733795166,
+      "learning_rate": 4.884410195613515e-06,
+      "loss": 0.1942,
+      "step": 3827
+    },
+    {
+      "epoch": 2.2684444444444445,
+      "grad_norm": 1.997601866722107,
+      "learning_rate": 4.880458407429362e-06,
+      "loss": 0.1514,
+      "step": 3828
+    },
+    {
+      "epoch": 2.269037037037037,
+      "grad_norm": 1.6891758441925049,
+      "learning_rate": 4.876506619245209e-06,
+      "loss": 0.1412,
+      "step": 3829
+    },
+    {
+      "epoch": 2.2696296296296294,
+      "grad_norm": 2.0039258003234863,
+      "learning_rate": 4.872554831061056e-06,
+      "loss": 0.1559,
+      "step": 3830
+    },
+    {
+      "epoch": 2.2702222222222224,
+      "grad_norm": 2.4039483070373535,
+      "learning_rate": 4.868603042876902e-06,
+      "loss": 0.1924,
+      "step": 3831
+    },
+    {
+      "epoch": 2.270814814814815,
+      "grad_norm": 3.061664581298828,
+      "learning_rate": 4.864651254692749e-06,
+      "loss": 0.1943,
+      "step": 3832
+    },
+    {
+      "epoch": 2.2714074074074073,
+      "grad_norm": 1.8178247213363647,
+      "learning_rate": 4.860699466508595e-06,
+      "loss": 0.2309,
+      "step": 3833
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 2.1719141006469727,
+      "learning_rate": 4.856747678324442e-06,
+      "loss": 0.1832,
+      "step": 3834
+    },
+    {
+      "epoch": 2.2725925925925927,
+      "grad_norm": 2.1511359214782715,
+      "learning_rate": 4.852795890140289e-06,
+      "loss": 0.1779,
+      "step": 3835
+    },
+    {
+      "epoch": 2.273185185185185,
+      "grad_norm": 1.5954891443252563,
+      "learning_rate": 4.848844101956136e-06,
+      "loss": 0.1508,
+      "step": 3836
+    },
+    {
+      "epoch": 2.2737777777777777,
+      "grad_norm": 2.364960193634033,
+      "learning_rate": 4.844892313771982e-06,
+      "loss": 0.181,
+      "step": 3837
+    },
+    {
+      "epoch": 2.27437037037037,
+      "grad_norm": 2.059490203857422,
+      "learning_rate": 4.840940525587829e-06,
+      "loss": 0.1908,
+      "step": 3838
+    },
+    {
+      "epoch": 2.274962962962963,
+      "grad_norm": 2.467972993850708,
+      "learning_rate": 4.836988737403675e-06,
+      "loss": 0.2566,
+      "step": 3839
+    },
+    {
+      "epoch": 2.2755555555555556,
+      "grad_norm": 2.466747522354126,
+      "learning_rate": 4.833036949219522e-06,
+      "loss": 0.2126,
+      "step": 3840
+    },
+    {
+      "epoch": 2.276148148148148,
+      "grad_norm": 2.630584955215454,
+      "learning_rate": 4.829085161035369e-06,
+      "loss": 0.2629,
+      "step": 3841
+    },
+    {
+      "epoch": 2.276740740740741,
+      "grad_norm": 2.310358762741089,
+      "learning_rate": 4.825133372851216e-06,
+      "loss": 0.2034,
+      "step": 3842
+    },
+    {
+      "epoch": 2.2773333333333334,
+      "grad_norm": 2.2148029804229736,
+      "learning_rate": 4.821181584667062e-06,
+      "loss": 0.2174,
+      "step": 3843
+    },
+    {
+      "epoch": 2.277925925925926,
+      "grad_norm": 1.9453212022781372,
+      "learning_rate": 4.817229796482909e-06,
+      "loss": 0.1466,
+      "step": 3844
+    },
+    {
+      "epoch": 2.2785185185185184,
+      "grad_norm": 1.762149453163147,
+      "learning_rate": 4.813278008298755e-06,
+      "loss": 0.1956,
+      "step": 3845
+    },
+    {
+      "epoch": 2.279111111111111,
+      "grad_norm": 2.313040018081665,
+      "learning_rate": 4.809326220114602e-06,
+      "loss": 0.1799,
+      "step": 3846
+    },
+    {
+      "epoch": 2.279703703703704,
+      "grad_norm": 1.8406537771224976,
+      "learning_rate": 4.8053744319304495e-06,
+      "loss": 0.2217,
+      "step": 3847
+    },
+    {
+      "epoch": 2.2802962962962963,
+      "grad_norm": 1.6775718927383423,
+      "learning_rate": 4.801422643746296e-06,
+      "loss": 0.1678,
+      "step": 3848
+    },
+    {
+      "epoch": 2.2808888888888887,
+      "grad_norm": 1.8512030839920044,
+      "learning_rate": 4.797470855562143e-06,
+      "loss": 0.1767,
+      "step": 3849
+    },
+    {
+      "epoch": 2.2814814814814817,
+      "grad_norm": 1.9720958471298218,
+      "learning_rate": 4.793519067377989e-06,
+      "loss": 0.1573,
+      "step": 3850
+    },
+    {
+      "epoch": 2.282074074074074,
+      "grad_norm": 2.2439403533935547,
+      "learning_rate": 4.789567279193836e-06,
+      "loss": 0.1389,
+      "step": 3851
+    },
+    {
+      "epoch": 2.2826666666666666,
+      "grad_norm": 1.730584740638733,
+      "learning_rate": 4.785615491009682e-06,
+      "loss": 0.167,
+      "step": 3852
+    },
+    {
+      "epoch": 2.283259259259259,
+      "grad_norm": 2.0175139904022217,
+      "learning_rate": 4.7816637028255295e-06,
+      "loss": 0.2288,
+      "step": 3853
+    },
+    {
+      "epoch": 2.283851851851852,
+      "grad_norm": 1.4321789741516113,
+      "learning_rate": 4.777711914641376e-06,
+      "loss": 0.1466,
+      "step": 3854
+    },
+    {
+      "epoch": 2.2844444444444445,
+      "grad_norm": 2.052764654159546,
+      "learning_rate": 4.773760126457223e-06,
+      "loss": 0.1633,
+      "step": 3855
+    },
+    {
+      "epoch": 2.285037037037037,
+      "grad_norm": 2.136003017425537,
+      "learning_rate": 4.769808338273069e-06,
+      "loss": 0.1272,
+      "step": 3856
+    },
+    {
+      "epoch": 2.2856296296296295,
+      "grad_norm": 1.6906026601791382,
+      "learning_rate": 4.765856550088916e-06,
+      "loss": 0.1885,
+      "step": 3857
+    },
+    {
+      "epoch": 2.2862222222222224,
+      "grad_norm": 1.8479292392730713,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.187,
+      "step": 3858
+    },
+    {
+      "epoch": 2.286814814814815,
+      "grad_norm": 2.0483200550079346,
+      "learning_rate": 4.757952973720609e-06,
+      "loss": 0.1729,
+      "step": 3859
+    },
+    {
+      "epoch": 2.2874074074074073,
+      "grad_norm": 2.185852289199829,
+      "learning_rate": 4.754001185536456e-06,
+      "loss": 0.3464,
+      "step": 3860
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 2.207460880279541,
+      "learning_rate": 4.750049397352302e-06,
+      "loss": 0.2513,
+      "step": 3861
+    },
+    {
+      "epoch": 2.2885925925925927,
+      "grad_norm": 1.882690191268921,
+      "learning_rate": 4.746097609168149e-06,
+      "loss": 0.1667,
+      "step": 3862
+    },
+    {
+      "epoch": 2.289185185185185,
+      "grad_norm": 2.254002094268799,
+      "learning_rate": 4.742145820983995e-06,
+      "loss": 0.1799,
+      "step": 3863
+    },
+    {
+      "epoch": 2.2897777777777777,
+      "grad_norm": 2.3053503036499023,
+      "learning_rate": 4.738194032799842e-06,
+      "loss": 0.1739,
+      "step": 3864
+    },
+    {
+      "epoch": 2.29037037037037,
+      "grad_norm": 4.591277599334717,
+      "learning_rate": 4.734242244615689e-06,
+      "loss": 0.3668,
+      "step": 3865
+    },
+    {
+      "epoch": 2.290962962962963,
+      "grad_norm": 1.5380115509033203,
+      "learning_rate": 4.730290456431536e-06,
+      "loss": 0.1915,
+      "step": 3866
+    },
+    {
+      "epoch": 2.2915555555555556,
+      "grad_norm": 2.7418806552886963,
+      "learning_rate": 4.726338668247382e-06,
+      "loss": 0.2134,
+      "step": 3867
+    },
+    {
+      "epoch": 2.292148148148148,
+      "grad_norm": 1.6568118333816528,
+      "learning_rate": 4.722386880063229e-06,
+      "loss": 0.1425,
+      "step": 3868
+    },
+    {
+      "epoch": 2.292740740740741,
+      "grad_norm": 1.5352150201797485,
+      "learning_rate": 4.718435091879075e-06,
+      "loss": 0.1949,
+      "step": 3869
+    },
+    {
+      "epoch": 2.2933333333333334,
+      "grad_norm": 1.7663222551345825,
+      "learning_rate": 4.714483303694922e-06,
+      "loss": 0.306,
+      "step": 3870
+    },
+    {
+      "epoch": 2.293925925925926,
+      "grad_norm": 2.3892364501953125,
+      "learning_rate": 4.710531515510769e-06,
+      "loss": 0.3188,
+      "step": 3871
+    },
+    {
+      "epoch": 2.2945185185185184,
+      "grad_norm": 3.0626468658447266,
+      "learning_rate": 4.706579727326616e-06,
+      "loss": 0.1807,
+      "step": 3872
+    },
+    {
+      "epoch": 2.295111111111111,
+      "grad_norm": 3.74135684967041,
+      "learning_rate": 4.702627939142462e-06,
+      "loss": 0.1953,
+      "step": 3873
+    },
+    {
+      "epoch": 2.295703703703704,
+      "grad_norm": 1.783265471458435,
+      "learning_rate": 4.698676150958309e-06,
+      "loss": 0.1911,
+      "step": 3874
+    },
+    {
+      "epoch": 2.2962962962962963,
+      "grad_norm": 2.3043906688690186,
+      "learning_rate": 4.694724362774155e-06,
+      "loss": 0.2071,
+      "step": 3875
+    },
+    {
+      "epoch": 2.2968888888888888,
+      "grad_norm": 3.0500855445861816,
+      "learning_rate": 4.690772574590002e-06,
+      "loss": 0.2593,
+      "step": 3876
+    },
+    {
+      "epoch": 2.2974814814814817,
+      "grad_norm": 3.3497331142425537,
+      "learning_rate": 4.686820786405849e-06,
+      "loss": 0.3018,
+      "step": 3877
+    },
+    {
+      "epoch": 2.298074074074074,
+      "grad_norm": 2.2327804565429688,
+      "learning_rate": 4.682868998221696e-06,
+      "loss": 0.1887,
+      "step": 3878
+    },
+    {
+      "epoch": 2.2986666666666666,
+      "grad_norm": 1.9907739162445068,
+      "learning_rate": 4.678917210037543e-06,
+      "loss": 0.1367,
+      "step": 3879
+    },
+    {
+      "epoch": 2.299259259259259,
+      "grad_norm": 2.197568893432617,
+      "learning_rate": 4.674965421853389e-06,
+      "loss": 0.2296,
+      "step": 3880
+    },
+    {
+      "epoch": 2.299851851851852,
+      "grad_norm": 2.583909273147583,
+      "learning_rate": 4.671013633669236e-06,
+      "loss": 0.2262,
+      "step": 3881
+    },
+    {
+      "epoch": 2.3004444444444445,
+      "grad_norm": 2.4725706577301025,
+      "learning_rate": 4.667061845485082e-06,
+      "loss": 0.1936,
+      "step": 3882
+    },
+    {
+      "epoch": 2.301037037037037,
+      "grad_norm": 2.048927068710327,
+      "learning_rate": 4.6631100573009295e-06,
+      "loss": 0.2373,
+      "step": 3883
+    },
+    {
+      "epoch": 2.3016296296296295,
+      "grad_norm": 3.7143030166625977,
+      "learning_rate": 4.659158269116776e-06,
+      "loss": 0.2681,
+      "step": 3884
+    },
+    {
+      "epoch": 2.3022222222222224,
+      "grad_norm": 2.4559202194213867,
+      "learning_rate": 4.655206480932623e-06,
+      "loss": 0.167,
+      "step": 3885
+    },
+    {
+      "epoch": 2.302814814814815,
+      "grad_norm": 3.3731253147125244,
+      "learning_rate": 4.651254692748469e-06,
+      "loss": 0.1808,
+      "step": 3886
+    },
+    {
+      "epoch": 2.3034074074074073,
+      "grad_norm": 2.6064376831054688,
+      "learning_rate": 4.647302904564316e-06,
+      "loss": 0.2101,
+      "step": 3887
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 3.9235870838165283,
+      "learning_rate": 4.643351116380162e-06,
+      "loss": 0.3657,
+      "step": 3888
+    },
+    {
+      "epoch": 2.3045925925925927,
+      "grad_norm": 2.0200655460357666,
+      "learning_rate": 4.6393993281960095e-06,
+      "loss": 0.2068,
+      "step": 3889
+    },
+    {
+      "epoch": 2.3051851851851852,
+      "grad_norm": 1.7409112453460693,
+      "learning_rate": 4.635447540011856e-06,
+      "loss": 0.2183,
+      "step": 3890
+    },
+    {
+      "epoch": 2.3057777777777777,
+      "grad_norm": 1.7254568338394165,
+      "learning_rate": 4.631495751827703e-06,
+      "loss": 0.1566,
+      "step": 3891
+    },
+    {
+      "epoch": 2.30637037037037,
+      "grad_norm": 1.9702612161636353,
+      "learning_rate": 4.627543963643549e-06,
+      "loss": 0.1117,
+      "step": 3892
+    },
+    {
+      "epoch": 2.306962962962963,
+      "grad_norm": 1.6468108892440796,
+      "learning_rate": 4.623592175459396e-06,
+      "loss": 0.1383,
+      "step": 3893
+    },
+    {
+      "epoch": 2.3075555555555556,
+      "grad_norm": 2.045693874359131,
+      "learning_rate": 4.619640387275242e-06,
+      "loss": 0.193,
+      "step": 3894
+    },
+    {
+      "epoch": 2.308148148148148,
+      "grad_norm": 1.9262999296188354,
+      "learning_rate": 4.6156885990910895e-06,
+      "loss": 0.1892,
+      "step": 3895
+    },
+    {
+      "epoch": 2.308740740740741,
+      "grad_norm": 1.9302068948745728,
+      "learning_rate": 4.611736810906936e-06,
+      "loss": 0.2164,
+      "step": 3896
+    },
+    {
+      "epoch": 2.3093333333333335,
+      "grad_norm": 2.4222774505615234,
+      "learning_rate": 4.607785022722782e-06,
+      "loss": 0.1244,
+      "step": 3897
+    },
+    {
+      "epoch": 2.309925925925926,
+      "grad_norm": 1.8877326250076294,
+      "learning_rate": 4.603833234538629e-06,
+      "loss": 0.1671,
+      "step": 3898
+    },
+    {
+      "epoch": 2.3105185185185184,
+      "grad_norm": 2.219832181930542,
+      "learning_rate": 4.599881446354475e-06,
+      "loss": 0.2407,
+      "step": 3899
+    },
+    {
+      "epoch": 2.311111111111111,
+      "grad_norm": 2.1754283905029297,
+      "learning_rate": 4.595929658170322e-06,
+      "loss": 0.1597,
+      "step": 3900
+    },
+    {
+      "epoch": 2.311703703703704,
+      "grad_norm": 2.7520384788513184,
+      "learning_rate": 4.591977869986169e-06,
+      "loss": 0.2332,
+      "step": 3901
+    },
+    {
+      "epoch": 2.3122962962962963,
+      "grad_norm": 2.24584698677063,
+      "learning_rate": 4.588026081802016e-06,
+      "loss": 0.2943,
+      "step": 3902
+    },
+    {
+      "epoch": 2.3128888888888888,
+      "grad_norm": 2.5078046321868896,
+      "learning_rate": 4.584074293617862e-06,
+      "loss": 0.2145,
+      "step": 3903
+    },
+    {
+      "epoch": 2.3134814814814817,
+      "grad_norm": 2.2819461822509766,
+      "learning_rate": 4.580122505433709e-06,
+      "loss": 0.1847,
+      "step": 3904
+    },
+    {
+      "epoch": 2.314074074074074,
+      "grad_norm": 2.259018898010254,
+      "learning_rate": 4.576170717249555e-06,
+      "loss": 0.1615,
+      "step": 3905
+    },
+    {
+      "epoch": 2.3146666666666667,
+      "grad_norm": 2.335777997970581,
+      "learning_rate": 4.572218929065402e-06,
+      "loss": 0.2093,
+      "step": 3906
+    },
+    {
+      "epoch": 2.315259259259259,
+      "grad_norm": 1.878119707107544,
+      "learning_rate": 4.568267140881249e-06,
+      "loss": 0.1826,
+      "step": 3907
+    },
+    {
+      "epoch": 2.3158518518518516,
+      "grad_norm": 2.0731334686279297,
+      "learning_rate": 4.564315352697096e-06,
+      "loss": 0.1464,
+      "step": 3908
+    },
+    {
+      "epoch": 2.3164444444444445,
+      "grad_norm": 2.413719654083252,
+      "learning_rate": 4.560363564512943e-06,
+      "loss": 0.1694,
+      "step": 3909
+    },
+    {
+      "epoch": 2.317037037037037,
+      "grad_norm": 1.6520136594772339,
+      "learning_rate": 4.556411776328789e-06,
+      "loss": 0.1035,
+      "step": 3910
+    },
+    {
+      "epoch": 2.3176296296296295,
+      "grad_norm": 3.1925058364868164,
+      "learning_rate": 4.552459988144636e-06,
+      "loss": 0.1666,
+      "step": 3911
+    },
+    {
+      "epoch": 2.3182222222222224,
+      "grad_norm": 2.7200839519500732,
+      "learning_rate": 4.548508199960482e-06,
+      "loss": 0.189,
+      "step": 3912
+    },
+    {
+      "epoch": 2.318814814814815,
+      "grad_norm": 1.601671814918518,
+      "learning_rate": 4.5445564117763294e-06,
+      "loss": 0.1669,
+      "step": 3913
+    },
+    {
+      "epoch": 2.3194074074074074,
+      "grad_norm": 1.481281042098999,
+      "learning_rate": 4.540604623592176e-06,
+      "loss": 0.1083,
+      "step": 3914
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6465957164764404,
+      "learning_rate": 4.536652835408023e-06,
+      "loss": 0.1205,
+      "step": 3915
+    },
+    {
+      "epoch": 2.3205925925925928,
+      "grad_norm": 1.8722267150878906,
+      "learning_rate": 4.532701047223869e-06,
+      "loss": 0.1435,
+      "step": 3916
+    },
+    {
+      "epoch": 2.3211851851851852,
+      "grad_norm": 2.6942923069000244,
+      "learning_rate": 4.528749259039716e-06,
+      "loss": 0.235,
+      "step": 3917
+    },
+    {
+      "epoch": 2.3217777777777777,
+      "grad_norm": 2.348762273788452,
+      "learning_rate": 4.524797470855562e-06,
+      "loss": 0.1607,
+      "step": 3918
+    },
+    {
+      "epoch": 2.32237037037037,
+      "grad_norm": 3.0829899311065674,
+      "learning_rate": 4.5208456826714094e-06,
+      "loss": 0.2297,
+      "step": 3919
+    },
+    {
+      "epoch": 2.322962962962963,
+      "grad_norm": 2.11145281791687,
+      "learning_rate": 4.516893894487256e-06,
+      "loss": 0.1641,
+      "step": 3920
+    },
+    {
+      "epoch": 2.3235555555555556,
+      "grad_norm": 2.4929943084716797,
+      "learning_rate": 4.512942106303103e-06,
+      "loss": 0.2662,
+      "step": 3921
+    },
+    {
+      "epoch": 2.324148148148148,
+      "grad_norm": 1.9295878410339355,
+      "learning_rate": 4.508990318118949e-06,
+      "loss": 0.2226,
+      "step": 3922
+    },
+    {
+      "epoch": 2.3247407407407406,
+      "grad_norm": 1.7740594148635864,
+      "learning_rate": 4.505038529934796e-06,
+      "loss": 0.2192,
+      "step": 3923
+    },
+    {
+      "epoch": 2.3253333333333335,
+      "grad_norm": 2.6018447875976562,
+      "learning_rate": 4.501086741750642e-06,
+      "loss": 0.2728,
+      "step": 3924
+    },
+    {
+      "epoch": 2.325925925925926,
+      "grad_norm": 2.525810956954956,
+      "learning_rate": 4.4971349535664894e-06,
+      "loss": 0.2147,
+      "step": 3925
+    },
+    {
+      "epoch": 2.3265185185185184,
+      "grad_norm": 2.181826114654541,
+      "learning_rate": 4.493183165382336e-06,
+      "loss": 0.1631,
+      "step": 3926
+    },
+    {
+      "epoch": 2.327111111111111,
+      "grad_norm": 2.5912587642669678,
+      "learning_rate": 4.489231377198183e-06,
+      "loss": 0.3103,
+      "step": 3927
+    },
+    {
+      "epoch": 2.327703703703704,
+      "grad_norm": 2.147141695022583,
+      "learning_rate": 4.485279589014029e-06,
+      "loss": 0.2511,
+      "step": 3928
+    },
+    {
+      "epoch": 2.3282962962962963,
+      "grad_norm": 4.41713809967041,
+      "learning_rate": 4.481327800829876e-06,
+      "loss": 0.2244,
+      "step": 3929
+    },
+    {
+      "epoch": 2.328888888888889,
+      "grad_norm": 1.792189598083496,
+      "learning_rate": 4.477376012645722e-06,
+      "loss": 0.216,
+      "step": 3930
+    },
+    {
+      "epoch": 2.3294814814814817,
+      "grad_norm": 1.2979037761688232,
+      "learning_rate": 4.4734242244615694e-06,
+      "loss": 0.1312,
+      "step": 3931
+    },
+    {
+      "epoch": 2.330074074074074,
+      "grad_norm": 2.117457628250122,
+      "learning_rate": 4.4694724362774165e-06,
+      "loss": 0.2274,
+      "step": 3932
+    },
+    {
+      "epoch": 2.3306666666666667,
+      "grad_norm": 1.7824360132217407,
+      "learning_rate": 4.465520648093263e-06,
+      "loss": 0.0981,
+      "step": 3933
+    },
+    {
+      "epoch": 2.331259259259259,
+      "grad_norm": 2.077842950820923,
+      "learning_rate": 4.46156885990911e-06,
+      "loss": 0.2074,
+      "step": 3934
+    },
+    {
+      "epoch": 2.3318518518518516,
+      "grad_norm": 1.7281748056411743,
+      "learning_rate": 4.457617071724956e-06,
+      "loss": 0.1971,
+      "step": 3935
+    },
+    {
+      "epoch": 2.3324444444444445,
+      "grad_norm": 1.8666435480117798,
+      "learning_rate": 4.453665283540803e-06,
+      "loss": 0.187,
+      "step": 3936
+    },
+    {
+      "epoch": 2.333037037037037,
+      "grad_norm": 2.2802233695983887,
+      "learning_rate": 4.449713495356649e-06,
+      "loss": 0.1738,
+      "step": 3937
+    },
+    {
+      "epoch": 2.3336296296296295,
+      "grad_norm": 2.2502129077911377,
+      "learning_rate": 4.445761707172496e-06,
+      "loss": 0.143,
+      "step": 3938
+    },
+    {
+      "epoch": 2.3342222222222224,
+      "grad_norm": 1.457384467124939,
+      "learning_rate": 4.441809918988343e-06,
+      "loss": 0.1844,
+      "step": 3939
+    },
+    {
+      "epoch": 2.334814814814815,
+      "grad_norm": 1.887834072113037,
+      "learning_rate": 4.437858130804189e-06,
+      "loss": 0.1681,
+      "step": 3940
+    },
+    {
+      "epoch": 2.3354074074074074,
+      "grad_norm": 3.1487913131713867,
+      "learning_rate": 4.433906342620036e-06,
+      "loss": 0.1804,
+      "step": 3941
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 2.845705509185791,
+      "learning_rate": 4.429954554435882e-06,
+      "loss": 0.2153,
+      "step": 3942
+    },
+    {
+      "epoch": 2.3365925925925928,
+      "grad_norm": 1.7873482704162598,
+      "learning_rate": 4.4260027662517294e-06,
+      "loss": 0.1023,
+      "step": 3943
+    },
+    {
+      "epoch": 2.3371851851851853,
+      "grad_norm": 3.0304105281829834,
+      "learning_rate": 4.422050978067576e-06,
+      "loss": 0.1823,
+      "step": 3944
+    },
+    {
+      "epoch": 2.3377777777777777,
+      "grad_norm": 2.3834123611450195,
+      "learning_rate": 4.418099189883423e-06,
+      "loss": 0.1719,
+      "step": 3945
+    },
+    {
+      "epoch": 2.33837037037037,
+      "grad_norm": 1.9045099020004272,
+      "learning_rate": 4.414147401699269e-06,
+      "loss": 0.1328,
+      "step": 3946
+    },
+    {
+      "epoch": 2.338962962962963,
+      "grad_norm": 2.792884349822998,
+      "learning_rate": 4.410195613515116e-06,
+      "loss": 0.2108,
+      "step": 3947
+    },
+    {
+      "epoch": 2.3395555555555556,
+      "grad_norm": 2.871816873550415,
+      "learning_rate": 4.406243825330962e-06,
+      "loss": 0.1589,
+      "step": 3948
+    },
+    {
+      "epoch": 2.340148148148148,
+      "grad_norm": 2.1538708209991455,
+      "learning_rate": 4.4022920371468094e-06,
+      "loss": 0.1744,
+      "step": 3949
+    },
+    {
+      "epoch": 2.3407407407407406,
+      "grad_norm": 2.284189462661743,
+      "learning_rate": 4.398340248962656e-06,
+      "loss": 0.2359,
+      "step": 3950
+    },
+    {
+      "epoch": 2.3413333333333335,
+      "grad_norm": 2.5409209728240967,
+      "learning_rate": 4.394388460778503e-06,
+      "loss": 0.1889,
+      "step": 3951
+    },
+    {
+      "epoch": 2.341925925925926,
+      "grad_norm": 2.272437334060669,
+      "learning_rate": 4.390436672594349e-06,
+      "loss": 0.1801,
+      "step": 3952
+    },
+    {
+      "epoch": 2.3425185185185184,
+      "grad_norm": 2.631798267364502,
+      "learning_rate": 4.386484884410196e-06,
+      "loss": 0.256,
+      "step": 3953
+    },
+    {
+      "epoch": 2.343111111111111,
+      "grad_norm": 3.5744264125823975,
+      "learning_rate": 4.382533096226042e-06,
+      "loss": 0.2278,
+      "step": 3954
+    },
+    {
+      "epoch": 2.343703703703704,
+      "grad_norm": 3.458057403564453,
+      "learning_rate": 4.3785813080418894e-06,
+      "loss": 0.3187,
+      "step": 3955
+    },
+    {
+      "epoch": 2.3442962962962963,
+      "grad_norm": 2.435739278793335,
+      "learning_rate": 4.374629519857736e-06,
+      "loss": 0.2337,
+      "step": 3956
+    },
+    {
+      "epoch": 2.344888888888889,
+      "grad_norm": 2.8665759563446045,
+      "learning_rate": 4.370677731673583e-06,
+      "loss": 0.1719,
+      "step": 3957
+    },
+    {
+      "epoch": 2.3454814814814817,
+      "grad_norm": 2.090512275695801,
+      "learning_rate": 4.366725943489429e-06,
+      "loss": 0.16,
+      "step": 3958
+    },
+    {
+      "epoch": 2.346074074074074,
+      "grad_norm": 1.6313005685806274,
+      "learning_rate": 4.362774155305276e-06,
+      "loss": 0.1661,
+      "step": 3959
+    },
+    {
+      "epoch": 2.3466666666666667,
+      "grad_norm": 3.0109150409698486,
+      "learning_rate": 4.358822367121122e-06,
+      "loss": 0.2109,
+      "step": 3960
+    },
+    {
+      "epoch": 2.347259259259259,
+      "grad_norm": 2.2664315700531006,
+      "learning_rate": 4.3548705789369694e-06,
+      "loss": 0.2043,
+      "step": 3961
+    },
+    {
+      "epoch": 2.3478518518518516,
+      "grad_norm": 1.7514921426773071,
+      "learning_rate": 4.350918790752816e-06,
+      "loss": 0.1356,
+      "step": 3962
+    },
+    {
+      "epoch": 2.3484444444444446,
+      "grad_norm": 2.3291375637054443,
+      "learning_rate": 4.346967002568663e-06,
+      "loss": 0.1722,
+      "step": 3963
+    },
+    {
+      "epoch": 2.349037037037037,
+      "grad_norm": 2.577986717224121,
+      "learning_rate": 4.34301521438451e-06,
+      "loss": 0.2185,
+      "step": 3964
+    },
+    {
+      "epoch": 2.3496296296296295,
+      "grad_norm": 2.560969114303589,
+      "learning_rate": 4.339063426200356e-06,
+      "loss": 0.1409,
+      "step": 3965
+    },
+    {
+      "epoch": 2.3502222222222224,
+      "grad_norm": 2.5752460956573486,
+      "learning_rate": 4.335111638016203e-06,
+      "loss": 0.1777,
+      "step": 3966
+    },
+    {
+      "epoch": 2.350814814814815,
+      "grad_norm": 1.6048526763916016,
+      "learning_rate": 4.3311598498320494e-06,
+      "loss": 0.1382,
+      "step": 3967
+    },
+    {
+      "epoch": 2.3514074074074074,
+      "grad_norm": 1.9065654277801514,
+      "learning_rate": 4.3272080616478965e-06,
+      "loss": 0.168,
+      "step": 3968
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 1.6464698314666748,
+      "learning_rate": 4.323256273463743e-06,
+      "loss": 0.195,
+      "step": 3969
+    },
+    {
+      "epoch": 2.3525925925925923,
+      "grad_norm": 1.93190336227417,
+      "learning_rate": 4.31930448527959e-06,
+      "loss": 0.2302,
+      "step": 3970
+    },
+    {
+      "epoch": 2.3531851851851853,
+      "grad_norm": 2.1430890560150146,
+      "learning_rate": 4.315352697095436e-06,
+      "loss": 0.14,
+      "step": 3971
+    },
+    {
+      "epoch": 2.3537777777777777,
+      "grad_norm": 2.0593245029449463,
+      "learning_rate": 4.311400908911283e-06,
+      "loss": 0.1881,
+      "step": 3972
+    },
+    {
+      "epoch": 2.3543703703703702,
+      "grad_norm": 2.7232377529144287,
+      "learning_rate": 4.3074491207271294e-06,
+      "loss": 0.1626,
+      "step": 3973
+    },
+    {
+      "epoch": 2.354962962962963,
+      "grad_norm": 2.5886497497558594,
+      "learning_rate": 4.3034973325429765e-06,
+      "loss": 0.1573,
+      "step": 3974
+    },
+    {
+      "epoch": 2.3555555555555556,
+      "grad_norm": 1.6203527450561523,
+      "learning_rate": 4.299545544358823e-06,
+      "loss": 0.1919,
+      "step": 3975
+    },
+    {
+      "epoch": 2.356148148148148,
+      "grad_norm": 1.9429324865341187,
+      "learning_rate": 4.295593756174669e-06,
+      "loss": 0.1628,
+      "step": 3976
+    },
+    {
+      "epoch": 2.3567407407407406,
+      "grad_norm": 2.144810676574707,
+      "learning_rate": 4.291641967990516e-06,
+      "loss": 0.1851,
+      "step": 3977
+    },
+    {
+      "epoch": 2.3573333333333335,
+      "grad_norm": 3.05389404296875,
+      "learning_rate": 4.287690179806362e-06,
+      "loss": 0.2397,
+      "step": 3978
+    },
+    {
+      "epoch": 2.357925925925926,
+      "grad_norm": 1.9549895524978638,
+      "learning_rate": 4.2837383916222094e-06,
+      "loss": 0.179,
+      "step": 3979
+    },
+    {
+      "epoch": 2.3585185185185185,
+      "grad_norm": 2.7972350120544434,
+      "learning_rate": 4.279786603438056e-06,
+      "loss": 0.229,
+      "step": 3980
+    },
+    {
+      "epoch": 2.359111111111111,
+      "grad_norm": 2.386857509613037,
+      "learning_rate": 4.275834815253903e-06,
+      "loss": 0.2625,
+      "step": 3981
+    },
+    {
+      "epoch": 2.359703703703704,
+      "grad_norm": 1.7754262685775757,
+      "learning_rate": 4.271883027069749e-06,
+      "loss": 0.0968,
+      "step": 3982
+    },
+    {
+      "epoch": 2.3602962962962963,
+      "grad_norm": 3.946784496307373,
+      "learning_rate": 4.267931238885596e-06,
+      "loss": 0.2396,
+      "step": 3983
+    },
+    {
+      "epoch": 2.360888888888889,
+      "grad_norm": 2.949882984161377,
+      "learning_rate": 4.263979450701442e-06,
+      "loss": 0.1646,
+      "step": 3984
+    },
+    {
+      "epoch": 2.3614814814814813,
+      "grad_norm": 4.206084251403809,
+      "learning_rate": 4.2600276625172894e-06,
+      "loss": 0.3493,
+      "step": 3985
+    },
+    {
+      "epoch": 2.362074074074074,
+      "grad_norm": 1.9557431936264038,
+      "learning_rate": 4.256075874333136e-06,
+      "loss": 0.1888,
+      "step": 3986
+    },
+    {
+      "epoch": 2.3626666666666667,
+      "grad_norm": 2.3978123664855957,
+      "learning_rate": 4.252124086148983e-06,
+      "loss": 0.2217,
+      "step": 3987
+    },
+    {
+      "epoch": 2.363259259259259,
+      "grad_norm": 1.998841404914856,
+      "learning_rate": 4.248172297964829e-06,
+      "loss": 0.1588,
+      "step": 3988
+    },
+    {
+      "epoch": 2.3638518518518516,
+      "grad_norm": 1.903540849685669,
+      "learning_rate": 4.244220509780676e-06,
+      "loss": 0.1319,
+      "step": 3989
+    },
+    {
+      "epoch": 2.3644444444444446,
+      "grad_norm": 2.128915548324585,
+      "learning_rate": 4.240268721596522e-06,
+      "loss": 0.132,
+      "step": 3990
+    },
+    {
+      "epoch": 2.365037037037037,
+      "grad_norm": 2.6029810905456543,
+      "learning_rate": 4.2363169334123694e-06,
+      "loss": 0.1755,
+      "step": 3991
+    },
+    {
+      "epoch": 2.3656296296296295,
+      "grad_norm": 3.6023008823394775,
+      "learning_rate": 4.232365145228216e-06,
+      "loss": 0.2417,
+      "step": 3992
+    },
+    {
+      "epoch": 2.3662222222222224,
+      "grad_norm": 4.066946506500244,
+      "learning_rate": 4.228413357044063e-06,
+      "loss": 0.1872,
+      "step": 3993
+    },
+    {
+      "epoch": 2.366814814814815,
+      "grad_norm": 2.500814437866211,
+      "learning_rate": 4.22446156885991e-06,
+      "loss": 0.2129,
+      "step": 3994
+    },
+    {
+      "epoch": 2.3674074074074074,
+      "grad_norm": 2.5899198055267334,
+      "learning_rate": 4.220509780675756e-06,
+      "loss": 0.2842,
+      "step": 3995
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 1.7421834468841553,
+      "learning_rate": 4.216557992491603e-06,
+      "loss": 0.1928,
+      "step": 3996
+    },
+    {
+      "epoch": 2.3685925925925924,
+      "grad_norm": 2.0843505859375,
+      "learning_rate": 4.212606204307449e-06,
+      "loss": 0.2023,
+      "step": 3997
+    },
+    {
+      "epoch": 2.3691851851851853,
+      "grad_norm": 2.3495380878448486,
+      "learning_rate": 4.2086544161232965e-06,
+      "loss": 0.1716,
+      "step": 3998
+    },
+    {
+      "epoch": 2.3697777777777778,
+      "grad_norm": 2.6450843811035156,
+      "learning_rate": 4.204702627939143e-06,
+      "loss": 0.3041,
+      "step": 3999
+    },
+    {
+      "epoch": 2.3703703703703702,
+      "grad_norm": 3.5630300045013428,
+      "learning_rate": 4.20075083975499e-06,
+      "loss": 0.1976,
+      "step": 4000
+    },
+    {
+      "epoch": 2.3715555555555556,
+      "grad_norm": 2.3204588890075684,
+      "learning_rate": 4.196799051570836e-06,
+      "loss": 0.17,
+      "step": 4001
+    },
+    {
+      "epoch": 2.372148148148148,
+      "grad_norm": 2.72119402885437,
+      "learning_rate": 4.192847263386683e-06,
+      "loss": 0.1841,
+      "step": 4002
+    },
+    {
+      "epoch": 2.3727407407407406,
+      "grad_norm": 2.1960995197296143,
+      "learning_rate": 4.188895475202529e-06,
+      "loss": 0.1261,
+      "step": 4003
+    },
+    {
+      "epoch": 2.3733333333333335,
+      "grad_norm": 2.3310985565185547,
+      "learning_rate": 4.1849436870183765e-06,
+      "loss": 0.1374,
+      "step": 4004
+    },
+    {
+      "epoch": 2.373925925925926,
+      "grad_norm": 1.3955658674240112,
+      "learning_rate": 4.180991898834223e-06,
+      "loss": 0.1317,
+      "step": 4005
+    },
+    {
+      "epoch": 2.3745185185185185,
+      "grad_norm": 1.7689566612243652,
+      "learning_rate": 4.17704011065007e-06,
+      "loss": 0.2102,
+      "step": 4006
+    },
+    {
+      "epoch": 2.375111111111111,
+      "grad_norm": 1.6150625944137573,
+      "learning_rate": 4.173088322465916e-06,
+      "loss": 0.1192,
+      "step": 4007
+    },
+    {
+      "epoch": 2.375703703703704,
+      "grad_norm": 1.6348856687545776,
+      "learning_rate": 4.169136534281763e-06,
+      "loss": 0.1436,
+      "step": 4008
+    },
+    {
+      "epoch": 2.3762962962962964,
+      "grad_norm": 2.141201972961426,
+      "learning_rate": 4.165184746097609e-06,
+      "loss": 0.1678,
+      "step": 4009
+    },
+    {
+      "epoch": 2.376888888888889,
+      "grad_norm": 2.8617677688598633,
+      "learning_rate": 4.1612329579134565e-06,
+      "loss": 0.2445,
+      "step": 4010
+    },
+    {
+      "epoch": 2.3774814814814813,
+      "grad_norm": 3.914137363433838,
+      "learning_rate": 4.157281169729303e-06,
+      "loss": 0.2055,
+      "step": 4011
+    },
+    {
+      "epoch": 2.3780740740740742,
+      "grad_norm": 2.1020336151123047,
+      "learning_rate": 4.15332938154515e-06,
+      "loss": 0.1761,
+      "step": 4012
+    },
+    {
+      "epoch": 2.3786666666666667,
+      "grad_norm": 4.3048319816589355,
+      "learning_rate": 4.149377593360996e-06,
+      "loss": 0.159,
+      "step": 4013
+    },
+    {
+      "epoch": 2.379259259259259,
+      "grad_norm": 2.128159523010254,
+      "learning_rate": 4.145425805176843e-06,
+      "loss": 0.1866,
+      "step": 4014
+    },
+    {
+      "epoch": 2.3798518518518517,
+      "grad_norm": 1.8931909799575806,
+      "learning_rate": 4.141474016992689e-06,
+      "loss": 0.1574,
+      "step": 4015
+    },
+    {
+      "epoch": 2.3804444444444446,
+      "grad_norm": 2.162123918533325,
+      "learning_rate": 4.137522228808536e-06,
+      "loss": 0.1376,
+      "step": 4016
+    },
+    {
+      "epoch": 2.381037037037037,
+      "grad_norm": 2.306570291519165,
+      "learning_rate": 4.133570440624383e-06,
+      "loss": 0.232,
+      "step": 4017
+    },
+    {
+      "epoch": 2.3816296296296295,
+      "grad_norm": 2.103764057159424,
+      "learning_rate": 4.129618652440229e-06,
+      "loss": 0.1596,
+      "step": 4018
+    },
+    {
+      "epoch": 2.3822222222222225,
+      "grad_norm": 2.2318673133850098,
+      "learning_rate": 4.125666864256076e-06,
+      "loss": 0.2522,
+      "step": 4019
+    },
+    {
+      "epoch": 2.382814814814815,
+      "grad_norm": 2.235348701477051,
+      "learning_rate": 4.121715076071922e-06,
+      "loss": 0.1615,
+      "step": 4020
+    },
+    {
+      "epoch": 2.3834074074074074,
+      "grad_norm": 2.506279468536377,
+      "learning_rate": 4.117763287887769e-06,
+      "loss": 0.2138,
+      "step": 4021
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 1.6865050792694092,
+      "learning_rate": 4.113811499703616e-06,
+      "loss": 0.1235,
+      "step": 4022
+    },
+    {
+      "epoch": 2.3845925925925924,
+      "grad_norm": 1.7588621377944946,
+      "learning_rate": 4.109859711519463e-06,
+      "loss": 0.1737,
+      "step": 4023
+    },
+    {
+      "epoch": 2.3851851851851853,
+      "grad_norm": 2.071753740310669,
+      "learning_rate": 4.105907923335309e-06,
+      "loss": 0.2554,
+      "step": 4024
+    },
+    {
+      "epoch": 2.3857777777777778,
+      "grad_norm": 2.3983497619628906,
+      "learning_rate": 4.101956135151156e-06,
+      "loss": 0.3196,
+      "step": 4025
+    },
+    {
+      "epoch": 2.3863703703703703,
+      "grad_norm": 3.5294482707977295,
+      "learning_rate": 4.098004346967003e-06,
+      "loss": 0.2862,
+      "step": 4026
+    },
+    {
+      "epoch": 2.386962962962963,
+      "grad_norm": 2.5162949562072754,
+      "learning_rate": 4.094052558782849e-06,
+      "loss": 0.1586,
+      "step": 4027
+    },
+    {
+      "epoch": 2.3875555555555557,
+      "grad_norm": 1.8969959020614624,
+      "learning_rate": 4.0901007705986965e-06,
+      "loss": 0.1547,
+      "step": 4028
+    },
+    {
+      "epoch": 2.388148148148148,
+      "grad_norm": 1.9014815092086792,
+      "learning_rate": 4.086148982414543e-06,
+      "loss": 0.1711,
+      "step": 4029
+    },
+    {
+      "epoch": 2.3887407407407406,
+      "grad_norm": 2.161135673522949,
+      "learning_rate": 4.08219719423039e-06,
+      "loss": 0.2483,
+      "step": 4030
+    },
+    {
+      "epoch": 2.389333333333333,
+      "grad_norm": 1.9820996522903442,
+      "learning_rate": 4.078245406046236e-06,
+      "loss": 0.2348,
+      "step": 4031
+    },
+    {
+      "epoch": 2.389925925925926,
+      "grad_norm": 2.0012574195861816,
+      "learning_rate": 4.074293617862083e-06,
+      "loss": 0.157,
+      "step": 4032
+    },
+    {
+      "epoch": 2.3905185185185185,
+      "grad_norm": 2.4220783710479736,
+      "learning_rate": 4.070341829677929e-06,
+      "loss": 0.2142,
+      "step": 4033
+    },
+    {
+      "epoch": 2.391111111111111,
+      "grad_norm": 2.143470525741577,
+      "learning_rate": 4.0663900414937765e-06,
+      "loss": 0.213,
+      "step": 4034
+    },
+    {
+      "epoch": 2.391703703703704,
+      "grad_norm": 1.8890646696090698,
+      "learning_rate": 4.062438253309623e-06,
+      "loss": 0.1964,
+      "step": 4035
+    },
+    {
+      "epoch": 2.3922962962962964,
+      "grad_norm": 2.660048246383667,
+      "learning_rate": 4.05848646512547e-06,
+      "loss": 0.1788,
+      "step": 4036
+    },
+    {
+      "epoch": 2.392888888888889,
+      "grad_norm": 1.933821201324463,
+      "learning_rate": 4.054534676941316e-06,
+      "loss": 0.138,
+      "step": 4037
+    },
+    {
+      "epoch": 2.3934814814814813,
+      "grad_norm": 1.7812376022338867,
+      "learning_rate": 4.050582888757163e-06,
+      "loss": 0.144,
+      "step": 4038
+    },
+    {
+      "epoch": 2.3940740740740742,
+      "grad_norm": 3.4859302043914795,
+      "learning_rate": 4.046631100573009e-06,
+      "loss": 0.2827,
+      "step": 4039
+    },
+    {
+      "epoch": 2.3946666666666667,
+      "grad_norm": 2.6835832595825195,
+      "learning_rate": 4.0426793123888565e-06,
+      "loss": 0.2209,
+      "step": 4040
+    },
+    {
+      "epoch": 2.395259259259259,
+      "grad_norm": 2.0942728519439697,
+      "learning_rate": 4.038727524204703e-06,
+      "loss": 0.142,
+      "step": 4041
+    },
+    {
+      "epoch": 2.3958518518518517,
+      "grad_norm": 4.407095909118652,
+      "learning_rate": 4.03477573602055e-06,
+      "loss": 0.2568,
+      "step": 4042
+    },
+    {
+      "epoch": 2.3964444444444446,
+      "grad_norm": 2.5345211029052734,
+      "learning_rate": 4.030823947836396e-06,
+      "loss": 0.1702,
+      "step": 4043
+    },
+    {
+      "epoch": 2.397037037037037,
+      "grad_norm": 2.533353805541992,
+      "learning_rate": 4.026872159652243e-06,
+      "loss": 0.2036,
+      "step": 4044
+    },
+    {
+      "epoch": 2.3976296296296296,
+      "grad_norm": 1.9407941102981567,
+      "learning_rate": 4.022920371468089e-06,
+      "loss": 0.2315,
+      "step": 4045
+    },
+    {
+      "epoch": 2.398222222222222,
+      "grad_norm": 3.6190218925476074,
+      "learning_rate": 4.0189685832839365e-06,
+      "loss": 0.1697,
+      "step": 4046
+    },
+    {
+      "epoch": 2.398814814814815,
+      "grad_norm": 2.133800745010376,
+      "learning_rate": 4.015016795099783e-06,
+      "loss": 0.1783,
+      "step": 4047
+    },
+    {
+      "epoch": 2.3994074074074074,
+      "grad_norm": 2.2818777561187744,
+      "learning_rate": 4.01106500691563e-06,
+      "loss": 0.132,
+      "step": 4048
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.9051567316055298,
+      "learning_rate": 4.007113218731477e-06,
+      "loss": 0.1551,
+      "step": 4049
+    },
+    {
+      "epoch": 2.4005925925925924,
+      "grad_norm": 1.9545505046844482,
+      "learning_rate": 4.003161430547323e-06,
+      "loss": 0.1694,
+      "step": 4050
+    },
+    {
+      "epoch": 2.4011851851851853,
+      "grad_norm": 2.3476476669311523,
+      "learning_rate": 3.99920964236317e-06,
+      "loss": 0.1716,
+      "step": 4051
+    },
+    {
+      "epoch": 2.401777777777778,
+      "grad_norm": 2.7131636142730713,
+      "learning_rate": 3.9952578541790165e-06,
+      "loss": 0.2331,
+      "step": 4052
+    },
+    {
+      "epoch": 2.4023703703703703,
+      "grad_norm": 2.3288843631744385,
+      "learning_rate": 3.991306065994864e-06,
+      "loss": 0.2145,
+      "step": 4053
+    },
+    {
+      "epoch": 2.402962962962963,
+      "grad_norm": 2.141965866088867,
+      "learning_rate": 3.98735427781071e-06,
+      "loss": 0.1866,
+      "step": 4054
+    },
+    {
+      "epoch": 2.4035555555555557,
+      "grad_norm": 1.9893879890441895,
+      "learning_rate": 3.983402489626556e-06,
+      "loss": 0.1397,
+      "step": 4055
+    },
+    {
+      "epoch": 2.404148148148148,
+      "grad_norm": 2.18739652633667,
+      "learning_rate": 3.979450701442403e-06,
+      "loss": 0.155,
+      "step": 4056
+    },
+    {
+      "epoch": 2.4047407407407406,
+      "grad_norm": 1.9836313724517822,
+      "learning_rate": 3.975498913258249e-06,
+      "loss": 0.165,
+      "step": 4057
+    },
+    {
+      "epoch": 2.405333333333333,
+      "grad_norm": 1.6728217601776123,
+      "learning_rate": 3.9715471250740965e-06,
+      "loss": 0.1863,
+      "step": 4058
+    },
+    {
+      "epoch": 2.405925925925926,
+      "grad_norm": 1.924508810043335,
+      "learning_rate": 3.967595336889943e-06,
+      "loss": 0.175,
+      "step": 4059
+    },
+    {
+      "epoch": 2.4065185185185185,
+      "grad_norm": 2.3270931243896484,
+      "learning_rate": 3.96364354870579e-06,
+      "loss": 0.2072,
+      "step": 4060
+    },
+    {
+      "epoch": 2.407111111111111,
+      "grad_norm": 2.081918954849243,
+      "learning_rate": 3.959691760521636e-06,
+      "loss": 0.1644,
+      "step": 4061
+    },
+    {
+      "epoch": 2.407703703703704,
+      "grad_norm": 1.7902871370315552,
+      "learning_rate": 3.955739972337483e-06,
+      "loss": 0.1964,
+      "step": 4062
+    },
+    {
+      "epoch": 2.4082962962962964,
+      "grad_norm": 1.971307635307312,
+      "learning_rate": 3.951788184153329e-06,
+      "loss": 0.1649,
+      "step": 4063
+    },
+    {
+      "epoch": 2.408888888888889,
+      "grad_norm": 1.764816403388977,
+      "learning_rate": 3.9478363959691765e-06,
+      "loss": 0.172,
+      "step": 4064
+    },
+    {
+      "epoch": 2.4094814814814813,
+      "grad_norm": 2.5210297107696533,
+      "learning_rate": 3.943884607785023e-06,
+      "loss": 0.309,
+      "step": 4065
+    },
+    {
+      "epoch": 2.4100740740740743,
+      "grad_norm": 2.397313117980957,
+      "learning_rate": 3.93993281960087e-06,
+      "loss": 0.222,
+      "step": 4066
+    },
+    {
+      "epoch": 2.4106666666666667,
+      "grad_norm": 2.413499116897583,
+      "learning_rate": 3.935981031416716e-06,
+      "loss": 0.2621,
+      "step": 4067
+    },
+    {
+      "epoch": 2.411259259259259,
+      "grad_norm": 3.691664695739746,
+      "learning_rate": 3.932029243232563e-06,
+      "loss": 0.2365,
+      "step": 4068
+    },
+    {
+      "epoch": 2.4118518518518517,
+      "grad_norm": 2.646416187286377,
+      "learning_rate": 3.928077455048409e-06,
+      "loss": 0.1442,
+      "step": 4069
+    },
+    {
+      "epoch": 2.4124444444444446,
+      "grad_norm": 2.96108078956604,
+      "learning_rate": 3.9241256668642565e-06,
+      "loss": 0.2382,
+      "step": 4070
+    },
+    {
+      "epoch": 2.413037037037037,
+      "grad_norm": 2.0511248111724854,
+      "learning_rate": 3.920173878680103e-06,
+      "loss": 0.1467,
+      "step": 4071
+    },
+    {
+      "epoch": 2.4136296296296296,
+      "grad_norm": 2.124934673309326,
+      "learning_rate": 3.91622209049595e-06,
+      "loss": 0.1911,
+      "step": 4072
+    },
+    {
+      "epoch": 2.414222222222222,
+      "grad_norm": 2.347308397293091,
+      "learning_rate": 3.912270302311796e-06,
+      "loss": 0.1841,
+      "step": 4073
+    },
+    {
+      "epoch": 2.414814814814815,
+      "grad_norm": 1.9483251571655273,
+      "learning_rate": 3.908318514127643e-06,
+      "loss": 0.1948,
+      "step": 4074
+    },
+    {
+      "epoch": 2.4154074074074074,
+      "grad_norm": 2.0900020599365234,
+      "learning_rate": 3.904366725943489e-06,
+      "loss": 0.1298,
+      "step": 4075
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 3.2128078937530518,
+      "learning_rate": 3.9004149377593365e-06,
+      "loss": 0.3254,
+      "step": 4076
+    },
+    {
+      "epoch": 2.4165925925925924,
+      "grad_norm": 2.410444498062134,
+      "learning_rate": 3.896463149575183e-06,
+      "loss": 0.2844,
+      "step": 4077
+    },
+    {
+      "epoch": 2.4171851851851853,
+      "grad_norm": 3.427100896835327,
+      "learning_rate": 3.89251136139103e-06,
+      "loss": 0.1739,
+      "step": 4078
+    },
+    {
+      "epoch": 2.417777777777778,
+      "grad_norm": 1.8441640138626099,
+      "learning_rate": 3.888559573206877e-06,
+      "loss": 0.1155,
+      "step": 4079
+    },
+    {
+      "epoch": 2.4183703703703703,
+      "grad_norm": 2.0790140628814697,
+      "learning_rate": 3.884607785022723e-06,
+      "loss": 0.1455,
+      "step": 4080
+    },
+    {
+      "epoch": 2.418962962962963,
+      "grad_norm": 2.6726059913635254,
+      "learning_rate": 3.88065599683857e-06,
+      "loss": 0.1638,
+      "step": 4081
+    },
+    {
+      "epoch": 2.4195555555555557,
+      "grad_norm": 1.5551730394363403,
+      "learning_rate": 3.8767042086544165e-06,
+      "loss": 0.1751,
+      "step": 4082
+    },
+    {
+      "epoch": 2.420148148148148,
+      "grad_norm": 2.690722942352295,
+      "learning_rate": 3.872752420470264e-06,
+      "loss": 0.271,
+      "step": 4083
+    },
+    {
+      "epoch": 2.4207407407407406,
+      "grad_norm": 2.431910753250122,
+      "learning_rate": 3.86880063228611e-06,
+      "loss": 0.1621,
+      "step": 4084
+    },
+    {
+      "epoch": 2.421333333333333,
+      "grad_norm": 1.6369627714157104,
+      "learning_rate": 3.864848844101957e-06,
+      "loss": 0.1507,
+      "step": 4085
+    },
+    {
+      "epoch": 2.421925925925926,
+      "grad_norm": 1.5960049629211426,
+      "learning_rate": 3.860897055917803e-06,
+      "loss": 0.1194,
+      "step": 4086
+    },
+    {
+      "epoch": 2.4225185185185185,
+      "grad_norm": 1.5332497358322144,
+      "learning_rate": 3.85694526773365e-06,
+      "loss": 0.1462,
+      "step": 4087
+    },
+    {
+      "epoch": 2.423111111111111,
+      "grad_norm": 2.3000829219818115,
+      "learning_rate": 3.8529934795494965e-06,
+      "loss": 0.2462,
+      "step": 4088
+    },
+    {
+      "epoch": 2.423703703703704,
+      "grad_norm": 1.8755767345428467,
+      "learning_rate": 3.8490416913653436e-06,
+      "loss": 0.1488,
+      "step": 4089
+    },
+    {
+      "epoch": 2.4242962962962964,
+      "grad_norm": 3.2239487171173096,
+      "learning_rate": 3.84508990318119e-06,
+      "loss": 0.1885,
+      "step": 4090
+    },
+    {
+      "epoch": 2.424888888888889,
+      "grad_norm": 2.216632604598999,
+      "learning_rate": 3.841138114997037e-06,
+      "loss": 0.1494,
+      "step": 4091
+    },
+    {
+      "epoch": 2.4254814814814813,
+      "grad_norm": 2.056556224822998,
+      "learning_rate": 3.837186326812883e-06,
+      "loss": 0.1757,
+      "step": 4092
+    },
+    {
+      "epoch": 2.426074074074074,
+      "grad_norm": 2.0064709186553955,
+      "learning_rate": 3.83323453862873e-06,
+      "loss": 0.1847,
+      "step": 4093
+    },
+    {
+      "epoch": 2.4266666666666667,
+      "grad_norm": 2.25256085395813,
+      "learning_rate": 3.8292827504445765e-06,
+      "loss": 0.1796,
+      "step": 4094
+    },
+    {
+      "epoch": 2.4272592592592592,
+      "grad_norm": 2.3947856426239014,
+      "learning_rate": 3.825330962260423e-06,
+      "loss": 0.2358,
+      "step": 4095
+    },
+    {
+      "epoch": 2.4278518518518517,
+      "grad_norm": 1.8659683465957642,
+      "learning_rate": 3.82137917407627e-06,
+      "loss": 0.1291,
+      "step": 4096
+    },
+    {
+      "epoch": 2.4284444444444446,
+      "grad_norm": 3.434382677078247,
+      "learning_rate": 3.817427385892116e-06,
+      "loss": 0.1492,
+      "step": 4097
+    },
+    {
+      "epoch": 2.429037037037037,
+      "grad_norm": 2.346041202545166,
+      "learning_rate": 3.813475597707963e-06,
+      "loss": 0.1833,
+      "step": 4098
+    },
+    {
+      "epoch": 2.4296296296296296,
+      "grad_norm": 2.6594362258911133,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.1524,
+      "step": 4099
+    },
+    {
+      "epoch": 2.430222222222222,
+      "grad_norm": 1.6149851083755493,
+      "learning_rate": 3.8055720213396565e-06,
+      "loss": 0.1368,
+      "step": 4100
+    },
+    {
+      "epoch": 2.430814814814815,
+      "grad_norm": 2.1633846759796143,
+      "learning_rate": 3.801620233155503e-06,
+      "loss": 0.2122,
+      "step": 4101
+    },
+    {
+      "epoch": 2.4314074074074075,
+      "grad_norm": 2.4627339839935303,
+      "learning_rate": 3.79766844497135e-06,
+      "loss": 0.1942,
+      "step": 4102
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 2.0872690677642822,
+      "learning_rate": 3.7937166567871965e-06,
+      "loss": 0.2185,
+      "step": 4103
+    },
+    {
+      "epoch": 2.4325925925925924,
+      "grad_norm": 1.840808629989624,
+      "learning_rate": 3.789764868603043e-06,
+      "loss": 0.1502,
+      "step": 4104
+    },
+    {
+      "epoch": 2.4331851851851853,
+      "grad_norm": 2.278040647506714,
+      "learning_rate": 3.78581308041889e-06,
+      "loss": 0.2565,
+      "step": 4105
+    },
+    {
+      "epoch": 2.433777777777778,
+      "grad_norm": 1.9836419820785522,
+      "learning_rate": 3.7818612922347365e-06,
+      "loss": 0.1163,
+      "step": 4106
+    },
+    {
+      "epoch": 2.4343703703703703,
+      "grad_norm": 2.385202169418335,
+      "learning_rate": 3.777909504050583e-06,
+      "loss": 0.178,
+      "step": 4107
+    },
+    {
+      "epoch": 2.4349629629629628,
+      "grad_norm": 2.213113307952881,
+      "learning_rate": 3.77395771586643e-06,
+      "loss": 0.2652,
+      "step": 4108
+    },
+    {
+      "epoch": 2.4355555555555557,
+      "grad_norm": 3.631687641143799,
+      "learning_rate": 3.7700059276822765e-06,
+      "loss": 0.199,
+      "step": 4109
+    },
+    {
+      "epoch": 2.436148148148148,
+      "grad_norm": 2.0553882122039795,
+      "learning_rate": 3.766054139498123e-06,
+      "loss": 0.143,
+      "step": 4110
+    },
+    {
+      "epoch": 2.4367407407407407,
+      "grad_norm": 1.9475470781326294,
+      "learning_rate": 3.76210235131397e-06,
+      "loss": 0.1797,
+      "step": 4111
+    },
+    {
+      "epoch": 2.437333333333333,
+      "grad_norm": 2.260472297668457,
+      "learning_rate": 3.7581505631298165e-06,
+      "loss": 0.1791,
+      "step": 4112
+    },
+    {
+      "epoch": 2.437925925925926,
+      "grad_norm": 2.0690927505493164,
+      "learning_rate": 3.754198774945663e-06,
+      "loss": 0.1406,
+      "step": 4113
+    },
+    {
+      "epoch": 2.4385185185185185,
+      "grad_norm": 2.149660587310791,
+      "learning_rate": 3.75024698676151e-06,
+      "loss": 0.2103,
+      "step": 4114
+    },
+    {
+      "epoch": 2.439111111111111,
+      "grad_norm": 2.2990400791168213,
+      "learning_rate": 3.7462951985773565e-06,
+      "loss": 0.1472,
+      "step": 4115
+    },
+    {
+      "epoch": 2.439703703703704,
+      "grad_norm": 2.87217116355896,
+      "learning_rate": 3.742343410393203e-06,
+      "loss": 0.2014,
+      "step": 4116
+    },
+    {
+      "epoch": 2.4402962962962964,
+      "grad_norm": 2.267664670944214,
+      "learning_rate": 3.73839162220905e-06,
+      "loss": 0.211,
+      "step": 4117
+    },
+    {
+      "epoch": 2.440888888888889,
+      "grad_norm": 2.283905506134033,
+      "learning_rate": 3.7344398340248965e-06,
+      "loss": 0.2471,
+      "step": 4118
+    },
+    {
+      "epoch": 2.4414814814814814,
+      "grad_norm": 1.9193238019943237,
+      "learning_rate": 3.730488045840743e-06,
+      "loss": 0.1404,
+      "step": 4119
+    },
+    {
+      "epoch": 2.442074074074074,
+      "grad_norm": 1.8108962774276733,
+      "learning_rate": 3.72653625765659e-06,
+      "loss": 0.1718,
+      "step": 4120
+    },
+    {
+      "epoch": 2.4426666666666668,
+      "grad_norm": 1.561130404472351,
+      "learning_rate": 3.7225844694724365e-06,
+      "loss": 0.1367,
+      "step": 4121
+    },
+    {
+      "epoch": 2.4432592592592592,
+      "grad_norm": 1.5084766149520874,
+      "learning_rate": 3.7186326812882836e-06,
+      "loss": 0.114,
+      "step": 4122
+    },
+    {
+      "epoch": 2.4438518518518517,
+      "grad_norm": 1.8120487928390503,
+      "learning_rate": 3.7146808931041302e-06,
+      "loss": 0.137,
+      "step": 4123
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 2.2811598777770996,
+      "learning_rate": 3.710729104919977e-06,
+      "loss": 0.159,
+      "step": 4124
+    },
+    {
+      "epoch": 2.445037037037037,
+      "grad_norm": 2.7834620475769043,
+      "learning_rate": 3.7067773167358236e-06,
+      "loss": 0.234,
+      "step": 4125
+    },
+    {
+      "epoch": 2.4456296296296296,
+      "grad_norm": 2.4866907596588135,
+      "learning_rate": 3.7028255285516702e-06,
+      "loss": 0.2267,
+      "step": 4126
+    },
+    {
+      "epoch": 2.446222222222222,
+      "grad_norm": 1.9311749935150146,
+      "learning_rate": 3.698873740367517e-06,
+      "loss": 0.1459,
+      "step": 4127
+    },
+    {
+      "epoch": 2.446814814814815,
+      "grad_norm": 3.298818588256836,
+      "learning_rate": 3.6949219521833636e-06,
+      "loss": 0.3292,
+      "step": 4128
+    },
+    {
+      "epoch": 2.4474074074074075,
+      "grad_norm": 1.9539239406585693,
+      "learning_rate": 3.6909701639992102e-06,
+      "loss": 0.145,
+      "step": 4129
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 2.220761775970459,
+      "learning_rate": 3.687018375815057e-06,
+      "loss": 0.2111,
+      "step": 4130
+    },
+    {
+      "epoch": 2.4485925925925924,
+      "grad_norm": 2.2999558448791504,
+      "learning_rate": 3.6830665876309036e-06,
+      "loss": 0.146,
+      "step": 4131
+    },
+    {
+      "epoch": 2.4491851851851854,
+      "grad_norm": 2.866795539855957,
+      "learning_rate": 3.6791147994467502e-06,
+      "loss": 0.1925,
+      "step": 4132
+    },
+    {
+      "epoch": 2.449777777777778,
+      "grad_norm": 1.7971819639205933,
+      "learning_rate": 3.675163011262597e-06,
+      "loss": 0.1717,
+      "step": 4133
+    },
+    {
+      "epoch": 2.4503703703703703,
+      "grad_norm": 2.144143581390381,
+      "learning_rate": 3.671211223078443e-06,
+      "loss": 0.1569,
+      "step": 4134
+    },
+    {
+      "epoch": 2.450962962962963,
+      "grad_norm": 2.1367242336273193,
+      "learning_rate": 3.66725943489429e-06,
+      "loss": 0.163,
+      "step": 4135
+    },
+    {
+      "epoch": 2.4515555555555557,
+      "grad_norm": 2.8785946369171143,
+      "learning_rate": 3.6633076467101365e-06,
+      "loss": 0.1994,
+      "step": 4136
+    },
+    {
+      "epoch": 2.452148148148148,
+      "grad_norm": 2.2836389541625977,
+      "learning_rate": 3.659355858525983e-06,
+      "loss": 0.1589,
+      "step": 4137
+    },
+    {
+      "epoch": 2.4527407407407407,
+      "grad_norm": 2.677546739578247,
+      "learning_rate": 3.65540407034183e-06,
+      "loss": 0.2294,
+      "step": 4138
+    },
+    {
+      "epoch": 2.453333333333333,
+      "grad_norm": 1.6014606952667236,
+      "learning_rate": 3.6514522821576765e-06,
+      "loss": 0.1007,
+      "step": 4139
+    },
+    {
+      "epoch": 2.453925925925926,
+      "grad_norm": 2.3745310306549072,
+      "learning_rate": 3.647500493973523e-06,
+      "loss": 0.1595,
+      "step": 4140
+    },
+    {
+      "epoch": 2.4545185185185185,
+      "grad_norm": 2.246236801147461,
+      "learning_rate": 3.64354870578937e-06,
+      "loss": 0.199,
+      "step": 4141
+    },
+    {
+      "epoch": 2.455111111111111,
+      "grad_norm": 1.956180214881897,
+      "learning_rate": 3.6395969176052165e-06,
+      "loss": 0.1689,
+      "step": 4142
+    },
+    {
+      "epoch": 2.455703703703704,
+      "grad_norm": 2.0118823051452637,
+      "learning_rate": 3.635645129421063e-06,
+      "loss": 0.2137,
+      "step": 4143
+    },
+    {
+      "epoch": 2.4562962962962964,
+      "grad_norm": 2.224818706512451,
+      "learning_rate": 3.63169334123691e-06,
+      "loss": 0.121,
+      "step": 4144
+    },
+    {
+      "epoch": 2.456888888888889,
+      "grad_norm": 1.7537124156951904,
+      "learning_rate": 3.6277415530527565e-06,
+      "loss": 0.1298,
+      "step": 4145
+    },
+    {
+      "epoch": 2.4574814814814814,
+      "grad_norm": 2.9222817420959473,
+      "learning_rate": 3.623789764868603e-06,
+      "loss": 0.2562,
+      "step": 4146
+    },
+    {
+      "epoch": 2.458074074074074,
+      "grad_norm": 2.456953763961792,
+      "learning_rate": 3.61983797668445e-06,
+      "loss": 0.1023,
+      "step": 4147
+    },
+    {
+      "epoch": 2.458666666666667,
+      "grad_norm": 2.216259241104126,
+      "learning_rate": 3.6158861885002965e-06,
+      "loss": 0.2449,
+      "step": 4148
+    },
+    {
+      "epoch": 2.4592592592592593,
+      "grad_norm": 2.2803313732147217,
+      "learning_rate": 3.611934400316143e-06,
+      "loss": 0.1633,
+      "step": 4149
+    },
+    {
+      "epoch": 2.4598518518518517,
+      "grad_norm": 2.771428346633911,
+      "learning_rate": 3.60798261213199e-06,
+      "loss": 0.1691,
+      "step": 4150
+    },
+    {
+      "epoch": 2.4604444444444447,
+      "grad_norm": 2.057084560394287,
+      "learning_rate": 3.6040308239478365e-06,
+      "loss": 0.1413,
+      "step": 4151
+    },
+    {
+      "epoch": 2.461037037037037,
+      "grad_norm": 2.022669553756714,
+      "learning_rate": 3.600079035763683e-06,
+      "loss": 0.1828,
+      "step": 4152
+    },
+    {
+      "epoch": 2.4616296296296296,
+      "grad_norm": 2.7182552814483643,
+      "learning_rate": 3.5961272475795302e-06,
+      "loss": 0.16,
+      "step": 4153
+    },
+    {
+      "epoch": 2.462222222222222,
+      "grad_norm": 1.8587899208068848,
+      "learning_rate": 3.592175459395377e-06,
+      "loss": 0.1774,
+      "step": 4154
+    },
+    {
+      "epoch": 2.4628148148148146,
+      "grad_norm": 1.7296494245529175,
+      "learning_rate": 3.5882236712112236e-06,
+      "loss": 0.1751,
+      "step": 4155
+    },
+    {
+      "epoch": 2.4634074074074075,
+      "grad_norm": 1.955978512763977,
+      "learning_rate": 3.5842718830270702e-06,
+      "loss": 0.1909,
+      "step": 4156
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 2.0476341247558594,
+      "learning_rate": 3.580320094842917e-06,
+      "loss": 0.1729,
+      "step": 4157
+    },
+    {
+      "epoch": 2.4645925925925924,
+      "grad_norm": 2.3898861408233643,
+      "learning_rate": 3.5763683066587636e-06,
+      "loss": 0.1872,
+      "step": 4158
+    },
+    {
+      "epoch": 2.4651851851851854,
+      "grad_norm": 1.618474006652832,
+      "learning_rate": 3.5724165184746102e-06,
+      "loss": 0.1356,
+      "step": 4159
+    },
+    {
+      "epoch": 2.465777777777778,
+      "grad_norm": 1.8839960098266602,
+      "learning_rate": 3.568464730290457e-06,
+      "loss": 0.1953,
+      "step": 4160
+    },
+    {
+      "epoch": 2.4663703703703703,
+      "grad_norm": 2.3413257598876953,
+      "learning_rate": 3.5645129421063036e-06,
+      "loss": 0.1674,
+      "step": 4161
+    },
+    {
+      "epoch": 2.466962962962963,
+      "grad_norm": 2.0721278190612793,
+      "learning_rate": 3.5605611539221502e-06,
+      "loss": 0.1357,
+      "step": 4162
+    },
+    {
+      "epoch": 2.4675555555555557,
+      "grad_norm": 1.8851659297943115,
+      "learning_rate": 3.556609365737997e-06,
+      "loss": 0.2087,
+      "step": 4163
+    },
+    {
+      "epoch": 2.468148148148148,
+      "grad_norm": 2.1332974433898926,
+      "learning_rate": 3.5526575775538436e-06,
+      "loss": 0.2762,
+      "step": 4164
+    },
+    {
+      "epoch": 2.4687407407407407,
+      "grad_norm": 1.7659767866134644,
+      "learning_rate": 3.5487057893696902e-06,
+      "loss": 0.1486,
+      "step": 4165
+    },
+    {
+      "epoch": 2.469333333333333,
+      "grad_norm": 1.5994532108306885,
+      "learning_rate": 3.544754001185537e-06,
+      "loss": 0.1778,
+      "step": 4166
+    },
+    {
+      "epoch": 2.469925925925926,
+      "grad_norm": 2.6006150245666504,
+      "learning_rate": 3.5408022130013836e-06,
+      "loss": 0.2031,
+      "step": 4167
+    },
+    {
+      "epoch": 2.4705185185185186,
+      "grad_norm": 2.1145482063293457,
+      "learning_rate": 3.5368504248172302e-06,
+      "loss": 0.1579,
+      "step": 4168
+    },
+    {
+      "epoch": 2.471111111111111,
+      "grad_norm": 3.1392245292663574,
+      "learning_rate": 3.532898636633077e-06,
+      "loss": 0.2217,
+      "step": 4169
+    },
+    {
+      "epoch": 2.4717037037037035,
+      "grad_norm": 2.272494316101074,
+      "learning_rate": 3.5289468484489236e-06,
+      "loss": 0.2306,
+      "step": 4170
+    },
+    {
+      "epoch": 2.4722962962962964,
+      "grad_norm": 1.7262535095214844,
+      "learning_rate": 3.5249950602647702e-06,
+      "loss": 0.1276,
+      "step": 4171
+    },
+    {
+      "epoch": 2.472888888888889,
+      "grad_norm": 2.6594791412353516,
+      "learning_rate": 3.521043272080617e-06,
+      "loss": 0.2427,
+      "step": 4172
+    },
+    {
+      "epoch": 2.4734814814814814,
+      "grad_norm": 2.2355306148529053,
+      "learning_rate": 3.5170914838964635e-06,
+      "loss": 0.2025,
+      "step": 4173
+    },
+    {
+      "epoch": 2.474074074074074,
+      "grad_norm": 1.5815174579620361,
+      "learning_rate": 3.51313969571231e-06,
+      "loss": 0.1947,
+      "step": 4174
+    },
+    {
+      "epoch": 2.474666666666667,
+      "grad_norm": 2.4270639419555664,
+      "learning_rate": 3.5091879075281565e-06,
+      "loss": 0.2846,
+      "step": 4175
+    },
+    {
+      "epoch": 2.4752592592592593,
+      "grad_norm": 2.150735378265381,
+      "learning_rate": 3.505236119344003e-06,
+      "loss": 0.1797,
+      "step": 4176
+    },
+    {
+      "epoch": 2.4758518518518517,
+      "grad_norm": 3.6499950885772705,
+      "learning_rate": 3.50128433115985e-06,
+      "loss": 0.2883,
+      "step": 4177
+    },
+    {
+      "epoch": 2.4764444444444447,
+      "grad_norm": 3.5157811641693115,
+      "learning_rate": 3.4973325429756965e-06,
+      "loss": 0.3045,
+      "step": 4178
+    },
+    {
+      "epoch": 2.477037037037037,
+      "grad_norm": 2.6906490325927734,
+      "learning_rate": 3.493380754791543e-06,
+      "loss": 0.206,
+      "step": 4179
+    },
+    {
+      "epoch": 2.4776296296296296,
+      "grad_norm": 2.633241891860962,
+      "learning_rate": 3.48942896660739e-06,
+      "loss": 0.1718,
+      "step": 4180
+    },
+    {
+      "epoch": 2.478222222222222,
+      "grad_norm": 2.1251678466796875,
+      "learning_rate": 3.4854771784232365e-06,
+      "loss": 0.2544,
+      "step": 4181
+    },
+    {
+      "epoch": 2.4788148148148146,
+      "grad_norm": 1.9892303943634033,
+      "learning_rate": 3.481525390239083e-06,
+      "loss": 0.1688,
+      "step": 4182
+    },
+    {
+      "epoch": 2.4794074074074075,
+      "grad_norm": 2.3079633712768555,
+      "learning_rate": 3.47757360205493e-06,
+      "loss": 0.1263,
+      "step": 4183
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.027886390686035,
+      "learning_rate": 3.473621813870777e-06,
+      "loss": 0.1996,
+      "step": 4184
+    },
+    {
+      "epoch": 2.4805925925925925,
+      "grad_norm": 2.256727933883667,
+      "learning_rate": 3.4696700256866235e-06,
+      "loss": 0.1326,
+      "step": 4185
+    },
+    {
+      "epoch": 2.4811851851851854,
+      "grad_norm": 1.8859831094741821,
+      "learning_rate": 3.4657182375024702e-06,
+      "loss": 0.1547,
+      "step": 4186
+    },
+    {
+      "epoch": 2.481777777777778,
+      "grad_norm": 2.695997953414917,
+      "learning_rate": 3.461766449318317e-06,
+      "loss": 0.2081,
+      "step": 4187
+    },
+    {
+      "epoch": 2.4823703703703703,
+      "grad_norm": 2.6018621921539307,
+      "learning_rate": 3.4578146611341635e-06,
+      "loss": 0.1423,
+      "step": 4188
+    },
+    {
+      "epoch": 2.482962962962963,
+      "grad_norm": 1.3570506572723389,
+      "learning_rate": 3.45386287295001e-06,
+      "loss": 0.1427,
+      "step": 4189
+    },
+    {
+      "epoch": 2.4835555555555557,
+      "grad_norm": 2.2077324390411377,
+      "learning_rate": 3.449911084765857e-06,
+      "loss": 0.233,
+      "step": 4190
+    },
+    {
+      "epoch": 2.484148148148148,
+      "grad_norm": 2.7682642936706543,
+      "learning_rate": 3.4459592965817035e-06,
+      "loss": 0.2014,
+      "step": 4191
+    },
+    {
+      "epoch": 2.4847407407407407,
+      "grad_norm": 1.937550663948059,
+      "learning_rate": 3.44200750839755e-06,
+      "loss": 0.1848,
+      "step": 4192
+    },
+    {
+      "epoch": 2.485333333333333,
+      "grad_norm": 2.489137887954712,
+      "learning_rate": 3.438055720213397e-06,
+      "loss": 0.2719,
+      "step": 4193
+    },
+    {
+      "epoch": 2.485925925925926,
+      "grad_norm": 1.8809239864349365,
+      "learning_rate": 3.4341039320292435e-06,
+      "loss": 0.09,
+      "step": 4194
+    },
+    {
+      "epoch": 2.4865185185185186,
+      "grad_norm": 1.864810824394226,
+      "learning_rate": 3.43015214384509e-06,
+      "loss": 0.1436,
+      "step": 4195
+    },
+    {
+      "epoch": 2.487111111111111,
+      "grad_norm": 1.7839359045028687,
+      "learning_rate": 3.426200355660937e-06,
+      "loss": 0.1397,
+      "step": 4196
+    },
+    {
+      "epoch": 2.4877037037037035,
+      "grad_norm": 1.9025046825408936,
+      "learning_rate": 3.4222485674767835e-06,
+      "loss": 0.1398,
+      "step": 4197
+    },
+    {
+      "epoch": 2.4882962962962965,
+      "grad_norm": 2.0975615978240967,
+      "learning_rate": 3.41829677929263e-06,
+      "loss": 0.1544,
+      "step": 4198
+    },
+    {
+      "epoch": 2.488888888888889,
+      "grad_norm": 1.4334309101104736,
+      "learning_rate": 3.414344991108477e-06,
+      "loss": 0.0912,
+      "step": 4199
+    },
+    {
+      "epoch": 2.4894814814814814,
+      "grad_norm": 1.7467243671417236,
+      "learning_rate": 3.4103932029243235e-06,
+      "loss": 0.1233,
+      "step": 4200
+    },
+    {
+      "epoch": 2.490074074074074,
+      "grad_norm": 2.3934237957000732,
+      "learning_rate": 3.40644141474017e-06,
+      "loss": 0.1999,
+      "step": 4201
+    },
+    {
+      "epoch": 2.490666666666667,
+      "grad_norm": 2.1507158279418945,
+      "learning_rate": 3.402489626556017e-06,
+      "loss": 0.1464,
+      "step": 4202
+    },
+    {
+      "epoch": 2.4912592592592593,
+      "grad_norm": 2.505985736846924,
+      "learning_rate": 3.3985378383718635e-06,
+      "loss": 0.2232,
+      "step": 4203
+    },
+    {
+      "epoch": 2.4918518518518518,
+      "grad_norm": 2.5263736248016357,
+      "learning_rate": 3.39458605018771e-06,
+      "loss": 0.2903,
+      "step": 4204
+    },
+    {
+      "epoch": 2.4924444444444447,
+      "grad_norm": 2.2222094535827637,
+      "learning_rate": 3.390634262003557e-06,
+      "loss": 0.1193,
+      "step": 4205
+    },
+    {
+      "epoch": 2.493037037037037,
+      "grad_norm": 1.877442479133606,
+      "learning_rate": 3.3866824738194035e-06,
+      "loss": 0.1423,
+      "step": 4206
+    },
+    {
+      "epoch": 2.4936296296296296,
+      "grad_norm": 2.5431606769561768,
+      "learning_rate": 3.3827306856352506e-06,
+      "loss": 0.1854,
+      "step": 4207
+    },
+    {
+      "epoch": 2.494222222222222,
+      "grad_norm": 1.8383243083953857,
+      "learning_rate": 3.3787788974510973e-06,
+      "loss": 0.168,
+      "step": 4208
+    },
+    {
+      "epoch": 2.4948148148148146,
+      "grad_norm": 2.454209089279175,
+      "learning_rate": 3.374827109266944e-06,
+      "loss": 0.118,
+      "step": 4209
+    },
+    {
+      "epoch": 2.4954074074074075,
+      "grad_norm": 2.011122226715088,
+      "learning_rate": 3.3708753210827906e-06,
+      "loss": 0.1468,
+      "step": 4210
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 1.4830987453460693,
+      "learning_rate": 3.3669235328986373e-06,
+      "loss": 0.1484,
+      "step": 4211
+    },
+    {
+      "epoch": 2.4965925925925925,
+      "grad_norm": 1.8095155954360962,
+      "learning_rate": 3.362971744714484e-06,
+      "loss": 0.1885,
+      "step": 4212
+    },
+    {
+      "epoch": 2.4971851851851854,
+      "grad_norm": 2.19488263130188,
+      "learning_rate": 3.3590199565303298e-06,
+      "loss": 0.1921,
+      "step": 4213
+    },
+    {
+      "epoch": 2.497777777777778,
+      "grad_norm": 2.860172748565674,
+      "learning_rate": 3.3550681683461764e-06,
+      "loss": 0.221,
+      "step": 4214
+    },
+    {
+      "epoch": 2.4983703703703704,
+      "grad_norm": 1.5130575895309448,
+      "learning_rate": 3.3511163801620235e-06,
+      "loss": 0.1398,
+      "step": 4215
+    },
+    {
+      "epoch": 2.498962962962963,
+      "grad_norm": 2.4075801372528076,
+      "learning_rate": 3.34716459197787e-06,
+      "loss": 0.1742,
+      "step": 4216
+    },
+    {
+      "epoch": 2.4995555555555553,
+      "grad_norm": 2.344231367111206,
+      "learning_rate": 3.343212803793717e-06,
+      "loss": 0.2164,
+      "step": 4217
+    },
+    {
+      "epoch": 2.5001481481481482,
+      "grad_norm": 2.301759719848633,
+      "learning_rate": 3.3392610156095635e-06,
+      "loss": 0.2063,
+      "step": 4218
+    },
+    {
+      "epoch": 2.5007407407407407,
+      "grad_norm": 3.3594439029693604,
+      "learning_rate": 3.33530922742541e-06,
+      "loss": 0.1744,
+      "step": 4219
+    },
+    {
+      "epoch": 2.501333333333333,
+      "grad_norm": 2.070070505142212,
+      "learning_rate": 3.331357439241257e-06,
+      "loss": 0.1435,
+      "step": 4220
+    },
+    {
+      "epoch": 2.501925925925926,
+      "grad_norm": 1.7042511701583862,
+      "learning_rate": 3.3274056510571035e-06,
+      "loss": 0.0964,
+      "step": 4221
+    },
+    {
+      "epoch": 2.5025185185185186,
+      "grad_norm": 2.8826382160186768,
+      "learning_rate": 3.32345386287295e-06,
+      "loss": 0.2852,
+      "step": 4222
+    },
+    {
+      "epoch": 2.503111111111111,
+      "grad_norm": 1.8184891939163208,
+      "learning_rate": 3.319502074688797e-06,
+      "loss": 0.1577,
+      "step": 4223
+    },
+    {
+      "epoch": 2.5037037037037035,
+      "grad_norm": 2.5945305824279785,
+      "learning_rate": 3.3155502865046435e-06,
+      "loss": 0.201,
+      "step": 4224
+    },
+    {
+      "epoch": 2.504296296296296,
+      "grad_norm": 2.7452094554901123,
+      "learning_rate": 3.31159849832049e-06,
+      "loss": 0.2142,
+      "step": 4225
+    },
+    {
+      "epoch": 2.504888888888889,
+      "grad_norm": 2.599571704864502,
+      "learning_rate": 3.307646710136337e-06,
+      "loss": 0.1577,
+      "step": 4226
+    },
+    {
+      "epoch": 2.5054814814814814,
+      "grad_norm": 2.069239854812622,
+      "learning_rate": 3.3036949219521835e-06,
+      "loss": 0.1634,
+      "step": 4227
+    },
+    {
+      "epoch": 2.506074074074074,
+      "grad_norm": 2.2766902446746826,
+      "learning_rate": 3.29974313376803e-06,
+      "loss": 0.1098,
+      "step": 4228
+    },
+    {
+      "epoch": 2.506666666666667,
+      "grad_norm": 1.770334243774414,
+      "learning_rate": 3.295791345583877e-06,
+      "loss": 0.1512,
+      "step": 4229
+    },
+    {
+      "epoch": 2.5072592592592593,
+      "grad_norm": 1.9395313262939453,
+      "learning_rate": 3.2918395573997235e-06,
+      "loss": 0.1742,
+      "step": 4230
+    },
+    {
+      "epoch": 2.5078518518518518,
+      "grad_norm": 3.359776496887207,
+      "learning_rate": 3.28788776921557e-06,
+      "loss": 0.2191,
+      "step": 4231
+    },
+    {
+      "epoch": 2.5084444444444447,
+      "grad_norm": 2.248232841491699,
+      "learning_rate": 3.283935981031417e-06,
+      "loss": 0.1585,
+      "step": 4232
+    },
+    {
+      "epoch": 2.509037037037037,
+      "grad_norm": 1.6591346263885498,
+      "learning_rate": 3.2799841928472635e-06,
+      "loss": 0.1075,
+      "step": 4233
+    },
+    {
+      "epoch": 2.5096296296296297,
+      "grad_norm": 2.4015135765075684,
+      "learning_rate": 3.27603240466311e-06,
+      "loss": 0.1784,
+      "step": 4234
+    },
+    {
+      "epoch": 2.510222222222222,
+      "grad_norm": 2.1409759521484375,
+      "learning_rate": 3.272080616478957e-06,
+      "loss": 0.2,
+      "step": 4235
+    },
+    {
+      "epoch": 2.5108148148148146,
+      "grad_norm": 2.852079153060913,
+      "learning_rate": 3.2681288282948035e-06,
+      "loss": 0.2532,
+      "step": 4236
+    },
+    {
+      "epoch": 2.5114074074074075,
+      "grad_norm": 2.7875702381134033,
+      "learning_rate": 3.26417704011065e-06,
+      "loss": 0.2019,
+      "step": 4237
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 1.904557466506958,
+      "learning_rate": 3.2602252519264973e-06,
+      "loss": 0.1193,
+      "step": 4238
+    },
+    {
+      "epoch": 2.5125925925925925,
+      "grad_norm": 2.012772798538208,
+      "learning_rate": 3.256273463742344e-06,
+      "loss": 0.164,
+      "step": 4239
+    },
+    {
+      "epoch": 2.5131851851851854,
+      "grad_norm": 3.7314178943634033,
+      "learning_rate": 3.2523216755581906e-06,
+      "loss": 0.2101,
+      "step": 4240
+    },
+    {
+      "epoch": 2.513777777777778,
+      "grad_norm": 1.6154731512069702,
+      "learning_rate": 3.2483698873740373e-06,
+      "loss": 0.1559,
+      "step": 4241
+    },
+    {
+      "epoch": 2.5143703703703704,
+      "grad_norm": 1.8741594552993774,
+      "learning_rate": 3.244418099189884e-06,
+      "loss": 0.1236,
+      "step": 4242
+    },
+    {
+      "epoch": 2.514962962962963,
+      "grad_norm": 2.169672727584839,
+      "learning_rate": 3.2404663110057306e-06,
+      "loss": 0.1754,
+      "step": 4243
+    },
+    {
+      "epoch": 2.5155555555555553,
+      "grad_norm": 1.9006346464157104,
+      "learning_rate": 3.2365145228215773e-06,
+      "loss": 0.1914,
+      "step": 4244
+    },
+    {
+      "epoch": 2.5161481481481482,
+      "grad_norm": 2.1808791160583496,
+      "learning_rate": 3.232562734637424e-06,
+      "loss": 0.2178,
+      "step": 4245
+    },
+    {
+      "epoch": 2.5167407407407407,
+      "grad_norm": 2.005039930343628,
+      "learning_rate": 3.2286109464532706e-06,
+      "loss": 0.1335,
+      "step": 4246
+    },
+    {
+      "epoch": 2.517333333333333,
+      "grad_norm": 2.788930892944336,
+      "learning_rate": 3.2246591582691173e-06,
+      "loss": 0.2056,
+      "step": 4247
+    },
+    {
+      "epoch": 2.517925925925926,
+      "grad_norm": 2.177013635635376,
+      "learning_rate": 3.220707370084964e-06,
+      "loss": 0.1851,
+      "step": 4248
+    },
+    {
+      "epoch": 2.5185185185185186,
+      "grad_norm": 2.41276478767395,
+      "learning_rate": 3.2167555819008106e-06,
+      "loss": 0.1566,
+      "step": 4249
+    },
+    {
+      "epoch": 2.519111111111111,
+      "grad_norm": 1.9796184301376343,
+      "learning_rate": 3.2128037937166573e-06,
+      "loss": 0.111,
+      "step": 4250
+    },
+    {
+      "epoch": 2.5197037037037036,
+      "grad_norm": 2.2980735301971436,
+      "learning_rate": 3.208852005532504e-06,
+      "loss": 0.1672,
+      "step": 4251
+    },
+    {
+      "epoch": 2.520296296296296,
+      "grad_norm": 2.4502618312835693,
+      "learning_rate": 3.2049002173483506e-06,
+      "loss": 0.1825,
+      "step": 4252
+    },
+    {
+      "epoch": 2.520888888888889,
+      "grad_norm": 2.34709095954895,
+      "learning_rate": 3.200948429164197e-06,
+      "loss": 0.2349,
+      "step": 4253
+    },
+    {
+      "epoch": 2.5214814814814814,
+      "grad_norm": 2.18560528755188,
+      "learning_rate": 3.1969966409800435e-06,
+      "loss": 0.2215,
+      "step": 4254
+    },
+    {
+      "epoch": 2.522074074074074,
+      "grad_norm": 2.401235580444336,
+      "learning_rate": 3.19304485279589e-06,
+      "loss": 0.192,
+      "step": 4255
+    },
+    {
+      "epoch": 2.522666666666667,
+      "grad_norm": 1.798515796661377,
+      "learning_rate": 3.189093064611737e-06,
+      "loss": 0.1089,
+      "step": 4256
+    },
+    {
+      "epoch": 2.5232592592592593,
+      "grad_norm": 3.093364715576172,
+      "learning_rate": 3.1851412764275835e-06,
+      "loss": 0.2198,
+      "step": 4257
+    },
+    {
+      "epoch": 2.523851851851852,
+      "grad_norm": 2.3798632621765137,
+      "learning_rate": 3.18118948824343e-06,
+      "loss": 0.1542,
+      "step": 4258
+    },
+    {
+      "epoch": 2.5244444444444447,
+      "grad_norm": 2.241682291030884,
+      "learning_rate": 3.177237700059277e-06,
+      "loss": 0.2384,
+      "step": 4259
+    },
+    {
+      "epoch": 2.525037037037037,
+      "grad_norm": 1.828101634979248,
+      "learning_rate": 3.1732859118751235e-06,
+      "loss": 0.132,
+      "step": 4260
+    },
+    {
+      "epoch": 2.5256296296296297,
+      "grad_norm": 1.8989614248275757,
+      "learning_rate": 3.16933412369097e-06,
+      "loss": 0.1909,
+      "step": 4261
+    },
+    {
+      "epoch": 2.526222222222222,
+      "grad_norm": 2.556321859359741,
+      "learning_rate": 3.165382335506817e-06,
+      "loss": 0.2105,
+      "step": 4262
+    },
+    {
+      "epoch": 2.5268148148148146,
+      "grad_norm": 2.301823616027832,
+      "learning_rate": 3.1614305473226635e-06,
+      "loss": 0.2255,
+      "step": 4263
+    },
+    {
+      "epoch": 2.5274074074074075,
+      "grad_norm": 2.6590445041656494,
+      "learning_rate": 3.15747875913851e-06,
+      "loss": 0.1534,
+      "step": 4264
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 1.5971342325210571,
+      "learning_rate": 3.153526970954357e-06,
+      "loss": 0.1299,
+      "step": 4265
+    },
+    {
+      "epoch": 2.5285925925925925,
+      "grad_norm": 2.795004367828369,
+      "learning_rate": 3.1495751827702035e-06,
+      "loss": 0.1112,
+      "step": 4266
+    },
+    {
+      "epoch": 2.5291851851851854,
+      "grad_norm": 2.4354496002197266,
+      "learning_rate": 3.14562339458605e-06,
+      "loss": 0.1778,
+      "step": 4267
+    },
+    {
+      "epoch": 2.529777777777778,
+      "grad_norm": 1.3850913047790527,
+      "learning_rate": 3.141671606401897e-06,
+      "loss": 0.1063,
+      "step": 4268
+    },
+    {
+      "epoch": 2.5303703703703704,
+      "grad_norm": 1.7495551109313965,
+      "learning_rate": 3.137719818217744e-06,
+      "loss": 0.1208,
+      "step": 4269
+    },
+    {
+      "epoch": 2.530962962962963,
+      "grad_norm": 1.8498097658157349,
+      "learning_rate": 3.1337680300335906e-06,
+      "loss": 0.2553,
+      "step": 4270
+    },
+    {
+      "epoch": 2.5315555555555553,
+      "grad_norm": 2.6717538833618164,
+      "learning_rate": 3.1298162418494373e-06,
+      "loss": 0.1688,
+      "step": 4271
+    },
+    {
+      "epoch": 2.5321481481481483,
+      "grad_norm": 2.8559916019439697,
+      "learning_rate": 3.125864453665284e-06,
+      "loss": 0.2086,
+      "step": 4272
+    },
+    {
+      "epoch": 2.5327407407407407,
+      "grad_norm": 2.3744349479675293,
+      "learning_rate": 3.1219126654811306e-06,
+      "loss": 0.2565,
+      "step": 4273
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 2.236628532409668,
+      "learning_rate": 3.1179608772969773e-06,
+      "loss": 0.1356,
+      "step": 4274
+    },
+    {
+      "epoch": 2.533925925925926,
+      "grad_norm": 2.3324573040008545,
+      "learning_rate": 3.114009089112824e-06,
+      "loss": 0.1647,
+      "step": 4275
+    },
+    {
+      "epoch": 2.5345185185185186,
+      "grad_norm": 1.7959003448486328,
+      "learning_rate": 3.1100573009286706e-06,
+      "loss": 0.104,
+      "step": 4276
+    },
+    {
+      "epoch": 2.535111111111111,
+      "grad_norm": 4.586813926696777,
+      "learning_rate": 3.1061055127445173e-06,
+      "loss": 0.1981,
+      "step": 4277
+    },
+    {
+      "epoch": 2.5357037037037036,
+      "grad_norm": 1.8762515783309937,
+      "learning_rate": 3.102153724560364e-06,
+      "loss": 0.1651,
+      "step": 4278
+    },
+    {
+      "epoch": 2.536296296296296,
+      "grad_norm": 2.7932794094085693,
+      "learning_rate": 3.0982019363762106e-06,
+      "loss": 0.1832,
+      "step": 4279
+    },
+    {
+      "epoch": 2.536888888888889,
+      "grad_norm": 2.4161195755004883,
+      "learning_rate": 3.0942501481920573e-06,
+      "loss": 0.193,
+      "step": 4280
+    },
+    {
+      "epoch": 2.5374814814814814,
+      "grad_norm": 5.904382705688477,
+      "learning_rate": 3.090298360007904e-06,
+      "loss": 0.2949,
+      "step": 4281
+    },
+    {
+      "epoch": 2.538074074074074,
+      "grad_norm": 1.951417326927185,
+      "learning_rate": 3.0863465718237506e-06,
+      "loss": 0.1736,
+      "step": 4282
+    },
+    {
+      "epoch": 2.538666666666667,
+      "grad_norm": 2.2860565185546875,
+      "learning_rate": 3.0823947836395973e-06,
+      "loss": 0.1725,
+      "step": 4283
+    },
+    {
+      "epoch": 2.5392592592592593,
+      "grad_norm": 2.5856025218963623,
+      "learning_rate": 3.078442995455444e-06,
+      "loss": 0.1908,
+      "step": 4284
+    },
+    {
+      "epoch": 2.539851851851852,
+      "grad_norm": 1.7584205865859985,
+      "learning_rate": 3.0744912072712906e-06,
+      "loss": 0.1622,
+      "step": 4285
+    },
+    {
+      "epoch": 2.5404444444444443,
+      "grad_norm": 2.3891637325286865,
+      "learning_rate": 3.0705394190871373e-06,
+      "loss": 0.1636,
+      "step": 4286
+    },
+    {
+      "epoch": 2.5410370370370368,
+      "grad_norm": 2.560389995574951,
+      "learning_rate": 3.066587630902984e-06,
+      "loss": 0.155,
+      "step": 4287
+    },
+    {
+      "epoch": 2.5416296296296297,
+      "grad_norm": 2.4343836307525635,
+      "learning_rate": 3.0626358427188306e-06,
+      "loss": 0.1626,
+      "step": 4288
+    },
+    {
+      "epoch": 2.542222222222222,
+      "grad_norm": 2.6061394214630127,
+      "learning_rate": 3.0586840545346773e-06,
+      "loss": 0.2325,
+      "step": 4289
+    },
+    {
+      "epoch": 2.5428148148148146,
+      "grad_norm": 2.0041353702545166,
+      "learning_rate": 3.054732266350524e-06,
+      "loss": 0.1003,
+      "step": 4290
+    },
+    {
+      "epoch": 2.5434074074074076,
+      "grad_norm": 1.4518014192581177,
+      "learning_rate": 3.0507804781663706e-06,
+      "loss": 0.1065,
+      "step": 4291
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 2.763960123062134,
+      "learning_rate": 3.046828689982217e-06,
+      "loss": 0.2092,
+      "step": 4292
+    },
+    {
+      "epoch": 2.5445925925925925,
+      "grad_norm": 2.149127244949341,
+      "learning_rate": 3.0428769017980635e-06,
+      "loss": 0.1819,
+      "step": 4293
+    },
+    {
+      "epoch": 2.5451851851851854,
+      "grad_norm": 2.084062099456787,
+      "learning_rate": 3.03892511361391e-06,
+      "loss": 0.1363,
+      "step": 4294
+    },
+    {
+      "epoch": 2.545777777777778,
+      "grad_norm": 2.7990164756774902,
+      "learning_rate": 3.034973325429757e-06,
+      "loss": 0.1798,
+      "step": 4295
+    },
+    {
+      "epoch": 2.5463703703703704,
+      "grad_norm": 1.7398302555084229,
+      "learning_rate": 3.0310215372456035e-06,
+      "loss": 0.1278,
+      "step": 4296
+    },
+    {
+      "epoch": 2.546962962962963,
+      "grad_norm": 2.268254280090332,
+      "learning_rate": 3.02706974906145e-06,
+      "loss": 0.1722,
+      "step": 4297
+    },
+    {
+      "epoch": 2.5475555555555554,
+      "grad_norm": 2.4567043781280518,
+      "learning_rate": 3.023117960877297e-06,
+      "loss": 0.1799,
+      "step": 4298
+    },
+    {
+      "epoch": 2.5481481481481483,
+      "grad_norm": 1.9626532793045044,
+      "learning_rate": 3.0191661726931435e-06,
+      "loss": 0.1495,
+      "step": 4299
+    },
+    {
+      "epoch": 2.5487407407407408,
+      "grad_norm": 3.099856376647949,
+      "learning_rate": 3.0152143845089906e-06,
+      "loss": 0.2375,
+      "step": 4300
+    },
+    {
+      "epoch": 2.5493333333333332,
+      "grad_norm": 2.9612841606140137,
+      "learning_rate": 3.0112625963248373e-06,
+      "loss": 0.1905,
+      "step": 4301
+    },
+    {
+      "epoch": 2.549925925925926,
+      "grad_norm": 1.681656002998352,
+      "learning_rate": 3.007310808140684e-06,
+      "loss": 0.1022,
+      "step": 4302
+    },
+    {
+      "epoch": 2.5505185185185186,
+      "grad_norm": 1.6995004415512085,
+      "learning_rate": 3.0033590199565306e-06,
+      "loss": 0.1641,
+      "step": 4303
+    },
+    {
+      "epoch": 2.551111111111111,
+      "grad_norm": 3.1439783573150635,
+      "learning_rate": 2.9994072317723773e-06,
+      "loss": 0.2008,
+      "step": 4304
+    },
+    {
+      "epoch": 2.5517037037037036,
+      "grad_norm": 1.778039574623108,
+      "learning_rate": 2.995455443588224e-06,
+      "loss": 0.1515,
+      "step": 4305
+    },
+    {
+      "epoch": 2.552296296296296,
+      "grad_norm": 1.8307173252105713,
+      "learning_rate": 2.9915036554040706e-06,
+      "loss": 0.1055,
+      "step": 4306
+    },
+    {
+      "epoch": 2.552888888888889,
+      "grad_norm": 2.7373054027557373,
+      "learning_rate": 2.9875518672199173e-06,
+      "loss": 0.2135,
+      "step": 4307
+    },
+    {
+      "epoch": 2.5534814814814815,
+      "grad_norm": 2.0729808807373047,
+      "learning_rate": 2.983600079035764e-06,
+      "loss": 0.1914,
+      "step": 4308
+    },
+    {
+      "epoch": 2.554074074074074,
+      "grad_norm": 3.224432945251465,
+      "learning_rate": 2.9796482908516106e-06,
+      "loss": 0.1934,
+      "step": 4309
+    },
+    {
+      "epoch": 2.554666666666667,
+      "grad_norm": 2.378563165664673,
+      "learning_rate": 2.9756965026674573e-06,
+      "loss": 0.1179,
+      "step": 4310
+    },
+    {
+      "epoch": 2.5552592592592593,
+      "grad_norm": 2.3982090950012207,
+      "learning_rate": 2.971744714483304e-06,
+      "loss": 0.1852,
+      "step": 4311
+    },
+    {
+      "epoch": 2.555851851851852,
+      "grad_norm": 2.2060582637786865,
+      "learning_rate": 2.9677929262991506e-06,
+      "loss": 0.1842,
+      "step": 4312
+    },
+    {
+      "epoch": 2.5564444444444443,
+      "grad_norm": 2.6414408683776855,
+      "learning_rate": 2.9638411381149973e-06,
+      "loss": 0.2053,
+      "step": 4313
+    },
+    {
+      "epoch": 2.5570370370370368,
+      "grad_norm": 1.9975810050964355,
+      "learning_rate": 2.959889349930844e-06,
+      "loss": 0.1693,
+      "step": 4314
+    },
+    {
+      "epoch": 2.5576296296296297,
+      "grad_norm": 2.5404694080352783,
+      "learning_rate": 2.9559375617466906e-06,
+      "loss": 0.1578,
+      "step": 4315
+    },
+    {
+      "epoch": 2.558222222222222,
+      "grad_norm": 2.953011989593506,
+      "learning_rate": 2.9519857735625373e-06,
+      "loss": 0.1905,
+      "step": 4316
+    },
+    {
+      "epoch": 2.5588148148148147,
+      "grad_norm": 2.570457935333252,
+      "learning_rate": 2.948033985378384e-06,
+      "loss": 0.2844,
+      "step": 4317
+    },
+    {
+      "epoch": 2.5594074074074076,
+      "grad_norm": 1.8348406553268433,
+      "learning_rate": 2.9440821971942306e-06,
+      "loss": 0.2064,
+      "step": 4318
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.17959451675415,
+      "learning_rate": 2.9401304090100773e-06,
+      "loss": 0.2558,
+      "step": 4319
+    },
+    {
+      "epoch": 2.5605925925925925,
+      "grad_norm": 2.0839970111846924,
+      "learning_rate": 2.936178620825924e-06,
+      "loss": 0.1859,
+      "step": 4320
+    },
+    {
+      "epoch": 2.5611851851851855,
+      "grad_norm": 2.625849485397339,
+      "learning_rate": 2.9322268326417706e-06,
+      "loss": 0.2091,
+      "step": 4321
+    },
+    {
+      "epoch": 2.561777777777778,
+      "grad_norm": 2.483279228210449,
+      "learning_rate": 2.9282750444576173e-06,
+      "loss": 0.1817,
+      "step": 4322
+    },
+    {
+      "epoch": 2.5623703703703704,
+      "grad_norm": 1.471397876739502,
+      "learning_rate": 2.9243232562734644e-06,
+      "loss": 0.1091,
+      "step": 4323
+    },
+    {
+      "epoch": 2.562962962962963,
+      "grad_norm": 2.8086116313934326,
+      "learning_rate": 2.920371468089311e-06,
+      "loss": 0.2157,
+      "step": 4324
+    },
+    {
+      "epoch": 2.5635555555555554,
+      "grad_norm": 1.9945788383483887,
+      "learning_rate": 2.9164196799051577e-06,
+      "loss": 0.1486,
+      "step": 4325
+    },
+    {
+      "epoch": 2.5641481481481483,
+      "grad_norm": 1.9480401277542114,
+      "learning_rate": 2.9124678917210044e-06,
+      "loss": 0.1477,
+      "step": 4326
+    },
+    {
+      "epoch": 2.5647407407407408,
+      "grad_norm": 1.9323956966400146,
+      "learning_rate": 2.908516103536851e-06,
+      "loss": 0.0925,
+      "step": 4327
+    },
+    {
+      "epoch": 2.5653333333333332,
+      "grad_norm": 2.6505465507507324,
+      "learning_rate": 2.9045643153526977e-06,
+      "loss": 0.1833,
+      "step": 4328
+    },
+    {
+      "epoch": 2.565925925925926,
+      "grad_norm": 2.141648530960083,
+      "learning_rate": 2.9006125271685443e-06,
+      "loss": 0.1592,
+      "step": 4329
+    },
+    {
+      "epoch": 2.5665185185185186,
+      "grad_norm": 2.3575592041015625,
+      "learning_rate": 2.896660738984391e-06,
+      "loss": 0.1714,
+      "step": 4330
+    },
+    {
+      "epoch": 2.567111111111111,
+      "grad_norm": 1.664380669593811,
+      "learning_rate": 2.8927089508002377e-06,
+      "loss": 0.1552,
+      "step": 4331
+    },
+    {
+      "epoch": 2.5677037037037036,
+      "grad_norm": 2.0843770503997803,
+      "learning_rate": 2.888757162616084e-06,
+      "loss": 0.1968,
+      "step": 4332
+    },
+    {
+      "epoch": 2.568296296296296,
+      "grad_norm": 2.0367414951324463,
+      "learning_rate": 2.8848053744319306e-06,
+      "loss": 0.1382,
+      "step": 4333
+    },
+    {
+      "epoch": 2.568888888888889,
+      "grad_norm": 2.319258213043213,
+      "learning_rate": 2.8808535862477773e-06,
+      "loss": 0.2229,
+      "step": 4334
+    },
+    {
+      "epoch": 2.5694814814814815,
+      "grad_norm": 2.6562561988830566,
+      "learning_rate": 2.876901798063624e-06,
+      "loss": 0.2076,
+      "step": 4335
+    },
+    {
+      "epoch": 2.570074074074074,
+      "grad_norm": 3.1283771991729736,
+      "learning_rate": 2.8729500098794706e-06,
+      "loss": 0.1427,
+      "step": 4336
+    },
+    {
+      "epoch": 2.570666666666667,
+      "grad_norm": 1.8350578546524048,
+      "learning_rate": 2.8689982216953173e-06,
+      "loss": 0.1719,
+      "step": 4337
+    },
+    {
+      "epoch": 2.5712592592592594,
+      "grad_norm": 2.111640691757202,
+      "learning_rate": 2.865046433511164e-06,
+      "loss": 0.1387,
+      "step": 4338
+    },
+    {
+      "epoch": 2.571851851851852,
+      "grad_norm": 2.728557825088501,
+      "learning_rate": 2.8610946453270106e-06,
+      "loss": 0.2312,
+      "step": 4339
+    },
+    {
+      "epoch": 2.5724444444444443,
+      "grad_norm": 1.2628298997879028,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.1333,
+      "step": 4340
+    },
+    {
+      "epoch": 2.573037037037037,
+      "grad_norm": 1.6589550971984863,
+      "learning_rate": 2.853191068958704e-06,
+      "loss": 0.1638,
+      "step": 4341
+    },
+    {
+      "epoch": 2.5736296296296297,
+      "grad_norm": 2.3568878173828125,
+      "learning_rate": 2.8492392807745506e-06,
+      "loss": 0.2254,
+      "step": 4342
+    },
+    {
+      "epoch": 2.574222222222222,
+      "grad_norm": 2.6705262660980225,
+      "learning_rate": 2.8452874925903973e-06,
+      "loss": 0.1844,
+      "step": 4343
+    },
+    {
+      "epoch": 2.5748148148148147,
+      "grad_norm": 2.7045845985412598,
+      "learning_rate": 2.841335704406244e-06,
+      "loss": 0.31,
+      "step": 4344
+    },
+    {
+      "epoch": 2.5754074074074076,
+      "grad_norm": 2.440399408340454,
+      "learning_rate": 2.8373839162220906e-06,
+      "loss": 0.1889,
+      "step": 4345
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 3.3908028602600098,
+      "learning_rate": 2.8334321280379373e-06,
+      "loss": 0.2453,
+      "step": 4346
+    },
+    {
+      "epoch": 2.5765925925925925,
+      "grad_norm": 2.259267807006836,
+      "learning_rate": 2.829480339853784e-06,
+      "loss": 0.2281,
+      "step": 4347
+    },
+    {
+      "epoch": 2.577185185185185,
+      "grad_norm": 1.8492752313613892,
+      "learning_rate": 2.8255285516696306e-06,
+      "loss": 0.1087,
+      "step": 4348
+    },
+    {
+      "epoch": 2.5777777777777775,
+      "grad_norm": 2.7265090942382812,
+      "learning_rate": 2.8215767634854773e-06,
+      "loss": 0.2085,
+      "step": 4349
+    },
+    {
+      "epoch": 2.5783703703703704,
+      "grad_norm": 2.1760473251342773,
+      "learning_rate": 2.817624975301324e-06,
+      "loss": 0.1855,
+      "step": 4350
+    },
+    {
+      "epoch": 2.578962962962963,
+      "grad_norm": 1.8974623680114746,
+      "learning_rate": 2.8136731871171706e-06,
+      "loss": 0.1091,
+      "step": 4351
+    },
+    {
+      "epoch": 2.5795555555555554,
+      "grad_norm": 3.5990757942199707,
+      "learning_rate": 2.8097213989330173e-06,
+      "loss": 0.2574,
+      "step": 4352
+    },
+    {
+      "epoch": 2.5801481481481483,
+      "grad_norm": 1.9179513454437256,
+      "learning_rate": 2.805769610748864e-06,
+      "loss": 0.1763,
+      "step": 4353
+    },
+    {
+      "epoch": 2.580740740740741,
+      "grad_norm": 2.137296438217163,
+      "learning_rate": 2.801817822564711e-06,
+      "loss": 0.2679,
+      "step": 4354
+    },
+    {
+      "epoch": 2.5813333333333333,
+      "grad_norm": 2.236233949661255,
+      "learning_rate": 2.7978660343805577e-06,
+      "loss": 0.1826,
+      "step": 4355
+    },
+    {
+      "epoch": 2.581925925925926,
+      "grad_norm": 2.8810155391693115,
+      "learning_rate": 2.7939142461964043e-06,
+      "loss": 0.2217,
+      "step": 4356
+    },
+    {
+      "epoch": 2.5825185185185187,
+      "grad_norm": 2.5734057426452637,
+      "learning_rate": 2.789962458012251e-06,
+      "loss": 0.1523,
+      "step": 4357
+    },
+    {
+      "epoch": 2.583111111111111,
+      "grad_norm": 2.1182069778442383,
+      "learning_rate": 2.7860106698280977e-06,
+      "loss": 0.1806,
+      "step": 4358
+    },
+    {
+      "epoch": 2.5837037037037036,
+      "grad_norm": 2.2527225017547607,
+      "learning_rate": 2.7820588816439443e-06,
+      "loss": 0.2184,
+      "step": 4359
+    },
+    {
+      "epoch": 2.584296296296296,
+      "grad_norm": 3.1823692321777344,
+      "learning_rate": 2.778107093459791e-06,
+      "loss": 0.1941,
+      "step": 4360
+    },
+    {
+      "epoch": 2.584888888888889,
+      "grad_norm": 1.9925264120101929,
+      "learning_rate": 2.7741553052756377e-06,
+      "loss": 0.1298,
+      "step": 4361
+    },
+    {
+      "epoch": 2.5854814814814815,
+      "grad_norm": 2.25683856010437,
+      "learning_rate": 2.7702035170914843e-06,
+      "loss": 0.1461,
+      "step": 4362
+    },
+    {
+      "epoch": 2.586074074074074,
+      "grad_norm": 1.7445710897445679,
+      "learning_rate": 2.766251728907331e-06,
+      "loss": 0.0884,
+      "step": 4363
+    },
+    {
+      "epoch": 2.586666666666667,
+      "grad_norm": 1.7677922248840332,
+      "learning_rate": 2.7622999407231777e-06,
+      "loss": 0.1775,
+      "step": 4364
+    },
+    {
+      "epoch": 2.5872592592592594,
+      "grad_norm": 2.4415879249572754,
+      "learning_rate": 2.7583481525390243e-06,
+      "loss": 0.1728,
+      "step": 4365
+    },
+    {
+      "epoch": 2.587851851851852,
+      "grad_norm": 1.807874321937561,
+      "learning_rate": 2.754396364354871e-06,
+      "loss": 0.1085,
+      "step": 4366
+    },
+    {
+      "epoch": 2.5884444444444443,
+      "grad_norm": 2.2421629428863525,
+      "learning_rate": 2.7504445761707177e-06,
+      "loss": 0.187,
+      "step": 4367
+    },
+    {
+      "epoch": 2.589037037037037,
+      "grad_norm": 1.884169101715088,
+      "learning_rate": 2.7464927879865643e-06,
+      "loss": 0.1795,
+      "step": 4368
+    },
+    {
+      "epoch": 2.5896296296296297,
+      "grad_norm": 2.1121695041656494,
+      "learning_rate": 2.742540999802411e-06,
+      "loss": 0.1563,
+      "step": 4369
+    },
+    {
+      "epoch": 2.590222222222222,
+      "grad_norm": 1.5342961549758911,
+      "learning_rate": 2.7385892116182577e-06,
+      "loss": 0.1306,
+      "step": 4370
+    },
+    {
+      "epoch": 2.5908148148148147,
+      "grad_norm": 2.9431326389312744,
+      "learning_rate": 2.7346374234341043e-06,
+      "loss": 0.2311,
+      "step": 4371
+    },
+    {
+      "epoch": 2.5914074074074076,
+      "grad_norm": 2.169734239578247,
+      "learning_rate": 2.7306856352499506e-06,
+      "loss": 0.1742,
+      "step": 4372
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 1.9881715774536133,
+      "learning_rate": 2.7267338470657972e-06,
+      "loss": 0.1767,
+      "step": 4373
+    },
+    {
+      "epoch": 2.5925925925925926,
+      "grad_norm": 2.3565988540649414,
+      "learning_rate": 2.722782058881644e-06,
+      "loss": 0.1641,
+      "step": 4374
+    },
+    {
+      "epoch": 2.593185185185185,
+      "grad_norm": 1.6555629968643188,
+      "learning_rate": 2.7188302706974906e-06,
+      "loss": 0.1116,
+      "step": 4375
+    },
+    {
+      "epoch": 2.5937777777777775,
+      "grad_norm": 2.1352310180664062,
+      "learning_rate": 2.7148784825133372e-06,
+      "loss": 0.1642,
+      "step": 4376
+    },
+    {
+      "epoch": 2.5943703703703704,
+      "grad_norm": 2.3028602600097656,
+      "learning_rate": 2.710926694329184e-06,
+      "loss": 0.1828,
+      "step": 4377
+    },
+    {
+      "epoch": 2.594962962962963,
+      "grad_norm": 1.984410047531128,
+      "learning_rate": 2.7069749061450306e-06,
+      "loss": 0.1571,
+      "step": 4378
+    },
+    {
+      "epoch": 2.5955555555555554,
+      "grad_norm": 2.5795676708221436,
+      "learning_rate": 2.7030231179608772e-06,
+      "loss": 0.1373,
+      "step": 4379
+    },
+    {
+      "epoch": 2.5961481481481483,
+      "grad_norm": 2.767988920211792,
+      "learning_rate": 2.699071329776724e-06,
+      "loss": 0.2727,
+      "step": 4380
+    },
+    {
+      "epoch": 2.596740740740741,
+      "grad_norm": 1.8249950408935547,
+      "learning_rate": 2.6951195415925706e-06,
+      "loss": 0.1177,
+      "step": 4381
+    },
+    {
+      "epoch": 2.5973333333333333,
+      "grad_norm": 2.031402587890625,
+      "learning_rate": 2.6911677534084172e-06,
+      "loss": 0.1872,
+      "step": 4382
+    },
+    {
+      "epoch": 2.597925925925926,
+      "grad_norm": 1.9612226486206055,
+      "learning_rate": 2.687215965224264e-06,
+      "loss": 0.1677,
+      "step": 4383
+    },
+    {
+      "epoch": 2.5985185185185187,
+      "grad_norm": 2.4470157623291016,
+      "learning_rate": 2.6832641770401106e-06,
+      "loss": 0.1818,
+      "step": 4384
+    },
+    {
+      "epoch": 2.599111111111111,
+      "grad_norm": 2.19901704788208,
+      "learning_rate": 2.6793123888559577e-06,
+      "loss": 0.1439,
+      "step": 4385
+    },
+    {
+      "epoch": 2.5997037037037036,
+      "grad_norm": 1.5493342876434326,
+      "learning_rate": 2.6753606006718043e-06,
+      "loss": 0.0853,
+      "step": 4386
+    },
+    {
+      "epoch": 2.600296296296296,
+      "grad_norm": 1.6149637699127197,
+      "learning_rate": 2.671408812487651e-06,
+      "loss": 0.1662,
+      "step": 4387
+    },
+    {
+      "epoch": 2.600888888888889,
+      "grad_norm": 2.27258038520813,
+      "learning_rate": 2.6674570243034977e-06,
+      "loss": 0.1489,
+      "step": 4388
+    },
+    {
+      "epoch": 2.6014814814814815,
+      "grad_norm": 2.239976167678833,
+      "learning_rate": 2.6635052361193443e-06,
+      "loss": 0.18,
+      "step": 4389
+    },
+    {
+      "epoch": 2.602074074074074,
+      "grad_norm": 2.0881447792053223,
+      "learning_rate": 2.659553447935191e-06,
+      "loss": 0.1804,
+      "step": 4390
+    },
+    {
+      "epoch": 2.602666666666667,
+      "grad_norm": 2.861260175704956,
+      "learning_rate": 2.6556016597510377e-06,
+      "loss": 0.2463,
+      "step": 4391
+    },
+    {
+      "epoch": 2.6032592592592594,
+      "grad_norm": 1.9773190021514893,
+      "learning_rate": 2.6516498715668843e-06,
+      "loss": 0.2043,
+      "step": 4392
+    },
+    {
+      "epoch": 2.603851851851852,
+      "grad_norm": 2.3338072299957275,
+      "learning_rate": 2.647698083382731e-06,
+      "loss": 0.1821,
+      "step": 4393
+    },
+    {
+      "epoch": 2.6044444444444443,
+      "grad_norm": 1.7663716077804565,
+      "learning_rate": 2.6437462951985777e-06,
+      "loss": 0.1116,
+      "step": 4394
+    },
+    {
+      "epoch": 2.605037037037037,
+      "grad_norm": 2.585449457168579,
+      "learning_rate": 2.6397945070144243e-06,
+      "loss": 0.2575,
+      "step": 4395
+    },
+    {
+      "epoch": 2.6056296296296297,
+      "grad_norm": 2.1271591186523438,
+      "learning_rate": 2.635842718830271e-06,
+      "loss": 0.1734,
+      "step": 4396
+    },
+    {
+      "epoch": 2.606222222222222,
+      "grad_norm": 2.046389579772949,
+      "learning_rate": 2.6318909306461177e-06,
+      "loss": 0.1671,
+      "step": 4397
+    },
+    {
+      "epoch": 2.6068148148148147,
+      "grad_norm": 2.8735663890838623,
+      "learning_rate": 2.6279391424619643e-06,
+      "loss": 0.2918,
+      "step": 4398
+    },
+    {
+      "epoch": 2.6074074074074076,
+      "grad_norm": 1.5807490348815918,
+      "learning_rate": 2.623987354277811e-06,
+      "loss": 0.1272,
+      "step": 4399
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 1.7587268352508545,
+      "learning_rate": 2.6200355660936577e-06,
+      "loss": 0.147,
+      "step": 4400
+    },
+    {
+      "epoch": 2.6085925925925926,
+      "grad_norm": 2.479522705078125,
+      "learning_rate": 2.6160837779095043e-06,
+      "loss": 0.1651,
+      "step": 4401
+    },
+    {
+      "epoch": 2.609185185185185,
+      "grad_norm": 2.5728180408477783,
+      "learning_rate": 2.612131989725351e-06,
+      "loss": 0.1922,
+      "step": 4402
+    },
+    {
+      "epoch": 2.6097777777777775,
+      "grad_norm": 5.804673194885254,
+      "learning_rate": 2.6081802015411977e-06,
+      "loss": 0.4437,
+      "step": 4403
+    },
+    {
+      "epoch": 2.6103703703703705,
+      "grad_norm": 2.3335342407226562,
+      "learning_rate": 2.6042284133570443e-06,
+      "loss": 0.1751,
+      "step": 4404
+    },
+    {
+      "epoch": 2.610962962962963,
+      "grad_norm": 2.546496629714966,
+      "learning_rate": 2.600276625172891e-06,
+      "loss": 0.2225,
+      "step": 4405
+    },
+    {
+      "epoch": 2.6115555555555554,
+      "grad_norm": 2.782857656478882,
+      "learning_rate": 2.5963248369887377e-06,
+      "loss": 0.1477,
+      "step": 4406
+    },
+    {
+      "epoch": 2.6121481481481483,
+      "grad_norm": 1.9255659580230713,
+      "learning_rate": 2.5923730488045843e-06,
+      "loss": 0.1499,
+      "step": 4407
+    },
+    {
+      "epoch": 2.612740740740741,
+      "grad_norm": 3.303765296936035,
+      "learning_rate": 2.588421260620431e-06,
+      "loss": 0.2789,
+      "step": 4408
+    },
+    {
+      "epoch": 2.6133333333333333,
+      "grad_norm": 1.9909552335739136,
+      "learning_rate": 2.584469472436278e-06,
+      "loss": 0.1451,
+      "step": 4409
+    },
+    {
+      "epoch": 2.6139259259259258,
+      "grad_norm": 1.6894673109054565,
+      "learning_rate": 2.5805176842521247e-06,
+      "loss": 0.139,
+      "step": 4410
+    },
+    {
+      "epoch": 2.6145185185185182,
+      "grad_norm": 2.4024429321289062,
+      "learning_rate": 2.5765658960679706e-06,
+      "loss": 0.227,
+      "step": 4411
+    },
+    {
+      "epoch": 2.615111111111111,
+      "grad_norm": 2.2435741424560547,
+      "learning_rate": 2.5726141078838172e-06,
+      "loss": 0.118,
+      "step": 4412
+    },
+    {
+      "epoch": 2.6157037037037036,
+      "grad_norm": 2.261883020401001,
+      "learning_rate": 2.568662319699664e-06,
+      "loss": 0.1195,
+      "step": 4413
+    },
+    {
+      "epoch": 2.616296296296296,
+      "grad_norm": 2.8017239570617676,
+      "learning_rate": 2.5647105315155106e-06,
+      "loss": 0.2406,
+      "step": 4414
+    },
+    {
+      "epoch": 2.616888888888889,
+      "grad_norm": 2.057105302810669,
+      "learning_rate": 2.5607587433313572e-06,
+      "loss": 0.1625,
+      "step": 4415
+    },
+    {
+      "epoch": 2.6174814814814815,
+      "grad_norm": 2.3273842334747314,
+      "learning_rate": 2.5568069551472043e-06,
+      "loss": 0.1543,
+      "step": 4416
+    },
+    {
+      "epoch": 2.618074074074074,
+      "grad_norm": 1.7575011253356934,
+      "learning_rate": 2.552855166963051e-06,
+      "loss": 0.1094,
+      "step": 4417
+    },
+    {
+      "epoch": 2.618666666666667,
+      "grad_norm": 1.9503506422042847,
+      "learning_rate": 2.5489033787788977e-06,
+      "loss": 0.1121,
+      "step": 4418
+    },
+    {
+      "epoch": 2.6192592592592594,
+      "grad_norm": 2.3850314617156982,
+      "learning_rate": 2.5449515905947443e-06,
+      "loss": 0.1948,
+      "step": 4419
+    },
+    {
+      "epoch": 2.619851851851852,
+      "grad_norm": 2.0420467853546143,
+      "learning_rate": 2.540999802410591e-06,
+      "loss": 0.1755,
+      "step": 4420
+    },
+    {
+      "epoch": 2.6204444444444444,
+      "grad_norm": 2.2870841026306152,
+      "learning_rate": 2.5370480142264377e-06,
+      "loss": 0.1898,
+      "step": 4421
+    },
+    {
+      "epoch": 2.621037037037037,
+      "grad_norm": 1.8287529945373535,
+      "learning_rate": 2.5330962260422843e-06,
+      "loss": 0.0958,
+      "step": 4422
+    },
+    {
+      "epoch": 2.6216296296296298,
+      "grad_norm": 1.9196137189865112,
+      "learning_rate": 2.529144437858131e-06,
+      "loss": 0.1493,
+      "step": 4423
+    },
+    {
+      "epoch": 2.6222222222222222,
+      "grad_norm": 3.5164895057678223,
+      "learning_rate": 2.5251926496739777e-06,
+      "loss": 0.1761,
+      "step": 4424
+    },
+    {
+      "epoch": 2.6228148148148147,
+      "grad_norm": 2.648453712463379,
+      "learning_rate": 2.5212408614898243e-06,
+      "loss": 0.237,
+      "step": 4425
+    },
+    {
+      "epoch": 2.6234074074074076,
+      "grad_norm": 2.133957862854004,
+      "learning_rate": 2.517289073305671e-06,
+      "loss": 0.1079,
+      "step": 4426
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 2.1411309242248535,
+      "learning_rate": 2.5133372851215177e-06,
+      "loss": 0.1629,
+      "step": 4427
+    },
+    {
+      "epoch": 2.6245925925925926,
+      "grad_norm": 2.88657546043396,
+      "learning_rate": 2.5093854969373643e-06,
+      "loss": 0.2188,
+      "step": 4428
+    },
+    {
+      "epoch": 2.625185185185185,
+      "grad_norm": 2.3607327938079834,
+      "learning_rate": 2.505433708753211e-06,
+      "loss": 0.2421,
+      "step": 4429
+    },
+    {
+      "epoch": 2.6257777777777775,
+      "grad_norm": 2.5874359607696533,
+      "learning_rate": 2.5014819205690577e-06,
+      "loss": 0.1397,
+      "step": 4430
+    },
+    {
+      "epoch": 2.6263703703703705,
+      "grad_norm": 2.2484798431396484,
+      "learning_rate": 2.4975301323849043e-06,
+      "loss": 0.2793,
+      "step": 4431
+    },
+    {
+      "epoch": 2.626962962962963,
+      "grad_norm": 1.9048869609832764,
+      "learning_rate": 2.493578344200751e-06,
+      "loss": 0.1298,
+      "step": 4432
+    },
+    {
+      "epoch": 2.6275555555555554,
+      "grad_norm": 2.3358843326568604,
+      "learning_rate": 2.4896265560165977e-06,
+      "loss": 0.2059,
+      "step": 4433
+    },
+    {
+      "epoch": 2.6281481481481483,
+      "grad_norm": 1.7234694957733154,
+      "learning_rate": 2.4856747678324443e-06,
+      "loss": 0.1513,
+      "step": 4434
+    },
+    {
+      "epoch": 2.628740740740741,
+      "grad_norm": 2.380103588104248,
+      "learning_rate": 2.481722979648291e-06,
+      "loss": 0.1359,
+      "step": 4435
+    },
+    {
+      "epoch": 2.6293333333333333,
+      "grad_norm": 2.8172924518585205,
+      "learning_rate": 2.4777711914641376e-06,
+      "loss": 0.1737,
+      "step": 4436
+    },
+    {
+      "epoch": 2.629925925925926,
+      "grad_norm": 2.3499224185943604,
+      "learning_rate": 2.4738194032799843e-06,
+      "loss": 0.1701,
+      "step": 4437
+    },
+    {
+      "epoch": 2.6305185185185183,
+      "grad_norm": 1.4200701713562012,
+      "learning_rate": 2.469867615095831e-06,
+      "loss": 0.1034,
+      "step": 4438
+    },
+    {
+      "epoch": 2.631111111111111,
+      "grad_norm": 2.2259466648101807,
+      "learning_rate": 2.4659158269116776e-06,
+      "loss": 0.2179,
+      "step": 4439
+    },
+    {
+      "epoch": 2.6317037037037037,
+      "grad_norm": 1.7188929319381714,
+      "learning_rate": 2.4619640387275247e-06,
+      "loss": 0.142,
+      "step": 4440
+    },
+    {
+      "epoch": 2.632296296296296,
+      "grad_norm": 2.5917530059814453,
+      "learning_rate": 2.458012250543371e-06,
+      "loss": 0.2198,
+      "step": 4441
+    },
+    {
+      "epoch": 2.632888888888889,
+      "grad_norm": 2.8435721397399902,
+      "learning_rate": 2.4540604623592176e-06,
+      "loss": 0.1666,
+      "step": 4442
+    },
+    {
+      "epoch": 2.6334814814814815,
+      "grad_norm": 2.210184335708618,
+      "learning_rate": 2.4501086741750643e-06,
+      "loss": 0.187,
+      "step": 4443
+    },
+    {
+      "epoch": 2.634074074074074,
+      "grad_norm": 1.7209514379501343,
+      "learning_rate": 2.446156885990911e-06,
+      "loss": 0.1898,
+      "step": 4444
+    },
+    {
+      "epoch": 2.634666666666667,
+      "grad_norm": 3.0017340183258057,
+      "learning_rate": 2.4422050978067576e-06,
+      "loss": 0.2335,
+      "step": 4445
+    },
+    {
+      "epoch": 2.6352592592592594,
+      "grad_norm": 1.8342418670654297,
+      "learning_rate": 2.4382533096226043e-06,
+      "loss": 0.1469,
+      "step": 4446
+    },
+    {
+      "epoch": 2.635851851851852,
+      "grad_norm": 1.996302843093872,
+      "learning_rate": 2.434301521438451e-06,
+      "loss": 0.1884,
+      "step": 4447
+    },
+    {
+      "epoch": 2.6364444444444444,
+      "grad_norm": 1.327152132987976,
+      "learning_rate": 2.4303497332542976e-06,
+      "loss": 0.0976,
+      "step": 4448
+    },
+    {
+      "epoch": 2.637037037037037,
+      "grad_norm": 2.583371877670288,
+      "learning_rate": 2.4263979450701443e-06,
+      "loss": 0.1403,
+      "step": 4449
+    },
+    {
+      "epoch": 2.6376296296296298,
+      "grad_norm": 1.9933137893676758,
+      "learning_rate": 2.422446156885991e-06,
+      "loss": 0.1662,
+      "step": 4450
+    },
+    {
+      "epoch": 2.6382222222222222,
+      "grad_norm": 3.231527328491211,
+      "learning_rate": 2.4184943687018376e-06,
+      "loss": 0.1545,
+      "step": 4451
+    },
+    {
+      "epoch": 2.6388148148148147,
+      "grad_norm": 2.433105230331421,
+      "learning_rate": 2.4145425805176843e-06,
+      "loss": 0.108,
+      "step": 4452
+    },
+    {
+      "epoch": 2.6394074074074076,
+      "grad_norm": 2.8619019985198975,
+      "learning_rate": 2.410590792333531e-06,
+      "loss": 0.2947,
+      "step": 4453
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.367396354675293,
+      "learning_rate": 2.4066390041493776e-06,
+      "loss": 0.1429,
+      "step": 4454
+    },
+    {
+      "epoch": 2.6405925925925926,
+      "grad_norm": 2.9464821815490723,
+      "learning_rate": 2.4026872159652247e-06,
+      "loss": 0.1822,
+      "step": 4455
+    },
+    {
+      "epoch": 2.641185185185185,
+      "grad_norm": 3.5560429096221924,
+      "learning_rate": 2.3987354277810714e-06,
+      "loss": 0.2141,
+      "step": 4456
+    },
+    {
+      "epoch": 2.6417777777777776,
+      "grad_norm": 2.24246883392334,
+      "learning_rate": 2.394783639596918e-06,
+      "loss": 0.1858,
+      "step": 4457
+    },
+    {
+      "epoch": 2.6423703703703705,
+      "grad_norm": 3.281970977783203,
+      "learning_rate": 2.3908318514127647e-06,
+      "loss": 0.2047,
+      "step": 4458
+    },
+    {
+      "epoch": 2.642962962962963,
+      "grad_norm": 3.0465235710144043,
+      "learning_rate": 2.3868800632286114e-06,
+      "loss": 0.1669,
+      "step": 4459
+    },
+    {
+      "epoch": 2.6435555555555554,
+      "grad_norm": 2.2855560779571533,
+      "learning_rate": 2.382928275044458e-06,
+      "loss": 0.1626,
+      "step": 4460
+    },
+    {
+      "epoch": 2.6441481481481484,
+      "grad_norm": 2.3366198539733887,
+      "learning_rate": 2.3789764868603043e-06,
+      "loss": 0.1594,
+      "step": 4461
+    },
+    {
+      "epoch": 2.644740740740741,
+      "grad_norm": 1.7545344829559326,
+      "learning_rate": 2.375024698676151e-06,
+      "loss": 0.131,
+      "step": 4462
+    },
+    {
+      "epoch": 2.6453333333333333,
+      "grad_norm": 1.8465489149093628,
+      "learning_rate": 2.3710729104919976e-06,
+      "loss": 0.1651,
+      "step": 4463
+    },
+    {
+      "epoch": 2.645925925925926,
+      "grad_norm": 2.0527098178863525,
+      "learning_rate": 2.3671211223078443e-06,
+      "loss": 0.1894,
+      "step": 4464
+    },
+    {
+      "epoch": 2.6465185185185183,
+      "grad_norm": 1.7606942653656006,
+      "learning_rate": 2.363169334123691e-06,
+      "loss": 0.2036,
+      "step": 4465
+    },
+    {
+      "epoch": 2.647111111111111,
+      "grad_norm": 2.8894660472869873,
+      "learning_rate": 2.3592175459395376e-06,
+      "loss": 0.1864,
+      "step": 4466
+    },
+    {
+      "epoch": 2.6477037037037037,
+      "grad_norm": 1.6252614259719849,
+      "learning_rate": 2.3552657577553843e-06,
+      "loss": 0.1323,
+      "step": 4467
+    },
+    {
+      "epoch": 2.648296296296296,
+      "grad_norm": 1.6212077140808105,
+      "learning_rate": 2.351313969571231e-06,
+      "loss": 0.129,
+      "step": 4468
+    },
+    {
+      "epoch": 2.648888888888889,
+      "grad_norm": 2.3272945880889893,
+      "learning_rate": 2.3473621813870776e-06,
+      "loss": 0.1432,
+      "step": 4469
+    },
+    {
+      "epoch": 2.6494814814814815,
+      "grad_norm": 2.052215576171875,
+      "learning_rate": 2.3434103932029243e-06,
+      "loss": 0.2111,
+      "step": 4470
+    },
+    {
+      "epoch": 2.650074074074074,
+      "grad_norm": 2.538839817047119,
+      "learning_rate": 2.3394586050187714e-06,
+      "loss": 0.2083,
+      "step": 4471
+    },
+    {
+      "epoch": 2.6506666666666665,
+      "grad_norm": 1.654728651046753,
+      "learning_rate": 2.335506816834618e-06,
+      "loss": 0.1773,
+      "step": 4472
+    },
+    {
+      "epoch": 2.651259259259259,
+      "grad_norm": 2.778099298477173,
+      "learning_rate": 2.3315550286504647e-06,
+      "loss": 0.2461,
+      "step": 4473
+    },
+    {
+      "epoch": 2.651851851851852,
+      "grad_norm": 1.9213098287582397,
+      "learning_rate": 2.3276032404663114e-06,
+      "loss": 0.0942,
+      "step": 4474
+    },
+    {
+      "epoch": 2.6524444444444444,
+      "grad_norm": 2.0558502674102783,
+      "learning_rate": 2.323651452282158e-06,
+      "loss": 0.1279,
+      "step": 4475
+    },
+    {
+      "epoch": 2.653037037037037,
+      "grad_norm": 2.055678606033325,
+      "learning_rate": 2.3196996640980047e-06,
+      "loss": 0.1733,
+      "step": 4476
+    },
+    {
+      "epoch": 2.65362962962963,
+      "grad_norm": 2.470088005065918,
+      "learning_rate": 2.3157478759138514e-06,
+      "loss": 0.1933,
+      "step": 4477
+    },
+    {
+      "epoch": 2.6542222222222223,
+      "grad_norm": 2.281691312789917,
+      "learning_rate": 2.311796087729698e-06,
+      "loss": 0.1664,
+      "step": 4478
+    },
+    {
+      "epoch": 2.6548148148148147,
+      "grad_norm": 2.7810380458831787,
+      "learning_rate": 2.3078442995455447e-06,
+      "loss": 0.2558,
+      "step": 4479
+    },
+    {
+      "epoch": 2.6554074074074077,
+      "grad_norm": 2.7241861820220947,
+      "learning_rate": 2.303892511361391e-06,
+      "loss": 0.1884,
+      "step": 4480
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 1.8549232482910156,
+      "learning_rate": 2.2999407231772376e-06,
+      "loss": 0.1926,
+      "step": 4481
+    },
+    {
+      "epoch": 2.6565925925925926,
+      "grad_norm": 1.5532821416854858,
+      "learning_rate": 2.2959889349930843e-06,
+      "loss": 0.1435,
+      "step": 4482
+    },
+    {
+      "epoch": 2.657185185185185,
+      "grad_norm": 1.9252372980117798,
+      "learning_rate": 2.292037146808931e-06,
+      "loss": 0.1183,
+      "step": 4483
+    },
+    {
+      "epoch": 2.6577777777777776,
+      "grad_norm": 2.8249101638793945,
+      "learning_rate": 2.2880853586247776e-06,
+      "loss": 0.2751,
+      "step": 4484
+    },
+    {
+      "epoch": 2.6583703703703705,
+      "grad_norm": 2.22585391998291,
+      "learning_rate": 2.2841335704406243e-06,
+      "loss": 0.1804,
+      "step": 4485
+    },
+    {
+      "epoch": 2.658962962962963,
+      "grad_norm": 2.7826626300811768,
+      "learning_rate": 2.2801817822564714e-06,
+      "loss": 0.2237,
+      "step": 4486
+    },
+    {
+      "epoch": 2.6595555555555555,
+      "grad_norm": 2.248694896697998,
+      "learning_rate": 2.276229994072318e-06,
+      "loss": 0.2134,
+      "step": 4487
+    },
+    {
+      "epoch": 2.6601481481481484,
+      "grad_norm": 1.9601655006408691,
+      "learning_rate": 2.2722782058881647e-06,
+      "loss": 0.1268,
+      "step": 4488
+    },
+    {
+      "epoch": 2.660740740740741,
+      "grad_norm": 1.9437243938446045,
+      "learning_rate": 2.2683264177040114e-06,
+      "loss": 0.1234,
+      "step": 4489
+    },
+    {
+      "epoch": 2.6613333333333333,
+      "grad_norm": 2.881258487701416,
+      "learning_rate": 2.264374629519858e-06,
+      "loss": 0.174,
+      "step": 4490
+    },
+    {
+      "epoch": 2.661925925925926,
+      "grad_norm": 1.8774049282073975,
+      "learning_rate": 2.2604228413357047e-06,
+      "loss": 0.1228,
+      "step": 4491
+    },
+    {
+      "epoch": 2.6625185185185183,
+      "grad_norm": 1.927869439125061,
+      "learning_rate": 2.2564710531515514e-06,
+      "loss": 0.1681,
+      "step": 4492
+    },
+    {
+      "epoch": 2.663111111111111,
+      "grad_norm": 1.6209850311279297,
+      "learning_rate": 2.252519264967398e-06,
+      "loss": 0.1624,
+      "step": 4493
+    },
+    {
+      "epoch": 2.6637037037037037,
+      "grad_norm": 2.110100030899048,
+      "learning_rate": 2.2485674767832447e-06,
+      "loss": 0.2803,
+      "step": 4494
+    },
+    {
+      "epoch": 2.664296296296296,
+      "grad_norm": 2.322153329849243,
+      "learning_rate": 2.2446156885990914e-06,
+      "loss": 0.2352,
+      "step": 4495
+    },
+    {
+      "epoch": 2.664888888888889,
+      "grad_norm": 1.8997403383255005,
+      "learning_rate": 2.240663900414938e-06,
+      "loss": 0.1625,
+      "step": 4496
+    },
+    {
+      "epoch": 2.6654814814814816,
+      "grad_norm": 1.7332967519760132,
+      "learning_rate": 2.2367121122307847e-06,
+      "loss": 0.1276,
+      "step": 4497
+    },
+    {
+      "epoch": 2.666074074074074,
+      "grad_norm": 2.0136148929595947,
+      "learning_rate": 2.2327603240466314e-06,
+      "loss": 0.1454,
+      "step": 4498
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 2.1241588592529297,
+      "learning_rate": 2.228808535862478e-06,
+      "loss": 0.1951,
+      "step": 4499
+    },
+    {
+      "epoch": 2.667259259259259,
+      "grad_norm": 1.7834563255310059,
+      "learning_rate": 2.2248567476783243e-06,
+      "loss": 0.1739,
+      "step": 4500
+    },
+    {
+      "epoch": 2.667851851851852,
+      "grad_norm": 2.6499037742614746,
+      "learning_rate": 2.2209049594941714e-06,
+      "loss": 0.2242,
+      "step": 4501
+    },
+    {
+      "epoch": 2.6684444444444444,
+      "grad_norm": 2.9972152709960938,
+      "learning_rate": 2.216953171310018e-06,
+      "loss": 0.2771,
+      "step": 4502
+    },
+    {
+      "epoch": 2.669037037037037,
+      "grad_norm": 1.6221729516983032,
+      "learning_rate": 2.2130013831258647e-06,
+      "loss": 0.1089,
+      "step": 4503
+    },
+    {
+      "epoch": 2.66962962962963,
+      "grad_norm": 2.0192480087280273,
+      "learning_rate": 2.2090495949417114e-06,
+      "loss": 0.1774,
+      "step": 4504
+    },
+    {
+      "epoch": 2.6702222222222223,
+      "grad_norm": 2.436612606048584,
+      "learning_rate": 2.205097806757558e-06,
+      "loss": 0.1686,
+      "step": 4505
+    },
+    {
+      "epoch": 2.6708148148148148,
+      "grad_norm": 1.785163402557373,
+      "learning_rate": 2.2011460185734047e-06,
+      "loss": 0.1495,
+      "step": 4506
+    },
+    {
+      "epoch": 2.6714074074074077,
+      "grad_norm": 2.4975168704986572,
+      "learning_rate": 2.1971942303892514e-06,
+      "loss": 0.1449,
+      "step": 4507
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 4.368551254272461,
+      "learning_rate": 2.193242442205098e-06,
+      "loss": 0.277,
+      "step": 4508
+    },
+    {
+      "epoch": 2.6725925925925926,
+      "grad_norm": 2.5058109760284424,
+      "learning_rate": 2.1892906540209447e-06,
+      "loss": 0.1534,
+      "step": 4509
+    },
+    {
+      "epoch": 2.673185185185185,
+      "grad_norm": 1.8006670475006104,
+      "learning_rate": 2.1853388658367914e-06,
+      "loss": 0.1254,
+      "step": 4510
+    },
+    {
+      "epoch": 2.6737777777777776,
+      "grad_norm": 2.335174798965454,
+      "learning_rate": 2.181387077652638e-06,
+      "loss": 0.2617,
+      "step": 4511
+    },
+    {
+      "epoch": 2.6743703703703705,
+      "grad_norm": 2.832084894180298,
+      "learning_rate": 2.1774352894684847e-06,
+      "loss": 0.275,
+      "step": 4512
+    },
+    {
+      "epoch": 2.674962962962963,
+      "grad_norm": 2.021031379699707,
+      "learning_rate": 2.1734835012843314e-06,
+      "loss": 0.15,
+      "step": 4513
+    },
+    {
+      "epoch": 2.6755555555555555,
+      "grad_norm": 2.844428539276123,
+      "learning_rate": 2.169531713100178e-06,
+      "loss": 0.16,
+      "step": 4514
+    },
+    {
+      "epoch": 2.6761481481481484,
+      "grad_norm": 2.233086109161377,
+      "learning_rate": 2.1655799249160247e-06,
+      "loss": 0.2097,
+      "step": 4515
+    },
+    {
+      "epoch": 2.676740740740741,
+      "grad_norm": 1.9256126880645752,
+      "learning_rate": 2.1616281367318714e-06,
+      "loss": 0.173,
+      "step": 4516
+    },
+    {
+      "epoch": 2.6773333333333333,
+      "grad_norm": 1.9357085227966309,
+      "learning_rate": 2.157676348547718e-06,
+      "loss": 0.1434,
+      "step": 4517
+    },
+    {
+      "epoch": 2.677925925925926,
+      "grad_norm": 1.7642444372177124,
+      "learning_rate": 2.1537245603635647e-06,
+      "loss": 0.14,
+      "step": 4518
+    },
+    {
+      "epoch": 2.6785185185185183,
+      "grad_norm": 2.5748186111450195,
+      "learning_rate": 2.1497727721794114e-06,
+      "loss": 0.261,
+      "step": 4519
+    },
+    {
+      "epoch": 2.679111111111111,
+      "grad_norm": 2.0427939891815186,
+      "learning_rate": 2.145820983995258e-06,
+      "loss": 0.1444,
+      "step": 4520
+    },
+    {
+      "epoch": 2.6797037037037037,
+      "grad_norm": 2.078808069229126,
+      "learning_rate": 2.1418691958111047e-06,
+      "loss": 0.1524,
+      "step": 4521
+    },
+    {
+      "epoch": 2.680296296296296,
+      "grad_norm": 2.2926816940307617,
+      "learning_rate": 2.1379174076269514e-06,
+      "loss": 0.207,
+      "step": 4522
+    },
+    {
+      "epoch": 2.680888888888889,
+      "grad_norm": 3.224795341491699,
+      "learning_rate": 2.133965619442798e-06,
+      "loss": 0.1982,
+      "step": 4523
+    },
+    {
+      "epoch": 2.6814814814814816,
+      "grad_norm": 1.7852598428726196,
+      "learning_rate": 2.1300138312586447e-06,
+      "loss": 0.1192,
+      "step": 4524
+    },
+    {
+      "epoch": 2.682074074074074,
+      "grad_norm": 2.348841905593872,
+      "learning_rate": 2.1260620430744914e-06,
+      "loss": 0.1991,
+      "step": 4525
+    },
+    {
+      "epoch": 2.6826666666666665,
+      "grad_norm": 2.353821039199829,
+      "learning_rate": 2.122110254890338e-06,
+      "loss": 0.1731,
+      "step": 4526
+    },
+    {
+      "epoch": 2.683259259259259,
+      "grad_norm": 2.322511672973633,
+      "learning_rate": 2.1181584667061847e-06,
+      "loss": 0.1619,
+      "step": 4527
+    },
+    {
+      "epoch": 2.683851851851852,
+      "grad_norm": 2.3092873096466064,
+      "learning_rate": 2.1142066785220314e-06,
+      "loss": 0.1626,
+      "step": 4528
+    },
+    {
+      "epoch": 2.6844444444444444,
+      "grad_norm": 2.9346890449523926,
+      "learning_rate": 2.110254890337878e-06,
+      "loss": 0.1775,
+      "step": 4529
+    },
+    {
+      "epoch": 2.685037037037037,
+      "grad_norm": 3.1220085620880127,
+      "learning_rate": 2.1063031021537247e-06,
+      "loss": 0.2208,
+      "step": 4530
+    },
+    {
+      "epoch": 2.68562962962963,
+      "grad_norm": 1.943772315979004,
+      "learning_rate": 2.1023513139695714e-06,
+      "loss": 0.1117,
+      "step": 4531
+    },
+    {
+      "epoch": 2.6862222222222223,
+      "grad_norm": 1.6514003276824951,
+      "learning_rate": 2.098399525785418e-06,
+      "loss": 0.1991,
+      "step": 4532
+    },
+    {
+      "epoch": 2.6868148148148148,
+      "grad_norm": 3.039912700653076,
+      "learning_rate": 2.0944477376012647e-06,
+      "loss": 0.2508,
+      "step": 4533
+    },
+    {
+      "epoch": 2.6874074074074072,
+      "grad_norm": 2.010874032974243,
+      "learning_rate": 2.0904959494171114e-06,
+      "loss": 0.1306,
+      "step": 4534
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 1.9898673295974731,
+      "learning_rate": 2.086544161232958e-06,
+      "loss": 0.0955,
+      "step": 4535
+    },
+    {
+      "epoch": 2.6885925925925926,
+      "grad_norm": 1.7316426038742065,
+      "learning_rate": 2.0825923730488047e-06,
+      "loss": 0.1479,
+      "step": 4536
+    },
+    {
+      "epoch": 2.689185185185185,
+      "grad_norm": 3.156449794769287,
+      "learning_rate": 2.0786405848646514e-06,
+      "loss": 0.2064,
+      "step": 4537
+    },
+    {
+      "epoch": 2.6897777777777776,
+      "grad_norm": 2.293269634246826,
+      "learning_rate": 2.074688796680498e-06,
+      "loss": 0.1515,
+      "step": 4538
+    },
+    {
+      "epoch": 2.6903703703703705,
+      "grad_norm": 2.6615850925445557,
+      "learning_rate": 2.0707370084963447e-06,
+      "loss": 0.1825,
+      "step": 4539
+    },
+    {
+      "epoch": 2.690962962962963,
+      "grad_norm": 2.140716075897217,
+      "learning_rate": 2.0667852203121914e-06,
+      "loss": 0.0964,
+      "step": 4540
+    },
+    {
+      "epoch": 2.6915555555555555,
+      "grad_norm": 1.7442823648452759,
+      "learning_rate": 2.062833432128038e-06,
+      "loss": 0.1444,
+      "step": 4541
+    },
+    {
+      "epoch": 2.6921481481481484,
+      "grad_norm": 1.9684875011444092,
+      "learning_rate": 2.0588816439438847e-06,
+      "loss": 0.1719,
+      "step": 4542
+    },
+    {
+      "epoch": 2.692740740740741,
+      "grad_norm": 2.5807888507843018,
+      "learning_rate": 2.0549298557597314e-06,
+      "loss": 0.2463,
+      "step": 4543
+    },
+    {
+      "epoch": 2.6933333333333334,
+      "grad_norm": 2.6398022174835205,
+      "learning_rate": 2.050978067575578e-06,
+      "loss": 0.126,
+      "step": 4544
+    },
+    {
+      "epoch": 2.693925925925926,
+      "grad_norm": 2.2126073837280273,
+      "learning_rate": 2.0470262793914247e-06,
+      "loss": 0.1683,
+      "step": 4545
+    },
+    {
+      "epoch": 2.6945185185185183,
+      "grad_norm": 2.5030357837677,
+      "learning_rate": 2.0430744912072714e-06,
+      "loss": 0.2001,
+      "step": 4546
+    },
+    {
+      "epoch": 2.6951111111111112,
+      "grad_norm": 2.090963363647461,
+      "learning_rate": 2.039122703023118e-06,
+      "loss": 0.1169,
+      "step": 4547
+    },
+    {
+      "epoch": 2.6957037037037037,
+      "grad_norm": 1.4238061904907227,
+      "learning_rate": 2.0351709148389647e-06,
+      "loss": 0.1519,
+      "step": 4548
+    },
+    {
+      "epoch": 2.696296296296296,
+      "grad_norm": 2.0755867958068848,
+      "learning_rate": 2.0312191266548114e-06,
+      "loss": 0.1837,
+      "step": 4549
+    },
+    {
+      "epoch": 2.696888888888889,
+      "grad_norm": 2.221808671951294,
+      "learning_rate": 2.027267338470658e-06,
+      "loss": 0.1531,
+      "step": 4550
+    },
+    {
+      "epoch": 2.6974814814814816,
+      "grad_norm": 2.492553472518921,
+      "learning_rate": 2.0233155502865047e-06,
+      "loss": 0.132,
+      "step": 4551
+    },
+    {
+      "epoch": 2.698074074074074,
+      "grad_norm": 2.436112642288208,
+      "learning_rate": 2.0193637621023514e-06,
+      "loss": 0.1968,
+      "step": 4552
+    },
+    {
+      "epoch": 2.6986666666666665,
+      "grad_norm": 1.824535846710205,
+      "learning_rate": 2.015411973918198e-06,
+      "loss": 0.1406,
+      "step": 4553
+    },
+    {
+      "epoch": 2.699259259259259,
+      "grad_norm": 1.855000376701355,
+      "learning_rate": 2.0114601857340447e-06,
+      "loss": 0.1322,
+      "step": 4554
+    },
+    {
+      "epoch": 2.699851851851852,
+      "grad_norm": 2.465022325515747,
+      "learning_rate": 2.0075083975498914e-06,
+      "loss": 0.1726,
+      "step": 4555
+    },
+    {
+      "epoch": 2.7004444444444444,
+      "grad_norm": 2.182109832763672,
+      "learning_rate": 2.0035566093657385e-06,
+      "loss": 0.1885,
+      "step": 4556
+    },
+    {
+      "epoch": 2.701037037037037,
+      "grad_norm": 1.7929614782333374,
+      "learning_rate": 1.999604821181585e-06,
+      "loss": 0.1345,
+      "step": 4557
+    },
+    {
+      "epoch": 2.70162962962963,
+      "grad_norm": 2.2678236961364746,
+      "learning_rate": 1.995653032997432e-06,
+      "loss": 0.3045,
+      "step": 4558
+    },
+    {
+      "epoch": 2.7022222222222223,
+      "grad_norm": 2.2354745864868164,
+      "learning_rate": 1.991701244813278e-06,
+      "loss": 0.1515,
+      "step": 4559
+    },
+    {
+      "epoch": 2.702814814814815,
+      "grad_norm": 1.820374846458435,
+      "learning_rate": 1.9877494566291247e-06,
+      "loss": 0.1053,
+      "step": 4560
+    },
+    {
+      "epoch": 2.7034074074074073,
+      "grad_norm": 1.463283896446228,
+      "learning_rate": 1.9837976684449714e-06,
+      "loss": 0.1544,
+      "step": 4561
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 2.56734299659729,
+      "learning_rate": 1.979845880260818e-06,
+      "loss": 0.1214,
+      "step": 4562
+    },
+    {
+      "epoch": 2.7045925925925927,
+      "grad_norm": 2.402292251586914,
+      "learning_rate": 1.9758940920766647e-06,
+      "loss": 0.2182,
+      "step": 4563
+    },
+    {
+      "epoch": 2.705185185185185,
+      "grad_norm": 1.8067166805267334,
+      "learning_rate": 1.9719423038925114e-06,
+      "loss": 0.1762,
+      "step": 4564
+    },
+    {
+      "epoch": 2.7057777777777776,
+      "grad_norm": 2.5517401695251465,
+      "learning_rate": 1.967990515708358e-06,
+      "loss": 0.1886,
+      "step": 4565
+    },
+    {
+      "epoch": 2.7063703703703705,
+      "grad_norm": 2.0550601482391357,
+      "learning_rate": 1.9640387275242047e-06,
+      "loss": 0.194,
+      "step": 4566
+    },
+    {
+      "epoch": 2.706962962962963,
+      "grad_norm": 2.770677328109741,
+      "learning_rate": 1.9600869393400514e-06,
+      "loss": 0.1975,
+      "step": 4567
+    },
+    {
+      "epoch": 2.7075555555555555,
+      "grad_norm": 2.4669711589813232,
+      "learning_rate": 1.956135151155898e-06,
+      "loss": 0.1831,
+      "step": 4568
+    },
+    {
+      "epoch": 2.7081481481481484,
+      "grad_norm": 2.4337127208709717,
+      "learning_rate": 1.9521833629717447e-06,
+      "loss": 0.1587,
+      "step": 4569
+    },
+    {
+      "epoch": 2.708740740740741,
+      "grad_norm": 1.9201147556304932,
+      "learning_rate": 1.9482315747875914e-06,
+      "loss": 0.1657,
+      "step": 4570
+    },
+    {
+      "epoch": 2.7093333333333334,
+      "grad_norm": 2.0961453914642334,
+      "learning_rate": 1.9442797866034385e-06,
+      "loss": 0.1805,
+      "step": 4571
+    },
+    {
+      "epoch": 2.709925925925926,
+      "grad_norm": 1.9056274890899658,
+      "learning_rate": 1.940327998419285e-06,
+      "loss": 0.1524,
+      "step": 4572
+    },
+    {
+      "epoch": 2.7105185185185183,
+      "grad_norm": 2.4778501987457275,
+      "learning_rate": 1.936376210235132e-06,
+      "loss": 0.1972,
+      "step": 4573
+    },
+    {
+      "epoch": 2.7111111111111112,
+      "grad_norm": 2.6449155807495117,
+      "learning_rate": 1.9324244220509785e-06,
+      "loss": 0.1866,
+      "step": 4574
+    },
+    {
+      "epoch": 2.7117037037037037,
+      "grad_norm": 2.327347993850708,
+      "learning_rate": 1.928472633866825e-06,
+      "loss": 0.2037,
+      "step": 4575
+    },
+    {
+      "epoch": 2.712296296296296,
+      "grad_norm": 1.75370192527771,
+      "learning_rate": 1.9245208456826718e-06,
+      "loss": 0.1399,
+      "step": 4576
+    },
+    {
+      "epoch": 2.712888888888889,
+      "grad_norm": 4.325387954711914,
+      "learning_rate": 1.9205690574985185e-06,
+      "loss": 0.2888,
+      "step": 4577
+    },
+    {
+      "epoch": 2.7134814814814816,
+      "grad_norm": 3.182898759841919,
+      "learning_rate": 1.916617269314365e-06,
+      "loss": 0.1473,
+      "step": 4578
+    },
+    {
+      "epoch": 2.714074074074074,
+      "grad_norm": 1.8220463991165161,
+      "learning_rate": 1.9126654811302114e-06,
+      "loss": 0.1573,
+      "step": 4579
+    },
+    {
+      "epoch": 2.7146666666666666,
+      "grad_norm": 3.1579673290252686,
+      "learning_rate": 1.908713692946058e-06,
+      "loss": 0.3047,
+      "step": 4580
+    },
+    {
+      "epoch": 2.715259259259259,
+      "grad_norm": 2.4011616706848145,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 0.1708,
+      "step": 4581
+    },
+    {
+      "epoch": 2.715851851851852,
+      "grad_norm": 2.283018112182617,
+      "learning_rate": 1.9008101165777516e-06,
+      "loss": 0.1395,
+      "step": 4582
+    },
+    {
+      "epoch": 2.7164444444444444,
+      "grad_norm": 1.8487507104873657,
+      "learning_rate": 1.8968583283935982e-06,
+      "loss": 0.1189,
+      "step": 4583
+    },
+    {
+      "epoch": 2.717037037037037,
+      "grad_norm": 2.0050628185272217,
+      "learning_rate": 1.892906540209445e-06,
+      "loss": 0.1775,
+      "step": 4584
+    },
+    {
+      "epoch": 2.71762962962963,
+      "grad_norm": 1.2560707330703735,
+      "learning_rate": 1.8889547520252916e-06,
+      "loss": 0.0718,
+      "step": 4585
+    },
+    {
+      "epoch": 2.7182222222222223,
+      "grad_norm": 1.9681180715560913,
+      "learning_rate": 1.8850029638411382e-06,
+      "loss": 0.2111,
+      "step": 4586
+    },
+    {
+      "epoch": 2.718814814814815,
+      "grad_norm": 2.6213908195495605,
+      "learning_rate": 1.881051175656985e-06,
+      "loss": 0.2152,
+      "step": 4587
+    },
+    {
+      "epoch": 2.7194074074074073,
+      "grad_norm": 2.979566812515259,
+      "learning_rate": 1.8770993874728316e-06,
+      "loss": 0.1362,
+      "step": 4588
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 3.044621706008911,
+      "learning_rate": 1.8731475992886782e-06,
+      "loss": 0.1841,
+      "step": 4589
+    },
+    {
+      "epoch": 2.7205925925925927,
+      "grad_norm": 2.491658926010132,
+      "learning_rate": 1.869195811104525e-06,
+      "loss": 0.1355,
+      "step": 4590
+    },
+    {
+      "epoch": 2.721185185185185,
+      "grad_norm": 2.585744857788086,
+      "learning_rate": 1.8652440229203716e-06,
+      "loss": 0.1732,
+      "step": 4591
+    },
+    {
+      "epoch": 2.7217777777777776,
+      "grad_norm": 2.6205077171325684,
+      "learning_rate": 1.8612922347362182e-06,
+      "loss": 0.2524,
+      "step": 4592
+    },
+    {
+      "epoch": 2.7223703703703706,
+      "grad_norm": 2.3224213123321533,
+      "learning_rate": 1.8573404465520651e-06,
+      "loss": 0.1343,
+      "step": 4593
+    },
+    {
+      "epoch": 2.722962962962963,
+      "grad_norm": 1.9678319692611694,
+      "learning_rate": 1.8533886583679118e-06,
+      "loss": 0.1933,
+      "step": 4594
+    },
+    {
+      "epoch": 2.7235555555555555,
+      "grad_norm": 1.5859960317611694,
+      "learning_rate": 1.8494368701837584e-06,
+      "loss": 0.1338,
+      "step": 4595
+    },
+    {
+      "epoch": 2.724148148148148,
+      "grad_norm": 1.4946932792663574,
+      "learning_rate": 1.8454850819996051e-06,
+      "loss": 0.1329,
+      "step": 4596
+    },
+    {
+      "epoch": 2.7247407407407405,
+      "grad_norm": 2.0409390926361084,
+      "learning_rate": 1.8415332938154518e-06,
+      "loss": 0.2164,
+      "step": 4597
+    },
+    {
+      "epoch": 2.7253333333333334,
+      "grad_norm": 2.1452596187591553,
+      "learning_rate": 1.8375815056312984e-06,
+      "loss": 0.1285,
+      "step": 4598
+    },
+    {
+      "epoch": 2.725925925925926,
+      "grad_norm": 1.3605915307998657,
+      "learning_rate": 1.833629717447145e-06,
+      "loss": 0.1086,
+      "step": 4599
+    },
+    {
+      "epoch": 2.7265185185185183,
+      "grad_norm": 2.1272053718566895,
+      "learning_rate": 1.8296779292629916e-06,
+      "loss": 0.2287,
+      "step": 4600
+    },
+    {
+      "epoch": 2.7271111111111113,
+      "grad_norm": 2.2933640480041504,
+      "learning_rate": 1.8257261410788382e-06,
+      "loss": 0.18,
+      "step": 4601
+    },
+    {
+      "epoch": 2.7277037037037037,
+      "grad_norm": 3.230990171432495,
+      "learning_rate": 1.821774352894685e-06,
+      "loss": 0.2808,
+      "step": 4602
+    },
+    {
+      "epoch": 2.728296296296296,
+      "grad_norm": 2.060382843017578,
+      "learning_rate": 1.8178225647105316e-06,
+      "loss": 0.1811,
+      "step": 4603
+    },
+    {
+      "epoch": 2.728888888888889,
+      "grad_norm": 2.1582517623901367,
+      "learning_rate": 1.8138707765263782e-06,
+      "loss": 0.1234,
+      "step": 4604
+    },
+    {
+      "epoch": 2.7294814814814816,
+      "grad_norm": 1.7739160060882568,
+      "learning_rate": 1.809918988342225e-06,
+      "loss": 0.1326,
+      "step": 4605
+    },
+    {
+      "epoch": 2.730074074074074,
+      "grad_norm": 1.9581027030944824,
+      "learning_rate": 1.8059672001580716e-06,
+      "loss": 0.2201,
+      "step": 4606
+    },
+    {
+      "epoch": 2.7306666666666666,
+      "grad_norm": 2.444216728210449,
+      "learning_rate": 1.8020154119739182e-06,
+      "loss": 0.1565,
+      "step": 4607
+    },
+    {
+      "epoch": 2.731259259259259,
+      "grad_norm": 1.5956839323043823,
+      "learning_rate": 1.7980636237897651e-06,
+      "loss": 0.1401,
+      "step": 4608
+    },
+    {
+      "epoch": 2.731851851851852,
+      "grad_norm": 2.4557571411132812,
+      "learning_rate": 1.7941118356056118e-06,
+      "loss": 0.1797,
+      "step": 4609
+    },
+    {
+      "epoch": 2.7324444444444445,
+      "grad_norm": 2.211939573287964,
+      "learning_rate": 1.7901600474214584e-06,
+      "loss": 0.2359,
+      "step": 4610
+    },
+    {
+      "epoch": 2.733037037037037,
+      "grad_norm": 1.828330159187317,
+      "learning_rate": 1.7862082592373051e-06,
+      "loss": 0.2309,
+      "step": 4611
+    },
+    {
+      "epoch": 2.73362962962963,
+      "grad_norm": 2.048224449157715,
+      "learning_rate": 1.7822564710531518e-06,
+      "loss": 0.151,
+      "step": 4612
+    },
+    {
+      "epoch": 2.7342222222222223,
+      "grad_norm": 2.296901226043701,
+      "learning_rate": 1.7783046828689984e-06,
+      "loss": 0.1828,
+      "step": 4613
+    },
+    {
+      "epoch": 2.734814814814815,
+      "grad_norm": 1.7712475061416626,
+      "learning_rate": 1.7743528946848451e-06,
+      "loss": 0.1772,
+      "step": 4614
+    },
+    {
+      "epoch": 2.7354074074074073,
+      "grad_norm": 2.871703624725342,
+      "learning_rate": 1.7704011065006918e-06,
+      "loss": 0.1868,
+      "step": 4615
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 1.9653100967407227,
+      "learning_rate": 1.7664493183165384e-06,
+      "loss": 0.113,
+      "step": 4616
+    },
+    {
+      "epoch": 2.7365925925925927,
+      "grad_norm": 2.105087995529175,
+      "learning_rate": 1.7624975301323851e-06,
+      "loss": 0.1873,
+      "step": 4617
+    },
+    {
+      "epoch": 2.737185185185185,
+      "grad_norm": 2.168862819671631,
+      "learning_rate": 1.7585457419482318e-06,
+      "loss": 0.242,
+      "step": 4618
+    },
+    {
+      "epoch": 2.7377777777777776,
+      "grad_norm": 2.781831741333008,
+      "learning_rate": 1.7545939537640782e-06,
+      "loss": 0.2834,
+      "step": 4619
+    },
+    {
+      "epoch": 2.7383703703703706,
+      "grad_norm": 2.1623806953430176,
+      "learning_rate": 1.750642165579925e-06,
+      "loss": 0.1963,
+      "step": 4620
+    },
+    {
+      "epoch": 2.738962962962963,
+      "grad_norm": 3.024712085723877,
+      "learning_rate": 1.7466903773957716e-06,
+      "loss": 0.1525,
+      "step": 4621
+    },
+    {
+      "epoch": 2.7395555555555555,
+      "grad_norm": 1.728251338005066,
+      "learning_rate": 1.7427385892116182e-06,
+      "loss": 0.1595,
+      "step": 4622
+    },
+    {
+      "epoch": 2.740148148148148,
+      "grad_norm": 2.8717832565307617,
+      "learning_rate": 1.738786801027465e-06,
+      "loss": 0.206,
+      "step": 4623
+    },
+    {
+      "epoch": 2.7407407407407405,
+      "grad_norm": 1.8412084579467773,
+      "learning_rate": 1.7348350128433118e-06,
+      "loss": 0.1457,
+      "step": 4624
+    },
+    {
+      "epoch": 2.7413333333333334,
+      "grad_norm": 1.783788800239563,
+      "learning_rate": 1.7308832246591584e-06,
+      "loss": 0.1568,
+      "step": 4625
+    },
+    {
+      "epoch": 2.741925925925926,
+      "grad_norm": 2.451167106628418,
+      "learning_rate": 1.726931436475005e-06,
+      "loss": 0.1832,
+      "step": 4626
+    },
+    {
+      "epoch": 2.7425185185185184,
+      "grad_norm": 2.0369420051574707,
+      "learning_rate": 1.7229796482908518e-06,
+      "loss": 0.1351,
+      "step": 4627
+    },
+    {
+      "epoch": 2.7431111111111113,
+      "grad_norm": 1.637770652770996,
+      "learning_rate": 1.7190278601066984e-06,
+      "loss": 0.1547,
+      "step": 4628
+    },
+    {
+      "epoch": 2.7437037037037038,
+      "grad_norm": 2.4029767513275146,
+      "learning_rate": 1.715076071922545e-06,
+      "loss": 0.1776,
+      "step": 4629
+    },
+    {
+      "epoch": 2.7442962962962962,
+      "grad_norm": 2.6292026042938232,
+      "learning_rate": 1.7111242837383918e-06,
+      "loss": 0.3194,
+      "step": 4630
+    },
+    {
+      "epoch": 2.744888888888889,
+      "grad_norm": 2.282381057739258,
+      "learning_rate": 1.7071724955542384e-06,
+      "loss": 0.2249,
+      "step": 4631
+    },
+    {
+      "epoch": 2.7454814814814816,
+      "grad_norm": 2.4644484519958496,
+      "learning_rate": 1.703220707370085e-06,
+      "loss": 0.1434,
+      "step": 4632
+    },
+    {
+      "epoch": 2.746074074074074,
+      "grad_norm": 1.557657241821289,
+      "learning_rate": 1.6992689191859318e-06,
+      "loss": 0.1902,
+      "step": 4633
+    },
+    {
+      "epoch": 2.7466666666666666,
+      "grad_norm": 1.2580288648605347,
+      "learning_rate": 1.6953171310017784e-06,
+      "loss": 0.0658,
+      "step": 4634
+    },
+    {
+      "epoch": 2.747259259259259,
+      "grad_norm": 3.51874041557312,
+      "learning_rate": 1.6913653428176253e-06,
+      "loss": 0.2283,
+      "step": 4635
+    },
+    {
+      "epoch": 2.747851851851852,
+      "grad_norm": 1.6314300298690796,
+      "learning_rate": 1.687413554633472e-06,
+      "loss": 0.1401,
+      "step": 4636
+    },
+    {
+      "epoch": 2.7484444444444445,
+      "grad_norm": 1.9490931034088135,
+      "learning_rate": 1.6834617664493186e-06,
+      "loss": 0.159,
+      "step": 4637
+    },
+    {
+      "epoch": 2.749037037037037,
+      "grad_norm": 2.2891125679016113,
+      "learning_rate": 1.6795099782651649e-06,
+      "loss": 0.1446,
+      "step": 4638
+    },
+    {
+      "epoch": 2.74962962962963,
+      "grad_norm": 2.395282745361328,
+      "learning_rate": 1.6755581900810118e-06,
+      "loss": 0.201,
+      "step": 4639
+    },
+    {
+      "epoch": 2.7502222222222223,
+      "grad_norm": 1.5275628566741943,
+      "learning_rate": 1.6716064018968584e-06,
+      "loss": 0.1157,
+      "step": 4640
+    },
+    {
+      "epoch": 2.750814814814815,
+      "grad_norm": 2.12072491645813,
+      "learning_rate": 1.667654613712705e-06,
+      "loss": 0.1478,
+      "step": 4641
+    },
+    {
+      "epoch": 2.7514074074074073,
+      "grad_norm": 1.8971222639083862,
+      "learning_rate": 1.6637028255285518e-06,
+      "loss": 0.1986,
+      "step": 4642
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 1.9521048069000244,
+      "learning_rate": 1.6597510373443984e-06,
+      "loss": 0.1701,
+      "step": 4643
+    },
+    {
+      "epoch": 2.7525925925925927,
+      "grad_norm": 1.8742293119430542,
+      "learning_rate": 1.655799249160245e-06,
+      "loss": 0.1243,
+      "step": 4644
+    },
+    {
+      "epoch": 2.753185185185185,
+      "grad_norm": 2.3207528591156006,
+      "learning_rate": 1.6518474609760918e-06,
+      "loss": 0.171,
+      "step": 4645
+    },
+    {
+      "epoch": 2.7537777777777777,
+      "grad_norm": 1.6360973119735718,
+      "learning_rate": 1.6478956727919384e-06,
+      "loss": 0.0914,
+      "step": 4646
+    },
+    {
+      "epoch": 2.7543703703703706,
+      "grad_norm": 1.9251879453659058,
+      "learning_rate": 1.643943884607785e-06,
+      "loss": 0.1168,
+      "step": 4647
+    },
+    {
+      "epoch": 2.754962962962963,
+      "grad_norm": 1.699432611465454,
+      "learning_rate": 1.6399920964236318e-06,
+      "loss": 0.1464,
+      "step": 4648
+    },
+    {
+      "epoch": 2.7555555555555555,
+      "grad_norm": 1.9950528144836426,
+      "learning_rate": 1.6360403082394784e-06,
+      "loss": 0.1569,
+      "step": 4649
+    },
+    {
+      "epoch": 2.756148148148148,
+      "grad_norm": 2.3842270374298096,
+      "learning_rate": 1.632088520055325e-06,
+      "loss": 0.2725,
+      "step": 4650
+    },
+    {
+      "epoch": 2.7567407407407405,
+      "grad_norm": 2.9149863719940186,
+      "learning_rate": 1.628136731871172e-06,
+      "loss": 0.2741,
+      "step": 4651
+    },
+    {
+      "epoch": 2.7573333333333334,
+      "grad_norm": 1.705946683883667,
+      "learning_rate": 1.6241849436870186e-06,
+      "loss": 0.1109,
+      "step": 4652
+    },
+    {
+      "epoch": 2.757925925925926,
+      "grad_norm": 2.9647819995880127,
+      "learning_rate": 1.6202331555028653e-06,
+      "loss": 0.1717,
+      "step": 4653
+    },
+    {
+      "epoch": 2.7585185185185184,
+      "grad_norm": 1.9328571557998657,
+      "learning_rate": 1.616281367318712e-06,
+      "loss": 0.2188,
+      "step": 4654
+    },
+    {
+      "epoch": 2.7591111111111113,
+      "grad_norm": 1.5065594911575317,
+      "learning_rate": 1.6123295791345586e-06,
+      "loss": 0.1651,
+      "step": 4655
+    },
+    {
+      "epoch": 2.7597037037037038,
+      "grad_norm": 2.095130681991577,
+      "learning_rate": 1.6083777909504053e-06,
+      "loss": 0.228,
+      "step": 4656
+    },
+    {
+      "epoch": 2.7602962962962962,
+      "grad_norm": 2.557656764984131,
+      "learning_rate": 1.604426002766252e-06,
+      "loss": 0.1723,
+      "step": 4657
+    },
+    {
+      "epoch": 2.7608888888888887,
+      "grad_norm": 1.8311078548431396,
+      "learning_rate": 1.6004742145820984e-06,
+      "loss": 0.14,
+      "step": 4658
+    },
+    {
+      "epoch": 2.7614814814814816,
+      "grad_norm": 2.6542649269104004,
+      "learning_rate": 1.596522426397945e-06,
+      "loss": 0.2003,
+      "step": 4659
+    },
+    {
+      "epoch": 2.762074074074074,
+      "grad_norm": 2.5936450958251953,
+      "learning_rate": 1.5925706382137918e-06,
+      "loss": 0.1328,
+      "step": 4660
+    },
+    {
+      "epoch": 2.7626666666666666,
+      "grad_norm": 2.6595942974090576,
+      "learning_rate": 1.5886188500296384e-06,
+      "loss": 0.1431,
+      "step": 4661
+    },
+    {
+      "epoch": 2.763259259259259,
+      "grad_norm": 2.3078432083129883,
+      "learning_rate": 1.584667061845485e-06,
+      "loss": 0.1435,
+      "step": 4662
+    },
+    {
+      "epoch": 2.763851851851852,
+      "grad_norm": 2.688915252685547,
+      "learning_rate": 1.5807152736613318e-06,
+      "loss": 0.199,
+      "step": 4663
+    },
+    {
+      "epoch": 2.7644444444444445,
+      "grad_norm": 2.3189353942871094,
+      "learning_rate": 1.5767634854771784e-06,
+      "loss": 0.1637,
+      "step": 4664
+    },
+    {
+      "epoch": 2.765037037037037,
+      "grad_norm": 2.436385154724121,
+      "learning_rate": 1.572811697293025e-06,
+      "loss": 0.1771,
+      "step": 4665
+    },
+    {
+      "epoch": 2.76562962962963,
+      "grad_norm": 2.400707483291626,
+      "learning_rate": 1.568859909108872e-06,
+      "loss": 0.1731,
+      "step": 4666
+    },
+    {
+      "epoch": 2.7662222222222224,
+      "grad_norm": 1.9705009460449219,
+      "learning_rate": 1.5649081209247186e-06,
+      "loss": 0.1535,
+      "step": 4667
+    },
+    {
+      "epoch": 2.766814814814815,
+      "grad_norm": 1.694298267364502,
+      "learning_rate": 1.5609563327405653e-06,
+      "loss": 0.177,
+      "step": 4668
+    },
+    {
+      "epoch": 2.7674074074074073,
+      "grad_norm": 2.565598487854004,
+      "learning_rate": 1.557004544556412e-06,
+      "loss": 0.245,
+      "step": 4669
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 2.4082212448120117,
+      "learning_rate": 1.5530527563722586e-06,
+      "loss": 0.1938,
+      "step": 4670
+    },
+    {
+      "epoch": 2.7685925925925927,
+      "grad_norm": 1.9520014524459839,
+      "learning_rate": 1.5491009681881053e-06,
+      "loss": 0.1264,
+      "step": 4671
+    },
+    {
+      "epoch": 2.769185185185185,
+      "grad_norm": 1.6156048774719238,
+      "learning_rate": 1.545149180003952e-06,
+      "loss": 0.1092,
+      "step": 4672
+    },
+    {
+      "epoch": 2.7697777777777777,
+      "grad_norm": 2.3268043994903564,
+      "learning_rate": 1.5411973918197986e-06,
+      "loss": 0.1439,
+      "step": 4673
+    },
+    {
+      "epoch": 2.7703703703703706,
+      "grad_norm": 2.262697696685791,
+      "learning_rate": 1.5372456036356453e-06,
+      "loss": 0.1827,
+      "step": 4674
+    },
+    {
+      "epoch": 2.770962962962963,
+      "grad_norm": 2.047475814819336,
+      "learning_rate": 1.533293815451492e-06,
+      "loss": 0.1279,
+      "step": 4675
+    },
+    {
+      "epoch": 2.7715555555555556,
+      "grad_norm": 3.1266074180603027,
+      "learning_rate": 1.5293420272673386e-06,
+      "loss": 0.1458,
+      "step": 4676
+    },
+    {
+      "epoch": 2.772148148148148,
+      "grad_norm": 1.7601218223571777,
+      "learning_rate": 1.5253902390831853e-06,
+      "loss": 0.1166,
+      "step": 4677
+    },
+    {
+      "epoch": 2.7727407407407405,
+      "grad_norm": 1.79490065574646,
+      "learning_rate": 1.5214384508990318e-06,
+      "loss": 0.1421,
+      "step": 4678
+    },
+    {
+      "epoch": 2.7733333333333334,
+      "grad_norm": 2.657358407974243,
+      "learning_rate": 1.5174866627148784e-06,
+      "loss": 0.2136,
+      "step": 4679
+    },
+    {
+      "epoch": 2.773925925925926,
+      "grad_norm": 2.7938284873962402,
+      "learning_rate": 1.513534874530725e-06,
+      "loss": 0.2652,
+      "step": 4680
+    },
+    {
+      "epoch": 2.7745185185185184,
+      "grad_norm": 1.4367034435272217,
+      "learning_rate": 1.5095830863465718e-06,
+      "loss": 0.0911,
+      "step": 4681
+    },
+    {
+      "epoch": 2.7751111111111113,
+      "grad_norm": 2.0133352279663086,
+      "learning_rate": 1.5056312981624186e-06,
+      "loss": 0.1027,
+      "step": 4682
+    },
+    {
+      "epoch": 2.775703703703704,
+      "grad_norm": 2.146340847015381,
+      "learning_rate": 1.5016795099782653e-06,
+      "loss": 0.1164,
+      "step": 4683
+    },
+    {
+      "epoch": 2.7762962962962963,
+      "grad_norm": 3.0068118572235107,
+      "learning_rate": 1.497727721794112e-06,
+      "loss": 0.4062,
+      "step": 4684
+    },
+    {
+      "epoch": 2.7768888888888887,
+      "grad_norm": 2.3259925842285156,
+      "learning_rate": 1.4937759336099586e-06,
+      "loss": 0.2454,
+      "step": 4685
+    },
+    {
+      "epoch": 2.777481481481481,
+      "grad_norm": 2.248741865158081,
+      "learning_rate": 1.4898241454258053e-06,
+      "loss": 0.2399,
+      "step": 4686
+    },
+    {
+      "epoch": 2.778074074074074,
+      "grad_norm": 2.3746299743652344,
+      "learning_rate": 1.485872357241652e-06,
+      "loss": 0.1547,
+      "step": 4687
+    },
+    {
+      "epoch": 2.7786666666666666,
+      "grad_norm": 2.455156087875366,
+      "learning_rate": 1.4819205690574986e-06,
+      "loss": 0.186,
+      "step": 4688
+    },
+    {
+      "epoch": 2.779259259259259,
+      "grad_norm": 2.4752140045166016,
+      "learning_rate": 1.4779687808733453e-06,
+      "loss": 0.1428,
+      "step": 4689
+    },
+    {
+      "epoch": 2.779851851851852,
+      "grad_norm": 1.9635252952575684,
+      "learning_rate": 1.474016992689192e-06,
+      "loss": 0.1408,
+      "step": 4690
+    },
+    {
+      "epoch": 2.7804444444444445,
+      "grad_norm": 2.6431477069854736,
+      "learning_rate": 1.4700652045050386e-06,
+      "loss": 0.2324,
+      "step": 4691
+    },
+    {
+      "epoch": 2.781037037037037,
+      "grad_norm": 2.9544341564178467,
+      "learning_rate": 1.4661134163208853e-06,
+      "loss": 0.1623,
+      "step": 4692
+    },
+    {
+      "epoch": 2.78162962962963,
+      "grad_norm": 2.2338876724243164,
+      "learning_rate": 1.4621616281367322e-06,
+      "loss": 0.1334,
+      "step": 4693
+    },
+    {
+      "epoch": 2.7822222222222224,
+      "grad_norm": 2.1532845497131348,
+      "learning_rate": 1.4582098399525788e-06,
+      "loss": 0.1276,
+      "step": 4694
+    },
+    {
+      "epoch": 2.782814814814815,
+      "grad_norm": 2.0238630771636963,
+      "learning_rate": 1.4542580517684255e-06,
+      "loss": 0.1688,
+      "step": 4695
+    },
+    {
+      "epoch": 2.7834074074074073,
+      "grad_norm": 1.7881401777267456,
+      "learning_rate": 1.4503062635842722e-06,
+      "loss": 0.1504,
+      "step": 4696
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 3.9476263523101807,
+      "learning_rate": 1.4463544754001188e-06,
+      "loss": 0.1765,
+      "step": 4697
+    },
+    {
+      "epoch": 2.7845925925925927,
+      "grad_norm": 3.1113617420196533,
+      "learning_rate": 1.4424026872159653e-06,
+      "loss": 0.2314,
+      "step": 4698
+    },
+    {
+      "epoch": 2.785185185185185,
+      "grad_norm": 2.500380754470825,
+      "learning_rate": 1.438450899031812e-06,
+      "loss": 0.0923,
+      "step": 4699
+    },
+    {
+      "epoch": 2.7857777777777777,
+      "grad_norm": 1.7939602136611938,
+      "learning_rate": 1.4344991108476586e-06,
+      "loss": 0.1134,
+      "step": 4700
+    },
+    {
+      "epoch": 2.7863703703703706,
+      "grad_norm": 5.522995471954346,
+      "learning_rate": 1.4305473226635053e-06,
+      "loss": 0.3859,
+      "step": 4701
+    },
+    {
+      "epoch": 2.786962962962963,
+      "grad_norm": 2.488788366317749,
+      "learning_rate": 1.426595534479352e-06,
+      "loss": 0.1559,
+      "step": 4702
+    },
+    {
+      "epoch": 2.7875555555555556,
+      "grad_norm": 2.9358859062194824,
+      "learning_rate": 1.4226437462951986e-06,
+      "loss": 0.2718,
+      "step": 4703
+    },
+    {
+      "epoch": 2.788148148148148,
+      "grad_norm": 2.6359994411468506,
+      "learning_rate": 1.4186919581110453e-06,
+      "loss": 0.2616,
+      "step": 4704
+    },
+    {
+      "epoch": 2.7887407407407405,
+      "grad_norm": 1.6303772926330566,
+      "learning_rate": 1.414740169926892e-06,
+      "loss": 0.1552,
+      "step": 4705
+    },
+    {
+      "epoch": 2.7893333333333334,
+      "grad_norm": 2.355513095855713,
+      "learning_rate": 1.4107883817427386e-06,
+      "loss": 0.1878,
+      "step": 4706
+    },
+    {
+      "epoch": 2.789925925925926,
+      "grad_norm": 3.6442155838012695,
+      "learning_rate": 1.4068365935585853e-06,
+      "loss": 0.1903,
+      "step": 4707
+    },
+    {
+      "epoch": 2.7905185185185184,
+      "grad_norm": 2.6087381839752197,
+      "learning_rate": 1.402884805374432e-06,
+      "loss": 0.2744,
+      "step": 4708
+    },
+    {
+      "epoch": 2.7911111111111113,
+      "grad_norm": 1.9299527406692505,
+      "learning_rate": 1.3989330171902788e-06,
+      "loss": 0.1415,
+      "step": 4709
+    },
+    {
+      "epoch": 2.791703703703704,
+      "grad_norm": 3.622784376144409,
+      "learning_rate": 1.3949812290061255e-06,
+      "loss": 0.2639,
+      "step": 4710
+    },
+    {
+      "epoch": 2.7922962962962963,
+      "grad_norm": 2.4350340366363525,
+      "learning_rate": 1.3910294408219722e-06,
+      "loss": 0.1894,
+      "step": 4711
+    },
+    {
+      "epoch": 2.7928888888888888,
+      "grad_norm": 2.839541435241699,
+      "learning_rate": 1.3870776526378188e-06,
+      "loss": 0.1842,
+      "step": 4712
+    },
+    {
+      "epoch": 2.7934814814814812,
+      "grad_norm": 2.834523916244507,
+      "learning_rate": 1.3831258644536655e-06,
+      "loss": 0.161,
+      "step": 4713
+    },
+    {
+      "epoch": 2.794074074074074,
+      "grad_norm": 2.2598514556884766,
+      "learning_rate": 1.3791740762695122e-06,
+      "loss": 0.1991,
+      "step": 4714
+    },
+    {
+      "epoch": 2.7946666666666666,
+      "grad_norm": 1.9701836109161377,
+      "learning_rate": 1.3752222880853588e-06,
+      "loss": 0.1278,
+      "step": 4715
+    },
+    {
+      "epoch": 2.795259259259259,
+      "grad_norm": 2.4823856353759766,
+      "learning_rate": 1.3712704999012055e-06,
+      "loss": 0.2033,
+      "step": 4716
+    },
+    {
+      "epoch": 2.795851851851852,
+      "grad_norm": 2.335707664489746,
+      "learning_rate": 1.3673187117170522e-06,
+      "loss": 0.3024,
+      "step": 4717
+    },
+    {
+      "epoch": 2.7964444444444445,
+      "grad_norm": 1.4774240255355835,
+      "learning_rate": 1.3633669235328986e-06,
+      "loss": 0.074,
+      "step": 4718
+    },
+    {
+      "epoch": 2.797037037037037,
+      "grad_norm": 2.8107171058654785,
+      "learning_rate": 1.3594151353487453e-06,
+      "loss": 0.1457,
+      "step": 4719
+    },
+    {
+      "epoch": 2.7976296296296295,
+      "grad_norm": 2.306912422180176,
+      "learning_rate": 1.355463347164592e-06,
+      "loss": 0.2186,
+      "step": 4720
+    },
+    {
+      "epoch": 2.7982222222222224,
+      "grad_norm": 2.18123197555542,
+      "learning_rate": 1.3515115589804386e-06,
+      "loss": 0.2455,
+      "step": 4721
+    },
+    {
+      "epoch": 2.798814814814815,
+      "grad_norm": 1.6927053928375244,
+      "learning_rate": 1.3475597707962853e-06,
+      "loss": 0.1225,
+      "step": 4722
+    },
+    {
+      "epoch": 2.7994074074074073,
+      "grad_norm": 3.125947952270508,
+      "learning_rate": 1.343607982612132e-06,
+      "loss": 0.1849,
+      "step": 4723
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.4035861492156982,
+      "learning_rate": 1.3396561944279788e-06,
+      "loss": 0.2546,
+      "step": 4724
+    },
+    {
+      "epoch": 2.8005925925925927,
+      "grad_norm": 1.6479917764663696,
+      "learning_rate": 1.3357044062438255e-06,
+      "loss": 0.0856,
+      "step": 4725
+    },
+    {
+      "epoch": 2.8011851851851852,
+      "grad_norm": 2.8310492038726807,
+      "learning_rate": 1.3317526180596722e-06,
+      "loss": 0.2304,
+      "step": 4726
+    },
+    {
+      "epoch": 2.8017777777777777,
+      "grad_norm": 3.9132392406463623,
+      "learning_rate": 1.3278008298755188e-06,
+      "loss": 0.4459,
+      "step": 4727
+    },
+    {
+      "epoch": 2.8023703703703706,
+      "grad_norm": 2.1705002784729004,
+      "learning_rate": 1.3238490416913655e-06,
+      "loss": 0.123,
+      "step": 4728
+    },
+    {
+      "epoch": 2.802962962962963,
+      "grad_norm": 2.283473014831543,
+      "learning_rate": 1.3198972535072122e-06,
+      "loss": 0.1794,
+      "step": 4729
+    },
+    {
+      "epoch": 2.8035555555555556,
+      "grad_norm": 1.7047606706619263,
+      "learning_rate": 1.3159454653230588e-06,
+      "loss": 0.0824,
+      "step": 4730
+    },
+    {
+      "epoch": 2.804148148148148,
+      "grad_norm": 1.4930534362792969,
+      "learning_rate": 1.3119936771389055e-06,
+      "loss": 0.0989,
+      "step": 4731
+    },
+    {
+      "epoch": 2.8047407407407405,
+      "grad_norm": 2.03371262550354,
+      "learning_rate": 1.3080418889547522e-06,
+      "loss": 0.1729,
+      "step": 4732
+    },
+    {
+      "epoch": 2.8053333333333335,
+      "grad_norm": 2.0074567794799805,
+      "learning_rate": 1.3040901007705988e-06,
+      "loss": 0.2216,
+      "step": 4733
+    },
+    {
+      "epoch": 2.805925925925926,
+      "grad_norm": 2.6846840381622314,
+      "learning_rate": 1.3001383125864455e-06,
+      "loss": 0.1953,
+      "step": 4734
+    },
+    {
+      "epoch": 2.8065185185185184,
+      "grad_norm": 1.6522836685180664,
+      "learning_rate": 1.2961865244022922e-06,
+      "loss": 0.163,
+      "step": 4735
+    },
+    {
+      "epoch": 2.8071111111111113,
+      "grad_norm": 1.6607626676559448,
+      "learning_rate": 1.292234736218139e-06,
+      "loss": 0.1816,
+      "step": 4736
+    },
+    {
+      "epoch": 2.807703703703704,
+      "grad_norm": 1.6997953653335571,
+      "learning_rate": 1.2882829480339853e-06,
+      "loss": 0.1396,
+      "step": 4737
+    },
+    {
+      "epoch": 2.8082962962962963,
+      "grad_norm": 3.0058717727661133,
+      "learning_rate": 1.284331159849832e-06,
+      "loss": 0.2176,
+      "step": 4738
+    },
+    {
+      "epoch": 2.8088888888888888,
+      "grad_norm": 1.6980392932891846,
+      "learning_rate": 1.2803793716656786e-06,
+      "loss": 0.1334,
+      "step": 4739
+    },
+    {
+      "epoch": 2.8094814814814812,
+      "grad_norm": 1.5808247327804565,
+      "learning_rate": 1.2764275834815255e-06,
+      "loss": 0.1485,
+      "step": 4740
+    },
+    {
+      "epoch": 2.810074074074074,
+      "grad_norm": 1.5990707874298096,
+      "learning_rate": 1.2724757952973722e-06,
+      "loss": 0.1056,
+      "step": 4741
+    },
+    {
+      "epoch": 2.8106666666666666,
+      "grad_norm": 2.700678586959839,
+      "learning_rate": 1.2685240071132188e-06,
+      "loss": 0.1883,
+      "step": 4742
+    },
+    {
+      "epoch": 2.811259259259259,
+      "grad_norm": 2.1362178325653076,
+      "learning_rate": 1.2645722189290655e-06,
+      "loss": 0.1665,
+      "step": 4743
+    },
+    {
+      "epoch": 2.811851851851852,
+      "grad_norm": 2.3904190063476562,
+      "learning_rate": 1.2606204307449122e-06,
+      "loss": 0.1547,
+      "step": 4744
+    },
+    {
+      "epoch": 2.8124444444444445,
+      "grad_norm": 3.1900997161865234,
+      "learning_rate": 1.2566686425607588e-06,
+      "loss": 0.1913,
+      "step": 4745
+    },
+    {
+      "epoch": 2.813037037037037,
+      "grad_norm": 2.8991341590881348,
+      "learning_rate": 1.2527168543766055e-06,
+      "loss": 0.2369,
+      "step": 4746
+    },
+    {
+      "epoch": 2.8136296296296295,
+      "grad_norm": 2.347842216491699,
+      "learning_rate": 1.2487650661924522e-06,
+      "loss": 0.2542,
+      "step": 4747
+    },
+    {
+      "epoch": 2.814222222222222,
+      "grad_norm": 2.3187992572784424,
+      "learning_rate": 1.2448132780082988e-06,
+      "loss": 0.1298,
+      "step": 4748
+    },
+    {
+      "epoch": 2.814814814814815,
+      "grad_norm": 3.0101237297058105,
+      "learning_rate": 1.2408614898241455e-06,
+      "loss": 0.1704,
+      "step": 4749
+    },
+    {
+      "epoch": 2.8154074074074074,
+      "grad_norm": 1.5903377532958984,
+      "learning_rate": 1.2369097016399922e-06,
+      "loss": 0.074,
+      "step": 4750
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 1.5171000957489014,
+      "learning_rate": 1.2329579134558388e-06,
+      "loss": 0.1205,
+      "step": 4751
+    },
+    {
+      "epoch": 2.8165925925925928,
+      "grad_norm": 2.763653516769409,
+      "learning_rate": 1.2290061252716855e-06,
+      "loss": 0.2095,
+      "step": 4752
+    },
+    {
+      "epoch": 2.8171851851851852,
+      "grad_norm": 1.9128062725067139,
+      "learning_rate": 1.2250543370875322e-06,
+      "loss": 0.1659,
+      "step": 4753
+    },
+    {
+      "epoch": 2.8177777777777777,
+      "grad_norm": 1.9702473878860474,
+      "learning_rate": 1.2211025489033788e-06,
+      "loss": 0.14,
+      "step": 4754
+    },
+    {
+      "epoch": 2.8183703703703706,
+      "grad_norm": 3.231428861618042,
+      "learning_rate": 1.2171507607192255e-06,
+      "loss": 0.1861,
+      "step": 4755
+    },
+    {
+      "epoch": 2.818962962962963,
+      "grad_norm": 2.0691776275634766,
+      "learning_rate": 1.2131989725350722e-06,
+      "loss": 0.2103,
+      "step": 4756
+    },
+    {
+      "epoch": 2.8195555555555556,
+      "grad_norm": 2.262699604034424,
+      "learning_rate": 1.2092471843509188e-06,
+      "loss": 0.1626,
+      "step": 4757
+    },
+    {
+      "epoch": 2.820148148148148,
+      "grad_norm": 1.8812389373779297,
+      "learning_rate": 1.2052953961667655e-06,
+      "loss": 0.1451,
+      "step": 4758
+    },
+    {
+      "epoch": 2.8207407407407405,
+      "grad_norm": 2.577340841293335,
+      "learning_rate": 1.2013436079826124e-06,
+      "loss": 0.1879,
+      "step": 4759
+    },
+    {
+      "epoch": 2.8213333333333335,
+      "grad_norm": 2.1442720890045166,
+      "learning_rate": 1.197391819798459e-06,
+      "loss": 0.1455,
+      "step": 4760
+    },
+    {
+      "epoch": 2.821925925925926,
+      "grad_norm": 2.4360365867614746,
+      "learning_rate": 1.1934400316143057e-06,
+      "loss": 0.2614,
+      "step": 4761
+    },
+    {
+      "epoch": 2.8225185185185184,
+      "grad_norm": 3.220851182937622,
+      "learning_rate": 1.1894882434301522e-06,
+      "loss": 0.3089,
+      "step": 4762
+    },
+    {
+      "epoch": 2.8231111111111113,
+      "grad_norm": 2.6141703128814697,
+      "learning_rate": 1.1855364552459988e-06,
+      "loss": 0.2068,
+      "step": 4763
+    },
+    {
+      "epoch": 2.823703703703704,
+      "grad_norm": 2.042703628540039,
+      "learning_rate": 1.1815846670618455e-06,
+      "loss": 0.1663,
+      "step": 4764
+    },
+    {
+      "epoch": 2.8242962962962963,
+      "grad_norm": 2.097191333770752,
+      "learning_rate": 1.1776328788776922e-06,
+      "loss": 0.1893,
+      "step": 4765
+    },
+    {
+      "epoch": 2.824888888888889,
+      "grad_norm": 2.682542324066162,
+      "learning_rate": 1.1736810906935388e-06,
+      "loss": 0.2577,
+      "step": 4766
+    },
+    {
+      "epoch": 2.8254814814814813,
+      "grad_norm": 3.129164695739746,
+      "learning_rate": 1.1697293025093857e-06,
+      "loss": 0.3025,
+      "step": 4767
+    },
+    {
+      "epoch": 2.826074074074074,
+      "grad_norm": 1.812888741493225,
+      "learning_rate": 1.1657775143252324e-06,
+      "loss": 0.1201,
+      "step": 4768
+    },
+    {
+      "epoch": 2.8266666666666667,
+      "grad_norm": 2.1552858352661133,
+      "learning_rate": 1.161825726141079e-06,
+      "loss": 0.1667,
+      "step": 4769
+    },
+    {
+      "epoch": 2.827259259259259,
+      "grad_norm": 2.3988449573516846,
+      "learning_rate": 1.1578739379569257e-06,
+      "loss": 0.1803,
+      "step": 4770
+    },
+    {
+      "epoch": 2.827851851851852,
+      "grad_norm": 2.003761053085327,
+      "learning_rate": 1.1539221497727724e-06,
+      "loss": 0.1998,
+      "step": 4771
+    },
+    {
+      "epoch": 2.8284444444444445,
+      "grad_norm": 2.147265672683716,
+      "learning_rate": 1.1499703615886188e-06,
+      "loss": 0.1222,
+      "step": 4772
+    },
+    {
+      "epoch": 2.829037037037037,
+      "grad_norm": 1.7306731939315796,
+      "learning_rate": 1.1460185734044655e-06,
+      "loss": 0.1164,
+      "step": 4773
+    },
+    {
+      "epoch": 2.8296296296296295,
+      "grad_norm": 2.0561461448669434,
+      "learning_rate": 1.1420667852203122e-06,
+      "loss": 0.1171,
+      "step": 4774
+    },
+    {
+      "epoch": 2.830222222222222,
+      "grad_norm": 2.1161744594573975,
+      "learning_rate": 1.138114997036159e-06,
+      "loss": 0.1152,
+      "step": 4775
+    },
+    {
+      "epoch": 2.830814814814815,
+      "grad_norm": 2.0896787643432617,
+      "learning_rate": 1.1341632088520057e-06,
+      "loss": 0.159,
+      "step": 4776
+    },
+    {
+      "epoch": 2.8314074074074074,
+      "grad_norm": 2.887883424758911,
+      "learning_rate": 1.1302114206678524e-06,
+      "loss": 0.1439,
+      "step": 4777
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 1.8060355186462402,
+      "learning_rate": 1.126259632483699e-06,
+      "loss": 0.1306,
+      "step": 4778
+    },
+    {
+      "epoch": 2.8325925925925928,
+      "grad_norm": 2.565831422805786,
+      "learning_rate": 1.1223078442995457e-06,
+      "loss": 0.147,
+      "step": 4779
+    },
+    {
+      "epoch": 2.8331851851851853,
+      "grad_norm": 2.571537733078003,
+      "learning_rate": 1.1183560561153924e-06,
+      "loss": 0.1405,
+      "step": 4780
+    },
+    {
+      "epoch": 2.8337777777777777,
+      "grad_norm": 1.910696029663086,
+      "learning_rate": 1.114404267931239e-06,
+      "loss": 0.179,
+      "step": 4781
+    },
+    {
+      "epoch": 2.83437037037037,
+      "grad_norm": 2.171130895614624,
+      "learning_rate": 1.1104524797470857e-06,
+      "loss": 0.1767,
+      "step": 4782
+    },
+    {
+      "epoch": 2.834962962962963,
+      "grad_norm": 2.121032238006592,
+      "learning_rate": 1.1065006915629324e-06,
+      "loss": 0.0812,
+      "step": 4783
+    },
+    {
+      "epoch": 2.8355555555555556,
+      "grad_norm": 4.565988540649414,
+      "learning_rate": 1.102548903378779e-06,
+      "loss": 0.1833,
+      "step": 4784
+    },
+    {
+      "epoch": 2.836148148148148,
+      "grad_norm": 2.4931325912475586,
+      "learning_rate": 1.0985971151946257e-06,
+      "loss": 0.218,
+      "step": 4785
+    },
+    {
+      "epoch": 2.8367407407407406,
+      "grad_norm": 2.1762523651123047,
+      "learning_rate": 1.0946453270104724e-06,
+      "loss": 0.2034,
+      "step": 4786
+    },
+    {
+      "epoch": 2.8373333333333335,
+      "grad_norm": 2.4544637203216553,
+      "learning_rate": 1.090693538826319e-06,
+      "loss": 0.2161,
+      "step": 4787
+    },
+    {
+      "epoch": 2.837925925925926,
+      "grad_norm": 1.6672571897506714,
+      "learning_rate": 1.0867417506421657e-06,
+      "loss": 0.1307,
+      "step": 4788
+    },
+    {
+      "epoch": 2.8385185185185184,
+      "grad_norm": 3.192408323287964,
+      "learning_rate": 1.0827899624580124e-06,
+      "loss": 0.1791,
+      "step": 4789
+    },
+    {
+      "epoch": 2.8391111111111114,
+      "grad_norm": 2.5257091522216797,
+      "learning_rate": 1.078838174273859e-06,
+      "loss": 0.1788,
+      "step": 4790
+    },
+    {
+      "epoch": 2.839703703703704,
+      "grad_norm": 2.221832036972046,
+      "learning_rate": 1.0748863860897057e-06,
+      "loss": 0.2015,
+      "step": 4791
+    },
+    {
+      "epoch": 2.8402962962962963,
+      "grad_norm": 2.047893524169922,
+      "learning_rate": 1.0709345979055524e-06,
+      "loss": 0.1777,
+      "step": 4792
+    },
+    {
+      "epoch": 2.840888888888889,
+      "grad_norm": 1.3505316972732544,
+      "learning_rate": 1.066982809721399e-06,
+      "loss": 0.1175,
+      "step": 4793
+    },
+    {
+      "epoch": 2.8414814814814813,
+      "grad_norm": 2.336308240890503,
+      "learning_rate": 1.0630310215372457e-06,
+      "loss": 0.156,
+      "step": 4794
+    },
+    {
+      "epoch": 2.842074074074074,
+      "grad_norm": 1.771143913269043,
+      "learning_rate": 1.0590792333530924e-06,
+      "loss": 0.145,
+      "step": 4795
+    },
+    {
+      "epoch": 2.8426666666666667,
+      "grad_norm": 2.002934455871582,
+      "learning_rate": 1.055127445168939e-06,
+      "loss": 0.0889,
+      "step": 4796
+    },
+    {
+      "epoch": 2.843259259259259,
+      "grad_norm": 1.8333005905151367,
+      "learning_rate": 1.0511756569847857e-06,
+      "loss": 0.1263,
+      "step": 4797
+    },
+    {
+      "epoch": 2.843851851851852,
+      "grad_norm": 2.631608724594116,
+      "learning_rate": 1.0472238688006324e-06,
+      "loss": 0.1428,
+      "step": 4798
+    },
+    {
+      "epoch": 2.8444444444444446,
+      "grad_norm": 2.6394340991973877,
+      "learning_rate": 1.043272080616479e-06,
+      "loss": 0.2039,
+      "step": 4799
+    },
+    {
+      "epoch": 2.845037037037037,
+      "grad_norm": 2.714688301086426,
+      "learning_rate": 1.0393202924323257e-06,
+      "loss": 0.2413,
+      "step": 4800
+    },
+    {
+      "epoch": 2.8456296296296295,
+      "grad_norm": 3.8780953884124756,
+      "learning_rate": 1.0353685042481724e-06,
+      "loss": 0.1586,
+      "step": 4801
+    },
+    {
+      "epoch": 2.846222222222222,
+      "grad_norm": 2.1824281215667725,
+      "learning_rate": 1.031416716064019e-06,
+      "loss": 0.235,
+      "step": 4802
+    },
+    {
+      "epoch": 2.846814814814815,
+      "grad_norm": 2.6991543769836426,
+      "learning_rate": 1.0274649278798657e-06,
+      "loss": 0.2452,
+      "step": 4803
+    },
+    {
+      "epoch": 2.8474074074074074,
+      "grad_norm": 2.2272067070007324,
+      "learning_rate": 1.0235131396957124e-06,
+      "loss": 0.1749,
+      "step": 4804
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 4.72216272354126,
+      "learning_rate": 1.019561351511559e-06,
+      "loss": 0.3052,
+      "step": 4805
+    },
+    {
+      "epoch": 2.848592592592593,
+      "grad_norm": 2.0649616718292236,
+      "learning_rate": 1.0156095633274057e-06,
+      "loss": 0.1242,
+      "step": 4806
+    },
+    {
+      "epoch": 2.8491851851851853,
+      "grad_norm": 1.3712372779846191,
+      "learning_rate": 1.0116577751432524e-06,
+      "loss": 0.0815,
+      "step": 4807
+    },
+    {
+      "epoch": 2.8497777777777777,
+      "grad_norm": 2.027071237564087,
+      "learning_rate": 1.007705986959099e-06,
+      "loss": 0.2041,
+      "step": 4808
+    },
+    {
+      "epoch": 2.85037037037037,
+      "grad_norm": 2.8849754333496094,
+      "learning_rate": 1.0037541987749457e-06,
+      "loss": 0.1766,
+      "step": 4809
+    },
+    {
+      "epoch": 2.8509629629629627,
+      "grad_norm": 3.5350592136383057,
+      "learning_rate": 9.998024105907926e-07,
+      "loss": 0.2416,
+      "step": 4810
+    },
+    {
+      "epoch": 2.8515555555555556,
+      "grad_norm": 2.0606908798217773,
+      "learning_rate": 9.95850622406639e-07,
+      "loss": 0.1578,
+      "step": 4811
+    },
+    {
+      "epoch": 2.852148148148148,
+      "grad_norm": 2.279351234436035,
+      "learning_rate": 9.918988342224857e-07,
+      "loss": 0.2711,
+      "step": 4812
+    },
+    {
+      "epoch": 2.8527407407407406,
+      "grad_norm": 2.1139285564422607,
+      "learning_rate": 9.879470460383324e-07,
+      "loss": 0.1958,
+      "step": 4813
+    },
+    {
+      "epoch": 2.8533333333333335,
+      "grad_norm": 2.347503900527954,
+      "learning_rate": 9.83995257854179e-07,
+      "loss": 0.154,
+      "step": 4814
+    },
+    {
+      "epoch": 2.853925925925926,
+      "grad_norm": 3.159675359725952,
+      "learning_rate": 9.800434696700257e-07,
+      "loss": 0.1907,
+      "step": 4815
+    },
+    {
+      "epoch": 2.8545185185185185,
+      "grad_norm": 2.5635077953338623,
+      "learning_rate": 9.760916814858723e-07,
+      "loss": 0.1896,
+      "step": 4816
+    },
+    {
+      "epoch": 2.8551111111111114,
+      "grad_norm": 1.818298578262329,
+      "learning_rate": 9.721398933017192e-07,
+      "loss": 0.1301,
+      "step": 4817
+    },
+    {
+      "epoch": 2.855703703703704,
+      "grad_norm": 2.099593162536621,
+      "learning_rate": 9.68188105117566e-07,
+      "loss": 0.1433,
+      "step": 4818
+    },
+    {
+      "epoch": 2.8562962962962963,
+      "grad_norm": 1.9376494884490967,
+      "learning_rate": 9.642363169334126e-07,
+      "loss": 0.2055,
+      "step": 4819
+    },
+    {
+      "epoch": 2.856888888888889,
+      "grad_norm": 3.5965585708618164,
+      "learning_rate": 9.602845287492592e-07,
+      "loss": 0.2075,
+      "step": 4820
+    },
+    {
+      "epoch": 2.8574814814814813,
+      "grad_norm": 1.8959167003631592,
+      "learning_rate": 9.563327405651057e-07,
+      "loss": 0.1278,
+      "step": 4821
+    },
+    {
+      "epoch": 2.858074074074074,
+      "grad_norm": 1.4960888624191284,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 0.1238,
+      "step": 4822
+    },
+    {
+      "epoch": 2.8586666666666667,
+      "grad_norm": 2.153970956802368,
+      "learning_rate": 9.484291641967991e-07,
+      "loss": 0.15,
+      "step": 4823
+    },
+    {
+      "epoch": 2.859259259259259,
+      "grad_norm": 2.5464766025543213,
+      "learning_rate": 9.444773760126458e-07,
+      "loss": 0.1952,
+      "step": 4824
+    },
+    {
+      "epoch": 2.859851851851852,
+      "grad_norm": 1.989761471748352,
+      "learning_rate": 9.405255878284925e-07,
+      "loss": 0.2584,
+      "step": 4825
+    },
+    {
+      "epoch": 2.8604444444444446,
+      "grad_norm": 1.905838131904602,
+      "learning_rate": 9.365737996443391e-07,
+      "loss": 0.1192,
+      "step": 4826
+    },
+    {
+      "epoch": 2.861037037037037,
+      "grad_norm": 2.599214553833008,
+      "learning_rate": 9.326220114601858e-07,
+      "loss": 0.184,
+      "step": 4827
+    },
+    {
+      "epoch": 2.8616296296296295,
+      "grad_norm": 1.8701496124267578,
+      "learning_rate": 9.286702232760326e-07,
+      "loss": 0.1816,
+      "step": 4828
+    },
+    {
+      "epoch": 2.862222222222222,
+      "grad_norm": 3.0911011695861816,
+      "learning_rate": 9.247184350918792e-07,
+      "loss": 0.2559,
+      "step": 4829
+    },
+    {
+      "epoch": 2.862814814814815,
+      "grad_norm": 1.8369282484054565,
+      "learning_rate": 9.207666469077259e-07,
+      "loss": 0.1842,
+      "step": 4830
+    },
+    {
+      "epoch": 2.8634074074074074,
+      "grad_norm": 2.2862184047698975,
+      "learning_rate": 9.168148587235725e-07,
+      "loss": 0.1666,
+      "step": 4831
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 2.619333505630493,
+      "learning_rate": 9.128630705394191e-07,
+      "loss": 0.1562,
+      "step": 4832
+    },
+    {
+      "epoch": 2.864592592592593,
+      "grad_norm": 2.0756847858428955,
+      "learning_rate": 9.089112823552658e-07,
+      "loss": 0.1947,
+      "step": 4833
+    },
+    {
+      "epoch": 2.8651851851851853,
+      "grad_norm": 2.5435774326324463,
+      "learning_rate": 9.049594941711125e-07,
+      "loss": 0.1389,
+      "step": 4834
+    },
+    {
+      "epoch": 2.8657777777777778,
+      "grad_norm": 1.6212078332901,
+      "learning_rate": 9.010077059869591e-07,
+      "loss": 0.1633,
+      "step": 4835
+    },
+    {
+      "epoch": 2.8663703703703702,
+      "grad_norm": 1.8553410768508911,
+      "learning_rate": 8.970559178028059e-07,
+      "loss": 0.0989,
+      "step": 4836
+    },
+    {
+      "epoch": 2.8669629629629627,
+      "grad_norm": 2.6329870223999023,
+      "learning_rate": 8.931041296186526e-07,
+      "loss": 0.2678,
+      "step": 4837
+    },
+    {
+      "epoch": 2.8675555555555556,
+      "grad_norm": 2.1349165439605713,
+      "learning_rate": 8.891523414344992e-07,
+      "loss": 0.1417,
+      "step": 4838
+    },
+    {
+      "epoch": 2.868148148148148,
+      "grad_norm": 2.014561891555786,
+      "learning_rate": 8.852005532503459e-07,
+      "loss": 0.1395,
+      "step": 4839
+    },
+    {
+      "epoch": 2.8687407407407406,
+      "grad_norm": 2.1259329319000244,
+      "learning_rate": 8.812487650661926e-07,
+      "loss": 0.1744,
+      "step": 4840
+    },
+    {
+      "epoch": 2.8693333333333335,
+      "grad_norm": 3.192042350769043,
+      "learning_rate": 8.772969768820391e-07,
+      "loss": 0.2065,
+      "step": 4841
+    },
+    {
+      "epoch": 2.869925925925926,
+      "grad_norm": 1.8384380340576172,
+      "learning_rate": 8.733451886978858e-07,
+      "loss": 0.1198,
+      "step": 4842
+    },
+    {
+      "epoch": 2.8705185185185185,
+      "grad_norm": 1.6659715175628662,
+      "learning_rate": 8.693934005137324e-07,
+      "loss": 0.1237,
+      "step": 4843
+    },
+    {
+      "epoch": 2.871111111111111,
+      "grad_norm": 2.1056721210479736,
+      "learning_rate": 8.654416123295792e-07,
+      "loss": 0.1517,
+      "step": 4844
+    },
+    {
+      "epoch": 2.871703703703704,
+      "grad_norm": 2.6888248920440674,
+      "learning_rate": 8.614898241454259e-07,
+      "loss": 0.1674,
+      "step": 4845
+    },
+    {
+      "epoch": 2.8722962962962963,
+      "grad_norm": 2.00899600982666,
+      "learning_rate": 8.575380359612726e-07,
+      "loss": 0.1362,
+      "step": 4846
+    },
+    {
+      "epoch": 2.872888888888889,
+      "grad_norm": 2.4349896907806396,
+      "learning_rate": 8.535862477771192e-07,
+      "loss": 0.15,
+      "step": 4847
+    },
+    {
+      "epoch": 2.8734814814814813,
+      "grad_norm": 1.7478106021881104,
+      "learning_rate": 8.496344595929659e-07,
+      "loss": 0.1871,
+      "step": 4848
+    },
+    {
+      "epoch": 2.8740740740740742,
+      "grad_norm": 1.8618096113204956,
+      "learning_rate": 8.456826714088127e-07,
+      "loss": 0.1773,
+      "step": 4849
+    },
+    {
+      "epoch": 2.8746666666666667,
+      "grad_norm": 2.218515396118164,
+      "learning_rate": 8.417308832246593e-07,
+      "loss": 0.1859,
+      "step": 4850
+    },
+    {
+      "epoch": 2.875259259259259,
+      "grad_norm": 1.8848520517349243,
+      "learning_rate": 8.377790950405059e-07,
+      "loss": 0.1539,
+      "step": 4851
+    },
+    {
+      "epoch": 2.875851851851852,
+      "grad_norm": 3.863361358642578,
+      "learning_rate": 8.338273068563526e-07,
+      "loss": 0.1856,
+      "step": 4852
+    },
+    {
+      "epoch": 2.8764444444444446,
+      "grad_norm": 2.7637200355529785,
+      "learning_rate": 8.298755186721992e-07,
+      "loss": 0.2757,
+      "step": 4853
+    },
+    {
+      "epoch": 2.877037037037037,
+      "grad_norm": 1.9595390558242798,
+      "learning_rate": 8.259237304880459e-07,
+      "loss": 0.1747,
+      "step": 4854
+    },
+    {
+      "epoch": 2.8776296296296295,
+      "grad_norm": 2.052361488342285,
+      "learning_rate": 8.219719423038925e-07,
+      "loss": 0.1787,
+      "step": 4855
+    },
+    {
+      "epoch": 2.878222222222222,
+      "grad_norm": 3.701378345489502,
+      "learning_rate": 8.180201541197392e-07,
+      "loss": 0.3444,
+      "step": 4856
+    },
+    {
+      "epoch": 2.878814814814815,
+      "grad_norm": 2.126941204071045,
+      "learning_rate": 8.14068365935586e-07,
+      "loss": 0.2663,
+      "step": 4857
+    },
+    {
+      "epoch": 2.8794074074074074,
+      "grad_norm": 1.68427312374115,
+      "learning_rate": 8.101165777514327e-07,
+      "loss": 0.094,
+      "step": 4858
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.702066659927368,
+      "learning_rate": 8.061647895672793e-07,
+      "loss": 0.217,
+      "step": 4859
+    },
+    {
+      "epoch": 2.880592592592593,
+      "grad_norm": 1.9515725374221802,
+      "learning_rate": 8.02213001383126e-07,
+      "loss": 0.1451,
+      "step": 4860
+    },
+    {
+      "epoch": 2.8811851851851853,
+      "grad_norm": 2.3716001510620117,
+      "learning_rate": 7.982612131989725e-07,
+      "loss": 0.1614,
+      "step": 4861
+    },
+    {
+      "epoch": 2.8817777777777778,
+      "grad_norm": 3.079545259475708,
+      "learning_rate": 7.943094250148192e-07,
+      "loss": 0.2077,
+      "step": 4862
+    },
+    {
+      "epoch": 2.8823703703703702,
+      "grad_norm": 2.954794406890869,
+      "learning_rate": 7.903576368306659e-07,
+      "loss": 0.2017,
+      "step": 4863
+    },
+    {
+      "epoch": 2.8829629629629627,
+      "grad_norm": 1.7923318147659302,
+      "learning_rate": 7.864058486465125e-07,
+      "loss": 0.1093,
+      "step": 4864
+    },
+    {
+      "epoch": 2.8835555555555556,
+      "grad_norm": 1.5826505422592163,
+      "learning_rate": 7.824540604623593e-07,
+      "loss": 0.0773,
+      "step": 4865
+    },
+    {
+      "epoch": 2.884148148148148,
+      "grad_norm": 2.9254913330078125,
+      "learning_rate": 7.78502272278206e-07,
+      "loss": 0.2214,
+      "step": 4866
+    },
+    {
+      "epoch": 2.8847407407407406,
+      "grad_norm": 2.350541591644287,
+      "learning_rate": 7.745504840940527e-07,
+      "loss": 0.1716,
+      "step": 4867
+    },
+    {
+      "epoch": 2.8853333333333335,
+      "grad_norm": 2.1893372535705566,
+      "learning_rate": 7.705986959098993e-07,
+      "loss": 0.1921,
+      "step": 4868
+    },
+    {
+      "epoch": 2.885925925925926,
+      "grad_norm": 1.9418907165527344,
+      "learning_rate": 7.66646907725746e-07,
+      "loss": 0.114,
+      "step": 4869
+    },
+    {
+      "epoch": 2.8865185185185185,
+      "grad_norm": 1.6134037971496582,
+      "learning_rate": 7.626951195415927e-07,
+      "loss": 0.1824,
+      "step": 4870
+    },
+    {
+      "epoch": 2.887111111111111,
+      "grad_norm": 3.019221782684326,
+      "learning_rate": 7.587433313574392e-07,
+      "loss": 0.1368,
+      "step": 4871
+    },
+    {
+      "epoch": 2.8877037037037034,
+      "grad_norm": 2.095237970352173,
+      "learning_rate": 7.547915431732859e-07,
+      "loss": 0.1991,
+      "step": 4872
+    },
+    {
+      "epoch": 2.8882962962962964,
+      "grad_norm": 1.8429075479507446,
+      "learning_rate": 7.508397549891327e-07,
+      "loss": 0.1507,
+      "step": 4873
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 2.155060291290283,
+      "learning_rate": 7.468879668049793e-07,
+      "loss": 0.1422,
+      "step": 4874
+    },
+    {
+      "epoch": 2.8894814814814813,
+      "grad_norm": 1.832108974456787,
+      "learning_rate": 7.42936178620826e-07,
+      "loss": 0.1162,
+      "step": 4875
+    },
+    {
+      "epoch": 2.8900740740740742,
+      "grad_norm": 2.943140983581543,
+      "learning_rate": 7.389843904366726e-07,
+      "loss": 0.2012,
+      "step": 4876
+    },
+    {
+      "epoch": 2.8906666666666667,
+      "grad_norm": 2.4728734493255615,
+      "learning_rate": 7.350326022525193e-07,
+      "loss": 0.1771,
+      "step": 4877
+    },
+    {
+      "epoch": 2.891259259259259,
+      "grad_norm": 2.2676339149475098,
+      "learning_rate": 7.310808140683661e-07,
+      "loss": 0.1871,
+      "step": 4878
+    },
+    {
+      "epoch": 2.891851851851852,
+      "grad_norm": 4.078519821166992,
+      "learning_rate": 7.271290258842128e-07,
+      "loss": 0.2846,
+      "step": 4879
+    },
+    {
+      "epoch": 2.8924444444444446,
+      "grad_norm": 1.5522600412368774,
+      "learning_rate": 7.231772377000594e-07,
+      "loss": 0.096,
+      "step": 4880
+    },
+    {
+      "epoch": 2.893037037037037,
+      "grad_norm": 4.0362868309021,
+      "learning_rate": 7.19225449515906e-07,
+      "loss": 0.213,
+      "step": 4881
+    },
+    {
+      "epoch": 2.8936296296296296,
+      "grad_norm": 2.102602481842041,
+      "learning_rate": 7.152736613317526e-07,
+      "loss": 0.1753,
+      "step": 4882
+    },
+    {
+      "epoch": 2.894222222222222,
+      "grad_norm": 1.9245034456253052,
+      "learning_rate": 7.113218731475993e-07,
+      "loss": 0.1081,
+      "step": 4883
+    },
+    {
+      "epoch": 2.894814814814815,
+      "grad_norm": 1.7824409008026123,
+      "learning_rate": 7.07370084963446e-07,
+      "loss": 0.1238,
+      "step": 4884
+    },
+    {
+      "epoch": 2.8954074074074074,
+      "grad_norm": 1.5373015403747559,
+      "learning_rate": 7.034182967792926e-07,
+      "loss": 0.1426,
+      "step": 4885
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 2.2574684619903564,
+      "learning_rate": 6.994665085951394e-07,
+      "loss": 0.2302,
+      "step": 4886
+    },
+    {
+      "epoch": 2.896592592592593,
+      "grad_norm": 2.2988884449005127,
+      "learning_rate": 6.955147204109861e-07,
+      "loss": 0.1747,
+      "step": 4887
+    },
+    {
+      "epoch": 2.8971851851851853,
+      "grad_norm": 1.8035458326339722,
+      "learning_rate": 6.915629322268328e-07,
+      "loss": 0.108,
+      "step": 4888
+    },
+    {
+      "epoch": 2.897777777777778,
+      "grad_norm": 2.250566244125366,
+      "learning_rate": 6.876111440426794e-07,
+      "loss": 0.2143,
+      "step": 4889
+    },
+    {
+      "epoch": 2.8983703703703703,
+      "grad_norm": 2.055515766143799,
+      "learning_rate": 6.836593558585261e-07,
+      "loss": 0.15,
+      "step": 4890
+    },
+    {
+      "epoch": 2.8989629629629627,
+      "grad_norm": 1.7858102321624756,
+      "learning_rate": 6.797075676743726e-07,
+      "loss": 0.1405,
+      "step": 4891
+    },
+    {
+      "epoch": 2.8995555555555557,
+      "grad_norm": 1.6616160869598389,
+      "learning_rate": 6.757557794902193e-07,
+      "loss": 0.088,
+      "step": 4892
+    },
+    {
+      "epoch": 2.900148148148148,
+      "grad_norm": 2.705400228500366,
+      "learning_rate": 6.71803991306066e-07,
+      "loss": 0.2379,
+      "step": 4893
+    },
+    {
+      "epoch": 2.9007407407407406,
+      "grad_norm": 1.9183498620986938,
+      "learning_rate": 6.678522031219127e-07,
+      "loss": 0.1782,
+      "step": 4894
+    },
+    {
+      "epoch": 2.9013333333333335,
+      "grad_norm": 1.5474779605865479,
+      "learning_rate": 6.639004149377594e-07,
+      "loss": 0.1151,
+      "step": 4895
+    },
+    {
+      "epoch": 2.901925925925926,
+      "grad_norm": 2.454240322113037,
+      "learning_rate": 6.599486267536061e-07,
+      "loss": 0.1754,
+      "step": 4896
+    },
+    {
+      "epoch": 2.9025185185185185,
+      "grad_norm": 2.0875301361083984,
+      "learning_rate": 6.559968385694527e-07,
+      "loss": 0.2359,
+      "step": 4897
+    },
+    {
+      "epoch": 2.903111111111111,
+      "grad_norm": 2.74885630607605,
+      "learning_rate": 6.520450503852994e-07,
+      "loss": 0.2371,
+      "step": 4898
+    },
+    {
+      "epoch": 2.9037037037037035,
+      "grad_norm": 1.8239715099334717,
+      "learning_rate": 6.480932622011461e-07,
+      "loss": 0.1482,
+      "step": 4899
+    },
+    {
+      "epoch": 2.9042962962962964,
+      "grad_norm": 3.179741859436035,
+      "learning_rate": 6.441414740169926e-07,
+      "loss": 0.2025,
+      "step": 4900
+    },
+    {
+      "epoch": 2.904888888888889,
+      "grad_norm": 2.015838861465454,
+      "learning_rate": 6.401896858328393e-07,
+      "loss": 0.1793,
+      "step": 4901
+    },
+    {
+      "epoch": 2.9054814814814813,
+      "grad_norm": 2.0077404975891113,
+      "learning_rate": 6.362378976486861e-07,
+      "loss": 0.1588,
+      "step": 4902
+    },
+    {
+      "epoch": 2.9060740740740743,
+      "grad_norm": 2.0545759201049805,
+      "learning_rate": 6.322861094645327e-07,
+      "loss": 0.116,
+      "step": 4903
+    },
+    {
+      "epoch": 2.9066666666666667,
+      "grad_norm": 2.1257147789001465,
+      "learning_rate": 6.283343212803794e-07,
+      "loss": 0.1834,
+      "step": 4904
+    },
+    {
+      "epoch": 2.907259259259259,
+      "grad_norm": 1.8179672956466675,
+      "learning_rate": 6.243825330962261e-07,
+      "loss": 0.1083,
+      "step": 4905
+    },
+    {
+      "epoch": 2.9078518518518517,
+      "grad_norm": 2.654956817626953,
+      "learning_rate": 6.204307449120727e-07,
+      "loss": 0.1711,
+      "step": 4906
+    },
+    {
+      "epoch": 2.9084444444444446,
+      "grad_norm": 2.215987205505371,
+      "learning_rate": 6.164789567279194e-07,
+      "loss": 0.1439,
+      "step": 4907
+    },
+    {
+      "epoch": 2.909037037037037,
+      "grad_norm": 1.927944540977478,
+      "learning_rate": 6.125271685437661e-07,
+      "loss": 0.1485,
+      "step": 4908
+    },
+    {
+      "epoch": 2.9096296296296296,
+      "grad_norm": 1.7326525449752808,
+      "learning_rate": 6.085753803596127e-07,
+      "loss": 0.0904,
+      "step": 4909
+    },
+    {
+      "epoch": 2.910222222222222,
+      "grad_norm": 3.809938907623291,
+      "learning_rate": 6.046235921754594e-07,
+      "loss": 0.2693,
+      "step": 4910
+    },
+    {
+      "epoch": 2.910814814814815,
+      "grad_norm": 1.726118564605713,
+      "learning_rate": 6.006718039913062e-07,
+      "loss": 0.1426,
+      "step": 4911
+    },
+    {
+      "epoch": 2.9114074074074074,
+      "grad_norm": 1.7889341115951538,
+      "learning_rate": 5.967200158071528e-07,
+      "loss": 0.1924,
+      "step": 4912
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 2.7135400772094727,
+      "learning_rate": 5.927682276229994e-07,
+      "loss": 0.2467,
+      "step": 4913
+    },
+    {
+      "epoch": 2.912592592592593,
+      "grad_norm": 2.1014249324798584,
+      "learning_rate": 5.888164394388461e-07,
+      "loss": 0.1318,
+      "step": 4914
+    },
+    {
+      "epoch": 2.9131851851851853,
+      "grad_norm": 3.0291953086853027,
+      "learning_rate": 5.848646512546928e-07,
+      "loss": 0.1516,
+      "step": 4915
+    },
+    {
+      "epoch": 2.913777777777778,
+      "grad_norm": 2.3011813163757324,
+      "learning_rate": 5.809128630705395e-07,
+      "loss": 0.1557,
+      "step": 4916
+    },
+    {
+      "epoch": 2.9143703703703703,
+      "grad_norm": 2.712193727493286,
+      "learning_rate": 5.769610748863862e-07,
+      "loss": 0.2106,
+      "step": 4917
+    },
+    {
+      "epoch": 2.9149629629629628,
+      "grad_norm": 2.7101569175720215,
+      "learning_rate": 5.730092867022327e-07,
+      "loss": 0.235,
+      "step": 4918
+    },
+    {
+      "epoch": 2.9155555555555557,
+      "grad_norm": 3.0768728256225586,
+      "learning_rate": 5.690574985180795e-07,
+      "loss": 0.179,
+      "step": 4919
+    },
+    {
+      "epoch": 2.916148148148148,
+      "grad_norm": 1.8413605690002441,
+      "learning_rate": 5.651057103339262e-07,
+      "loss": 0.1289,
+      "step": 4920
+    },
+    {
+      "epoch": 2.9167407407407406,
+      "grad_norm": 2.057438373565674,
+      "learning_rate": 5.611539221497728e-07,
+      "loss": 0.1644,
+      "step": 4921
+    },
+    {
+      "epoch": 2.9173333333333336,
+      "grad_norm": 3.2557129859924316,
+      "learning_rate": 5.572021339656195e-07,
+      "loss": 0.2034,
+      "step": 4922
+    },
+    {
+      "epoch": 2.917925925925926,
+      "grad_norm": 2.0549092292785645,
+      "learning_rate": 5.532503457814662e-07,
+      "loss": 0.1648,
+      "step": 4923
+    },
+    {
+      "epoch": 2.9185185185185185,
+      "grad_norm": 1.8670666217803955,
+      "learning_rate": 5.492985575973128e-07,
+      "loss": 0.0944,
+      "step": 4924
+    },
+    {
+      "epoch": 2.919111111111111,
+      "grad_norm": 2.1438488960266113,
+      "learning_rate": 5.453467694131595e-07,
+      "loss": 0.1125,
+      "step": 4925
+    },
+    {
+      "epoch": 2.9197037037037035,
+      "grad_norm": 1.3026388883590698,
+      "learning_rate": 5.413949812290062e-07,
+      "loss": 0.0792,
+      "step": 4926
+    },
+    {
+      "epoch": 2.9202962962962964,
+      "grad_norm": 2.1520702838897705,
+      "learning_rate": 5.374431930448528e-07,
+      "loss": 0.155,
+      "step": 4927
+    },
+    {
+      "epoch": 2.920888888888889,
+      "grad_norm": 2.002056121826172,
+      "learning_rate": 5.334914048606995e-07,
+      "loss": 0.1697,
+      "step": 4928
+    },
+    {
+      "epoch": 2.9214814814814813,
+      "grad_norm": 3.1042213439941406,
+      "learning_rate": 5.295396166765462e-07,
+      "loss": 0.1442,
+      "step": 4929
+    },
+    {
+      "epoch": 2.9220740740740743,
+      "grad_norm": 2.2624213695526123,
+      "learning_rate": 5.255878284923928e-07,
+      "loss": 0.2099,
+      "step": 4930
+    },
+    {
+      "epoch": 2.9226666666666667,
+      "grad_norm": 1.9602419137954712,
+      "learning_rate": 5.216360403082395e-07,
+      "loss": 0.1254,
+      "step": 4931
+    },
+    {
+      "epoch": 2.9232592592592592,
+      "grad_norm": 3.8344855308532715,
+      "learning_rate": 5.176842521240862e-07,
+      "loss": 0.2906,
+      "step": 4932
+    },
+    {
+      "epoch": 2.9238518518518517,
+      "grad_norm": 2.666562080383301,
+      "learning_rate": 5.137324639399328e-07,
+      "loss": 0.1418,
+      "step": 4933
+    },
+    {
+      "epoch": 2.924444444444444,
+      "grad_norm": 2.7941040992736816,
+      "learning_rate": 5.097806757557795e-07,
+      "loss": 0.1614,
+      "step": 4934
+    },
+    {
+      "epoch": 2.925037037037037,
+      "grad_norm": 2.1011955738067627,
+      "learning_rate": 5.058288875716262e-07,
+      "loss": 0.1277,
+      "step": 4935
+    },
+    {
+      "epoch": 2.9256296296296296,
+      "grad_norm": 2.0610525608062744,
+      "learning_rate": 5.018770993874728e-07,
+      "loss": 0.1421,
+      "step": 4936
+    },
+    {
+      "epoch": 2.926222222222222,
+      "grad_norm": 2.3354239463806152,
+      "learning_rate": 4.979253112033195e-07,
+      "loss": 0.2303,
+      "step": 4937
+    },
+    {
+      "epoch": 2.926814814814815,
+      "grad_norm": 2.860004186630249,
+      "learning_rate": 4.939735230191662e-07,
+      "loss": 0.1594,
+      "step": 4938
+    },
+    {
+      "epoch": 2.9274074074074075,
+      "grad_norm": 1.5813654661178589,
+      "learning_rate": 4.900217348350128e-07,
+      "loss": 0.1099,
+      "step": 4939
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 2.770392656326294,
+      "learning_rate": 4.860699466508596e-07,
+      "loss": 0.1353,
+      "step": 4940
+    },
+    {
+      "epoch": 2.928592592592593,
+      "grad_norm": 1.6746082305908203,
+      "learning_rate": 4.821181584667063e-07,
+      "loss": 0.1487,
+      "step": 4941
+    },
+    {
+      "epoch": 2.9291851851851853,
+      "grad_norm": 2.224069833755493,
+      "learning_rate": 4.781663702825528e-07,
+      "loss": 0.187,
+      "step": 4942
+    },
+    {
+      "epoch": 2.929777777777778,
+      "grad_norm": 2.5040793418884277,
+      "learning_rate": 4.7421458209839956e-07,
+      "loss": 0.2037,
+      "step": 4943
+    },
+    {
+      "epoch": 2.9303703703703703,
+      "grad_norm": 1.7711395025253296,
+      "learning_rate": 4.702627939142462e-07,
+      "loss": 0.132,
+      "step": 4944
+    },
+    {
+      "epoch": 2.9309629629629628,
+      "grad_norm": 2.700005531311035,
+      "learning_rate": 4.663110057300929e-07,
+      "loss": 0.1584,
+      "step": 4945
+    },
+    {
+      "epoch": 2.9315555555555557,
+      "grad_norm": 1.813720464706421,
+      "learning_rate": 4.623592175459396e-07,
+      "loss": 0.1281,
+      "step": 4946
+    },
+    {
+      "epoch": 2.932148148148148,
+      "grad_norm": 2.3632664680480957,
+      "learning_rate": 4.584074293617862e-07,
+      "loss": 0.1397,
+      "step": 4947
+    },
+    {
+      "epoch": 2.9327407407407406,
+      "grad_norm": 2.2680821418762207,
+      "learning_rate": 4.544556411776329e-07,
+      "loss": 0.1565,
+      "step": 4948
+    },
+    {
+      "epoch": 2.9333333333333336,
+      "grad_norm": 1.7013943195343018,
+      "learning_rate": 4.5050385299347956e-07,
+      "loss": 0.0995,
+      "step": 4949
+    },
+    {
+      "epoch": 2.933925925925926,
+      "grad_norm": 1.8724346160888672,
+      "learning_rate": 4.465520648093263e-07,
+      "loss": 0.1011,
+      "step": 4950
+    },
+    {
+      "epoch": 2.9345185185185185,
+      "grad_norm": 2.342607259750366,
+      "learning_rate": 4.4260027662517294e-07,
+      "loss": 0.1837,
+      "step": 4951
+    },
+    {
+      "epoch": 2.935111111111111,
+      "grad_norm": 3.3937201499938965,
+      "learning_rate": 4.3864848844101956e-07,
+      "loss": 0.2135,
+      "step": 4952
+    },
+    {
+      "epoch": 2.9357037037037035,
+      "grad_norm": 2.916027545928955,
+      "learning_rate": 4.346967002568662e-07,
+      "loss": 0.1302,
+      "step": 4953
+    },
+    {
+      "epoch": 2.9362962962962964,
+      "grad_norm": 2.0827295780181885,
+      "learning_rate": 4.3074491207271294e-07,
+      "loss": 0.1848,
+      "step": 4954
+    },
+    {
+      "epoch": 2.936888888888889,
+      "grad_norm": 2.8810484409332275,
+      "learning_rate": 4.267931238885596e-07,
+      "loss": 0.1992,
+      "step": 4955
+    },
+    {
+      "epoch": 2.9374814814814814,
+      "grad_norm": 1.6730306148529053,
+      "learning_rate": 4.2284133570440633e-07,
+      "loss": 0.1624,
+      "step": 4956
+    },
+    {
+      "epoch": 2.9380740740740743,
+      "grad_norm": 2.331732988357544,
+      "learning_rate": 4.1888954752025294e-07,
+      "loss": 0.1679,
+      "step": 4957
+    },
+    {
+      "epoch": 2.9386666666666668,
+      "grad_norm": 2.856339454650879,
+      "learning_rate": 4.149377593360996e-07,
+      "loss": 0.1932,
+      "step": 4958
+    },
+    {
+      "epoch": 2.9392592592592592,
+      "grad_norm": 2.3299496173858643,
+      "learning_rate": 4.109859711519463e-07,
+      "loss": 0.1967,
+      "step": 4959
+    },
+    {
+      "epoch": 2.9398518518518517,
+      "grad_norm": 2.028568744659424,
+      "learning_rate": 4.07034182967793e-07,
+      "loss": 0.1275,
+      "step": 4960
+    },
+    {
+      "epoch": 2.940444444444444,
+      "grad_norm": 3.309438467025757,
+      "learning_rate": 4.0308239478363966e-07,
+      "loss": 0.1321,
+      "step": 4961
+    },
+    {
+      "epoch": 2.941037037037037,
+      "grad_norm": 2.2231454849243164,
+      "learning_rate": 3.991306065994863e-07,
+      "loss": 0.1497,
+      "step": 4962
+    },
+    {
+      "epoch": 2.9416296296296296,
+      "grad_norm": 1.58977472782135,
+      "learning_rate": 3.9517881841533294e-07,
+      "loss": 0.0893,
+      "step": 4963
+    },
+    {
+      "epoch": 2.942222222222222,
+      "grad_norm": 1.8763960599899292,
+      "learning_rate": 3.9122703023117966e-07,
+      "loss": 0.1657,
+      "step": 4964
+    },
+    {
+      "epoch": 2.942814814814815,
+      "grad_norm": 2.337860345840454,
+      "learning_rate": 3.872752420470263e-07,
+      "loss": 0.2071,
+      "step": 4965
+    },
+    {
+      "epoch": 2.9434074074074075,
+      "grad_norm": 1.8062100410461426,
+      "learning_rate": 3.83323453862873e-07,
+      "loss": 0.1224,
+      "step": 4966
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 2.0571305751800537,
+      "learning_rate": 3.793716656787196e-07,
+      "loss": 0.1664,
+      "step": 4967
+    },
+    {
+      "epoch": 2.9445925925925924,
+      "grad_norm": 2.6339104175567627,
+      "learning_rate": 3.754198774945663e-07,
+      "loss": 0.2585,
+      "step": 4968
+    },
+    {
+      "epoch": 2.9451851851851854,
+      "grad_norm": 2.9098751544952393,
+      "learning_rate": 3.71468089310413e-07,
+      "loss": 0.2087,
+      "step": 4969
+    },
+    {
+      "epoch": 2.945777777777778,
+      "grad_norm": 1.8861726522445679,
+      "learning_rate": 3.6751630112625966e-07,
+      "loss": 0.2114,
+      "step": 4970
+    },
+    {
+      "epoch": 2.9463703703703703,
+      "grad_norm": 2.775742769241333,
+      "learning_rate": 3.635645129421064e-07,
+      "loss": 0.1793,
+      "step": 4971
+    },
+    {
+      "epoch": 2.946962962962963,
+      "grad_norm": 1.5235657691955566,
+      "learning_rate": 3.59612724757953e-07,
+      "loss": 0.1208,
+      "step": 4972
+    },
+    {
+      "epoch": 2.9475555555555557,
+      "grad_norm": 2.0947868824005127,
+      "learning_rate": 3.5566093657379966e-07,
+      "loss": 0.1378,
+      "step": 4973
+    },
+    {
+      "epoch": 2.948148148148148,
+      "grad_norm": 2.187411308288574,
+      "learning_rate": 3.517091483896463e-07,
+      "loss": 0.201,
+      "step": 4974
+    },
+    {
+      "epoch": 2.9487407407407407,
+      "grad_norm": 2.045353651046753,
+      "learning_rate": 3.4775736020549304e-07,
+      "loss": 0.1733,
+      "step": 4975
+    },
+    {
+      "epoch": 2.9493333333333336,
+      "grad_norm": 2.339953660964966,
+      "learning_rate": 3.438055720213397e-07,
+      "loss": 0.1676,
+      "step": 4976
+    },
+    {
+      "epoch": 2.949925925925926,
+      "grad_norm": 1.9999254941940308,
+      "learning_rate": 3.398537838371863e-07,
+      "loss": 0.1165,
+      "step": 4977
+    },
+    {
+      "epoch": 2.9505185185185185,
+      "grad_norm": 2.3185501098632812,
+      "learning_rate": 3.35901995653033e-07,
+      "loss": 0.1764,
+      "step": 4978
+    },
+    {
+      "epoch": 2.951111111111111,
+      "grad_norm": 2.234917163848877,
+      "learning_rate": 3.319502074688797e-07,
+      "loss": 0.1718,
+      "step": 4979
+    },
+    {
+      "epoch": 2.9517037037037035,
+      "grad_norm": 3.2966718673706055,
+      "learning_rate": 3.279984192847264e-07,
+      "loss": 0.1697,
+      "step": 4980
+    },
+    {
+      "epoch": 2.9522962962962964,
+      "grad_norm": 2.0434093475341797,
+      "learning_rate": 3.2404663110057304e-07,
+      "loss": 0.1646,
+      "step": 4981
+    },
+    {
+      "epoch": 2.952888888888889,
+      "grad_norm": 2.081824779510498,
+      "learning_rate": 3.2009484291641965e-07,
+      "loss": 0.1651,
+      "step": 4982
+    },
+    {
+      "epoch": 2.9534814814814814,
+      "grad_norm": 2.755833148956299,
+      "learning_rate": 3.1614305473226637e-07,
+      "loss": 0.1981,
+      "step": 4983
+    },
+    {
+      "epoch": 2.9540740740740743,
+      "grad_norm": 2.079185962677002,
+      "learning_rate": 3.1219126654811304e-07,
+      "loss": 0.0917,
+      "step": 4984
+    },
+    {
+      "epoch": 2.9546666666666668,
+      "grad_norm": 2.503878593444824,
+      "learning_rate": 3.082394783639597e-07,
+      "loss": 0.2027,
+      "step": 4985
+    },
+    {
+      "epoch": 2.9552592592592593,
+      "grad_norm": 3.3734192848205566,
+      "learning_rate": 3.0428769017980637e-07,
+      "loss": 0.2196,
+      "step": 4986
+    },
+    {
+      "epoch": 2.9558518518518517,
+      "grad_norm": 3.4649529457092285,
+      "learning_rate": 3.003359019956531e-07,
+      "loss": 0.2521,
+      "step": 4987
+    },
+    {
+      "epoch": 2.956444444444444,
+      "grad_norm": 5.58460807800293,
+      "learning_rate": 2.963841138114997e-07,
+      "loss": 0.4411,
+      "step": 4988
+    },
+    {
+      "epoch": 2.957037037037037,
+      "grad_norm": 4.901524066925049,
+      "learning_rate": 2.924323256273464e-07,
+      "loss": 0.2957,
+      "step": 4989
+    },
+    {
+      "epoch": 2.9576296296296296,
+      "grad_norm": 2.171149253845215,
+      "learning_rate": 2.884805374431931e-07,
+      "loss": 0.1884,
+      "step": 4990
+    },
+    {
+      "epoch": 2.958222222222222,
+      "grad_norm": 1.9079598188400269,
+      "learning_rate": 2.8452874925903976e-07,
+      "loss": 0.1484,
+      "step": 4991
+    },
+    {
+      "epoch": 2.958814814814815,
+      "grad_norm": 2.5619609355926514,
+      "learning_rate": 2.805769610748864e-07,
+      "loss": 0.169,
+      "step": 4992
+    },
+    {
+      "epoch": 2.9594074074074075,
+      "grad_norm": 1.612339973449707,
+      "learning_rate": 2.766251728907331e-07,
+      "loss": 0.1277,
+      "step": 4993
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.4796741008758545,
+      "learning_rate": 2.7267338470657976e-07,
+      "loss": 0.2009,
+      "step": 4994
+    },
+    {
+      "epoch": 2.9605925925925924,
+      "grad_norm": 1.8868672847747803,
+      "learning_rate": 2.687215965224264e-07,
+      "loss": 0.1239,
+      "step": 4995
+    },
+    {
+      "epoch": 2.961185185185185,
+      "grad_norm": 2.556361198425293,
+      "learning_rate": 2.647698083382731e-07,
+      "loss": 0.183,
+      "step": 4996
+    },
+    {
+      "epoch": 2.961777777777778,
+      "grad_norm": 2.6345226764678955,
+      "learning_rate": 2.6081802015411976e-07,
+      "loss": 0.0985,
+      "step": 4997
+    },
+    {
+      "epoch": 2.9623703703703703,
+      "grad_norm": 2.119414806365967,
+      "learning_rate": 2.568662319699664e-07,
+      "loss": 0.1229,
+      "step": 4998
+    },
+    {
+      "epoch": 2.962962962962963,
+      "grad_norm": 2.3750216960906982,
+      "learning_rate": 2.529144437858131e-07,
+      "loss": 0.2238,
+      "step": 4999
+    },
+    {
+      "epoch": 2.9635555555555557,
+      "grad_norm": 1.6443592309951782,
+      "learning_rate": 2.4896265560165975e-07,
+      "loss": 0.1284,
+      "step": 5000
+    },
+    {
+      "epoch": 2.964148148148148,
+      "grad_norm": 1.7027292251586914,
+      "learning_rate": 2.450108674175064e-07,
+      "loss": 0.1309,
+      "step": 5001
+    },
+    {
+      "epoch": 2.9647407407407407,
+      "grad_norm": 2.152165174484253,
+      "learning_rate": 2.4105907923335314e-07,
+      "loss": 0.2158,
+      "step": 5002
+    },
+    {
+      "epoch": 2.9653333333333336,
+      "grad_norm": 1.7640650272369385,
+      "learning_rate": 2.3710729104919978e-07,
+      "loss": 0.1304,
+      "step": 5003
+    },
+    {
+      "epoch": 2.965925925925926,
+      "grad_norm": 2.916719675064087,
+      "learning_rate": 2.3315550286504645e-07,
+      "loss": 0.2205,
+      "step": 5004
+    },
+    {
+      "epoch": 2.9665185185185186,
+      "grad_norm": 1.984520435333252,
+      "learning_rate": 2.292037146808931e-07,
+      "loss": 0.1607,
+      "step": 5005
+    },
+    {
+      "epoch": 2.967111111111111,
+      "grad_norm": 2.470721960067749,
+      "learning_rate": 2.2525192649673978e-07,
+      "loss": 0.178,
+      "step": 5006
+    },
+    {
+      "epoch": 2.9677037037037035,
+      "grad_norm": 1.961169958114624,
+      "learning_rate": 2.2130013831258647e-07,
+      "loss": 0.1386,
+      "step": 5007
+    },
+    {
+      "epoch": 2.9682962962962964,
+      "grad_norm": 2.50612473487854,
+      "learning_rate": 2.173483501284331e-07,
+      "loss": 0.2009,
+      "step": 5008
+    },
+    {
+      "epoch": 2.968888888888889,
+      "grad_norm": 2.6785099506378174,
+      "learning_rate": 2.133965619442798e-07,
+      "loss": 0.1767,
+      "step": 5009
+    },
+    {
+      "epoch": 2.9694814814814814,
+      "grad_norm": 2.2866294384002686,
+      "learning_rate": 2.0944477376012647e-07,
+      "loss": 0.1726,
+      "step": 5010
+    },
+    {
+      "epoch": 2.9700740740740743,
+      "grad_norm": 2.617375612258911,
+      "learning_rate": 2.0549298557597314e-07,
+      "loss": 0.1873,
+      "step": 5011
+    },
+    {
+      "epoch": 2.970666666666667,
+      "grad_norm": 1.5300716161727905,
+      "learning_rate": 2.0154119739181983e-07,
+      "loss": 0.118,
+      "step": 5012
+    },
+    {
+      "epoch": 2.9712592592592593,
+      "grad_norm": 2.5218632221221924,
+      "learning_rate": 1.9758940920766647e-07,
+      "loss": 0.2053,
+      "step": 5013
+    },
+    {
+      "epoch": 2.9718518518518517,
+      "grad_norm": 2.885709047317505,
+      "learning_rate": 1.9363762102351316e-07,
+      "loss": 0.136,
+      "step": 5014
+    },
+    {
+      "epoch": 2.9724444444444442,
+      "grad_norm": 1.7124849557876587,
+      "learning_rate": 1.896858328393598e-07,
+      "loss": 0.1362,
+      "step": 5015
+    },
+    {
+      "epoch": 2.973037037037037,
+      "grad_norm": 2.0018503665924072,
+      "learning_rate": 1.857340446552065e-07,
+      "loss": 0.1505,
+      "step": 5016
+    },
+    {
+      "epoch": 2.9736296296296296,
+      "grad_norm": 2.323570489883423,
+      "learning_rate": 1.817822564710532e-07,
+      "loss": 0.1156,
+      "step": 5017
+    },
+    {
+      "epoch": 2.974222222222222,
+      "grad_norm": 2.0642917156219482,
+      "learning_rate": 1.7783046828689983e-07,
+      "loss": 0.1796,
+      "step": 5018
+    },
+    {
+      "epoch": 2.974814814814815,
+      "grad_norm": 1.4094698429107666,
+      "learning_rate": 1.7387868010274652e-07,
+      "loss": 0.1159,
+      "step": 5019
+    },
+    {
+      "epoch": 2.9754074074074075,
+      "grad_norm": 2.5120468139648438,
+      "learning_rate": 1.6992689191859316e-07,
+      "loss": 0.1779,
+      "step": 5020
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 2.33035945892334,
+      "learning_rate": 1.6597510373443985e-07,
+      "loss": 0.3148,
+      "step": 5021
+    },
+    {
+      "epoch": 2.9765925925925925,
+      "grad_norm": 1.9817825555801392,
+      "learning_rate": 1.6202331555028652e-07,
+      "loss": 0.1192,
+      "step": 5022
+    },
+    {
+      "epoch": 2.977185185185185,
+      "grad_norm": 2.4603140354156494,
+      "learning_rate": 1.5807152736613319e-07,
+      "loss": 0.1868,
+      "step": 5023
+    },
+    {
+      "epoch": 2.977777777777778,
+      "grad_norm": 2.8225488662719727,
+      "learning_rate": 1.5411973918197985e-07,
+      "loss": 0.1912,
+      "step": 5024
+    },
+    {
+      "epoch": 2.9783703703703703,
+      "grad_norm": 2.1874630451202393,
+      "learning_rate": 1.5016795099782655e-07,
+      "loss": 0.163,
+      "step": 5025
+    },
+    {
+      "epoch": 2.978962962962963,
+      "grad_norm": 2.8842992782592773,
+      "learning_rate": 1.462161628136732e-07,
+      "loss": 0.1851,
+      "step": 5026
+    },
+    {
+      "epoch": 2.9795555555555557,
+      "grad_norm": 2.3311190605163574,
+      "learning_rate": 1.4226437462951988e-07,
+      "loss": 0.2171,
+      "step": 5027
+    },
+    {
+      "epoch": 2.980148148148148,
+      "grad_norm": 2.6044604778289795,
+      "learning_rate": 1.3831258644536654e-07,
+      "loss": 0.1453,
+      "step": 5028
+    },
+    {
+      "epoch": 2.9807407407407407,
+      "grad_norm": 2.49471116065979,
+      "learning_rate": 1.343607982612132e-07,
+      "loss": 0.1968,
+      "step": 5029
+    },
+    {
+      "epoch": 2.981333333333333,
+      "grad_norm": 1.9333173036575317,
+      "learning_rate": 1.3040901007705988e-07,
+      "loss": 0.1538,
+      "step": 5030
+    },
+    {
+      "epoch": 2.981925925925926,
+      "grad_norm": 3.000826835632324,
+      "learning_rate": 1.2645722189290654e-07,
+      "loss": 0.1967,
+      "step": 5031
+    },
+    {
+      "epoch": 2.9825185185185186,
+      "grad_norm": 2.2996904850006104,
+      "learning_rate": 1.225054337087532e-07,
+      "loss": 0.1394,
+      "step": 5032
+    },
+    {
+      "epoch": 2.983111111111111,
+      "grad_norm": 1.7709780931472778,
+      "learning_rate": 1.1855364552459989e-07,
+      "loss": 0.1352,
+      "step": 5033
+    },
+    {
+      "epoch": 2.9837037037037035,
+      "grad_norm": 1.8945822715759277,
+      "learning_rate": 1.1460185734044656e-07,
+      "loss": 0.1804,
+      "step": 5034
+    },
+    {
+      "epoch": 2.9842962962962964,
+      "grad_norm": 3.5741419792175293,
+      "learning_rate": 1.1065006915629324e-07,
+      "loss": 0.2047,
+      "step": 5035
+    },
+    {
+      "epoch": 2.984888888888889,
+      "grad_norm": 1.8715639114379883,
+      "learning_rate": 1.066982809721399e-07,
+      "loss": 0.1151,
+      "step": 5036
+    },
+    {
+      "epoch": 2.9854814814814814,
+      "grad_norm": 2.074307918548584,
+      "learning_rate": 1.0274649278798657e-07,
+      "loss": 0.1377,
+      "step": 5037
+    },
+    {
+      "epoch": 2.9860740740740743,
+      "grad_norm": 2.600599527359009,
+      "learning_rate": 9.879470460383324e-08,
+      "loss": 0.2719,
+      "step": 5038
+    },
+    {
+      "epoch": 2.986666666666667,
+      "grad_norm": 2.4986789226531982,
+      "learning_rate": 9.48429164196799e-08,
+      "loss": 0.1969,
+      "step": 5039
+    },
+    {
+      "epoch": 2.9872592592592593,
+      "grad_norm": 2.8238372802734375,
+      "learning_rate": 9.08911282355266e-08,
+      "loss": 0.1647,
+      "step": 5040
+    },
+    {
+      "epoch": 2.9878518518518518,
+      "grad_norm": 2.0154948234558105,
+      "learning_rate": 8.693934005137326e-08,
+      "loss": 0.1147,
+      "step": 5041
+    },
+    {
+      "epoch": 2.9884444444444442,
+      "grad_norm": 2.267366409301758,
+      "learning_rate": 8.298755186721993e-08,
+      "loss": 0.1947,
+      "step": 5042
+    },
+    {
+      "epoch": 2.989037037037037,
+      "grad_norm": 1.9071255922317505,
+      "learning_rate": 7.903576368306659e-08,
+      "loss": 0.1301,
+      "step": 5043
+    },
+    {
+      "epoch": 2.9896296296296296,
+      "grad_norm": 2.2883541584014893,
+      "learning_rate": 7.508397549891327e-08,
+      "loss": 0.2501,
+      "step": 5044
+    },
+    {
+      "epoch": 2.990222222222222,
+      "grad_norm": 1.8338241577148438,
+      "learning_rate": 7.113218731475994e-08,
+      "loss": 0.1799,
+      "step": 5045
+    },
+    {
+      "epoch": 2.990814814814815,
+      "grad_norm": 1.5573910474777222,
+      "learning_rate": 6.71803991306066e-08,
+      "loss": 0.2498,
+      "step": 5046
+    },
+    {
+      "epoch": 2.9914074074074075,
+      "grad_norm": 1.9234817028045654,
+      "learning_rate": 6.322861094645327e-08,
+      "loss": 0.0689,
+      "step": 5047
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 1.8653851747512817,
+      "learning_rate": 5.9276822762299945e-08,
+      "loss": 0.2013,
+      "step": 5048
+    },
+    {
+      "epoch": 2.9925925925925925,
+      "grad_norm": 1.896306037902832,
+      "learning_rate": 5.532503457814662e-08,
+      "loss": 0.171,
+      "step": 5049
+    },
+    {
+      "epoch": 2.993185185185185,
+      "grad_norm": 1.6657828092575073,
+      "learning_rate": 5.1373246393993284e-08,
+      "loss": 0.1378,
+      "step": 5050
+    },
+    {
+      "epoch": 2.993777777777778,
+      "grad_norm": 2.8170599937438965,
+      "learning_rate": 4.742145820983995e-08,
+      "loss": 0.1845,
+      "step": 5051
+    },
+    {
+      "epoch": 2.9943703703703703,
+      "grad_norm": 1.8336315155029297,
+      "learning_rate": 4.346967002568663e-08,
+      "loss": 0.155,
+      "step": 5052
+    },
+    {
+      "epoch": 2.994962962962963,
+      "grad_norm": 2.8716914653778076,
+      "learning_rate": 3.9517881841533297e-08,
+      "loss": 0.1675,
+      "step": 5053
+    },
+    {
+      "epoch": 2.9955555555555557,
+      "grad_norm": 2.1118741035461426,
+      "learning_rate": 3.556609365737997e-08,
+      "loss": 0.1489,
+      "step": 5054
+    },
+    {
+      "epoch": 2.9961481481481482,
+      "grad_norm": 2.355520009994507,
+      "learning_rate": 3.1614305473226636e-08,
+      "loss": 0.1239,
+      "step": 5055
+    },
+    {
+      "epoch": 2.9967407407407407,
+      "grad_norm": 1.7561947107315063,
+      "learning_rate": 2.766251728907331e-08,
+      "loss": 0.1454,
+      "step": 5056
+    },
+    {
+      "epoch": 2.997333333333333,
+      "grad_norm": 2.4879305362701416,
+      "learning_rate": 2.3710729104919975e-08,
+      "loss": 0.1153,
+      "step": 5057
+    },
+    {
+      "epoch": 2.9979259259259257,
+      "grad_norm": 2.6663336753845215,
+      "learning_rate": 1.9758940920766648e-08,
+      "loss": 0.1208,
+      "step": 5058
+    },
+    {
+      "epoch": 2.9985185185185186,
+      "grad_norm": 1.76002037525177,
+      "learning_rate": 1.5807152736613318e-08,
+      "loss": 0.1803,
+      "step": 5059
+    },
+    {
+      "epoch": 2.999111111111111,
+      "grad_norm": 2.402899742126465,
+      "learning_rate": 1.1855364552459988e-08,
+      "loss": 0.1237,
+      "step": 5060
+    },
+    {
+      "epoch": 2.9997037037037035,
+      "grad_norm": 2.013993740081787,
+      "learning_rate": 7.903576368306659e-09,
+      "loss": 0.142,
+      "step": 5061
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5061,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.4207814522463846e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}