diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4230 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 2976,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006720430107526882,
+      "grad_norm": 48.78054272427667,
+      "learning_rate": 6.711409395973155e-08,
+      "loss": 2.3825,
+      "step": 1
+    },
+    {
+      "epoch": 0.003360215053763441,
+      "grad_norm": 48.47840626370759,
+      "learning_rate": 3.3557046979865777e-07,
+      "loss": 2.376,
+      "step": 5
+    },
+    {
+      "epoch": 0.006720430107526882,
+      "grad_norm": 51.80697175894909,
+      "learning_rate": 6.711409395973155e-07,
+      "loss": 2.3226,
+      "step": 10
+    },
+    {
+      "epoch": 0.010080645161290322,
+      "grad_norm": 13.624773264265663,
+      "learning_rate": 1.006711409395973e-06,
+      "loss": 2.1053,
+      "step": 15
+    },
+    {
+      "epoch": 0.013440860215053764,
+      "grad_norm": 6.262298470265524,
+      "learning_rate": 1.342281879194631e-06,
+      "loss": 1.9272,
+      "step": 20
+    },
+    {
+      "epoch": 0.016801075268817203,
+      "grad_norm": 7.548559183569387,
+      "learning_rate": 1.6778523489932889e-06,
+      "loss": 1.7186,
+      "step": 25
+    },
+    {
+      "epoch": 0.020161290322580645,
+      "grad_norm": 2.6440968174788626,
+      "learning_rate": 2.013422818791946e-06,
+      "loss": 1.5223,
+      "step": 30
+    },
+    {
+      "epoch": 0.023521505376344086,
+      "grad_norm": 3.58492629574647,
+      "learning_rate": 2.348993288590604e-06,
+      "loss": 1.3232,
+      "step": 35
+    },
+    {
+      "epoch": 0.026881720430107527,
+      "grad_norm": 1.6886536264742815,
+      "learning_rate": 2.684563758389262e-06,
+      "loss": 1.1928,
+      "step": 40
+    },
+    {
+      "epoch": 0.03024193548387097,
+      "grad_norm": 1.217304086312586,
+      "learning_rate": 3.02013422818792e-06,
+      "loss": 1.0767,
+      "step": 45
+    },
+    {
+      "epoch": 0.033602150537634407,
+      "grad_norm": 1.1946964624516874,
+      "learning_rate": 3.3557046979865777e-06,
+      "loss": 1.0086,
+      "step": 50
+    },
+    {
+      "epoch": 0.03696236559139785,
+      "grad_norm": 1.3676335716854129,
+      "learning_rate": 3.6912751677852355e-06,
+      "loss": 0.9521,
+      "step": 55
+    },
+    {
+      "epoch": 0.04032258064516129,
+      "grad_norm": 1.15160172673339,
+      "learning_rate": 4.026845637583892e-06,
+      "loss": 0.9144,
+      "step": 60
+    },
+    {
+      "epoch": 0.043682795698924734,
+      "grad_norm": 1.1898281923774547,
+      "learning_rate": 4.362416107382551e-06,
+      "loss": 0.8891,
+      "step": 65
+    },
+    {
+      "epoch": 0.04704301075268817,
+      "grad_norm": 1.2290162400825384,
+      "learning_rate": 4.697986577181208e-06,
+      "loss": 0.8684,
+      "step": 70
+    },
+    {
+      "epoch": 0.05040322580645161,
+      "grad_norm": 1.1047127902421139,
+      "learning_rate": 5.033557046979867e-06,
+      "loss": 0.852,
+      "step": 75
+    },
+    {
+      "epoch": 0.053763440860215055,
+      "grad_norm": 1.3842650719759613,
+      "learning_rate": 5.369127516778524e-06,
+      "loss": 0.8421,
+      "step": 80
+    },
+    {
+      "epoch": 0.05712365591397849,
+      "grad_norm": 1.1630054678494466,
+      "learning_rate": 5.704697986577181e-06,
+      "loss": 0.8293,
+      "step": 85
+    },
+    {
+      "epoch": 0.06048387096774194,
+      "grad_norm": 1.2895915993551423,
+      "learning_rate": 6.04026845637584e-06,
+      "loss": 0.8155,
+      "step": 90
+    },
+    {
+      "epoch": 0.06384408602150538,
+      "grad_norm": 1.3357187881014583,
+      "learning_rate": 6.375838926174497e-06,
+      "loss": 0.8094,
+      "step": 95
+    },
+    {
+      "epoch": 0.06720430107526881,
+      "grad_norm": 1.4294338181391608,
+      "learning_rate": 6.711409395973155e-06,
+      "loss": 0.807,
+      "step": 100
+    },
+    {
+      "epoch": 0.07056451612903226,
+      "grad_norm": 0.9685428191519895,
+      "learning_rate": 7.046979865771813e-06,
+      "loss": 0.7963,
+      "step": 105
+    },
+    {
+      "epoch": 0.0739247311827957,
+      "grad_norm": 1.0685231337896743,
+      "learning_rate": 7.382550335570471e-06,
+      "loss": 0.7903,
+      "step": 110
+    },
+    {
+      "epoch": 0.07728494623655914,
+      "grad_norm": 1.3129982717287054,
+      "learning_rate": 7.718120805369127e-06,
+      "loss": 0.7932,
+      "step": 115
+    },
+    {
+      "epoch": 0.08064516129032258,
+      "grad_norm": 1.6648712375867527,
+      "learning_rate": 8.053691275167785e-06,
+      "loss": 0.7865,
+      "step": 120
+    },
+    {
+      "epoch": 0.08400537634408602,
+      "grad_norm": 1.2705854884448164,
+      "learning_rate": 8.389261744966444e-06,
+      "loss": 0.7799,
+      "step": 125
+    },
+    {
+      "epoch": 0.08736559139784947,
+      "grad_norm": 1.3845304392408557,
+      "learning_rate": 8.724832214765101e-06,
+      "loss": 0.7794,
+      "step": 130
+    },
+    {
+      "epoch": 0.0907258064516129,
+      "grad_norm": 1.0490487686523224,
+      "learning_rate": 9.060402684563759e-06,
+      "loss": 0.7762,
+      "step": 135
+    },
+    {
+      "epoch": 0.09408602150537634,
+      "grad_norm": 1.1349198286002737,
+      "learning_rate": 9.395973154362416e-06,
+      "loss": 0.7704,
+      "step": 140
+    },
+    {
+      "epoch": 0.09744623655913978,
+      "grad_norm": 1.311025567661312,
+      "learning_rate": 9.731543624161075e-06,
+      "loss": 0.7624,
+      "step": 145
+    },
+    {
+      "epoch": 0.10080645161290322,
+      "grad_norm": 1.0240371131291772,
+      "learning_rate": 1.0067114093959734e-05,
+      "loss": 0.7645,
+      "step": 150
+    },
+    {
+      "epoch": 0.10416666666666667,
+      "grad_norm": 1.0576784893797677,
+      "learning_rate": 1.040268456375839e-05,
+      "loss": 0.7612,
+      "step": 155
+    },
+    {
+      "epoch": 0.10752688172043011,
+      "grad_norm": 0.9872090722303354,
+      "learning_rate": 1.0738255033557049e-05,
+      "loss": 0.762,
+      "step": 160
+    },
+    {
+      "epoch": 0.11088709677419355,
+      "grad_norm": 1.1448278866865749,
+      "learning_rate": 1.1073825503355706e-05,
+      "loss": 0.7595,
+      "step": 165
+    },
+    {
+      "epoch": 0.11424731182795698,
+      "grad_norm": 0.9099648734864951,
+      "learning_rate": 1.1409395973154362e-05,
+      "loss": 0.7545,
+      "step": 170
+    },
+    {
+      "epoch": 0.11760752688172044,
+      "grad_norm": 1.2385558169257114,
+      "learning_rate": 1.174496644295302e-05,
+      "loss": 0.7559,
+      "step": 175
+    },
+    {
+      "epoch": 0.12096774193548387,
+      "grad_norm": 0.9417818627828941,
+      "learning_rate": 1.208053691275168e-05,
+      "loss": 0.7533,
+      "step": 180
+    },
+    {
+      "epoch": 0.12432795698924731,
+      "grad_norm": 1.4888584917162164,
+      "learning_rate": 1.2416107382550337e-05,
+      "loss": 0.75,
+      "step": 185
+    },
+    {
+      "epoch": 0.12768817204301075,
+      "grad_norm": 1.6094338507124168,
+      "learning_rate": 1.2751677852348994e-05,
+      "loss": 0.7516,
+      "step": 190
+    },
+    {
+      "epoch": 0.1310483870967742,
+      "grad_norm": 1.1484532669834084,
+      "learning_rate": 1.3087248322147652e-05,
+      "loss": 0.7528,
+      "step": 195
+    },
+    {
+      "epoch": 0.13440860215053763,
+      "grad_norm": 0.9606481759674919,
+      "learning_rate": 1.342281879194631e-05,
+      "loss": 0.7432,
+      "step": 200
+    },
+    {
+      "epoch": 0.13776881720430106,
+      "grad_norm": 0.90666325947728,
+      "learning_rate": 1.3758389261744966e-05,
+      "loss": 0.7456,
+      "step": 205
+    },
+    {
+      "epoch": 0.14112903225806453,
+      "grad_norm": 0.9003130723406445,
+      "learning_rate": 1.4093959731543626e-05,
+      "loss": 0.741,
+      "step": 210
+    },
+    {
+      "epoch": 0.14448924731182797,
+      "grad_norm": 0.9046243900141376,
+      "learning_rate": 1.4429530201342283e-05,
+      "loss": 0.7413,
+      "step": 215
+    },
+    {
+      "epoch": 0.1478494623655914,
+      "grad_norm": 0.9495916559973592,
+      "learning_rate": 1.4765100671140942e-05,
+      "loss": 0.7396,
+      "step": 220
+    },
+    {
+      "epoch": 0.15120967741935484,
+      "grad_norm": 0.8392501968484852,
+      "learning_rate": 1.5100671140939598e-05,
+      "loss": 0.7451,
+      "step": 225
+    },
+    {
+      "epoch": 0.15456989247311828,
+      "grad_norm": 1.0111465864476745,
+      "learning_rate": 1.5436241610738255e-05,
+      "loss": 0.7434,
+      "step": 230
+    },
+    {
+      "epoch": 0.15793010752688172,
+      "grad_norm": 0.9025884088477472,
+      "learning_rate": 1.5771812080536916e-05,
+      "loss": 0.7354,
+      "step": 235
+    },
+    {
+      "epoch": 0.16129032258064516,
+      "grad_norm": 0.9720296761608475,
+      "learning_rate": 1.610738255033557e-05,
+      "loss": 0.7371,
+      "step": 240
+    },
+    {
+      "epoch": 0.1646505376344086,
+      "grad_norm": 0.9947001241550435,
+      "learning_rate": 1.644295302013423e-05,
+      "loss": 0.7288,
+      "step": 245
+    },
+    {
+      "epoch": 0.16801075268817203,
+      "grad_norm": 0.9174436365138352,
+      "learning_rate": 1.6778523489932888e-05,
+      "loss": 0.7347,
+      "step": 250
+    },
+    {
+      "epoch": 0.17137096774193547,
+      "grad_norm": 1.0267733434401298,
+      "learning_rate": 1.7114093959731545e-05,
+      "loss": 0.7375,
+      "step": 255
+    },
+    {
+      "epoch": 0.17473118279569894,
+      "grad_norm": 0.8397895938406165,
+      "learning_rate": 1.7449664429530202e-05,
+      "loss": 0.7341,
+      "step": 260
+    },
+    {
+      "epoch": 0.17809139784946237,
+      "grad_norm": 0.76727099531327,
+      "learning_rate": 1.778523489932886e-05,
+      "loss": 0.7309,
+      "step": 265
+    },
+    {
+      "epoch": 0.1814516129032258,
+      "grad_norm": 0.7412065721447615,
+      "learning_rate": 1.8120805369127517e-05,
+      "loss": 0.7269,
+      "step": 270
+    },
+    {
+      "epoch": 0.18481182795698925,
+      "grad_norm": 0.7810498280270748,
+      "learning_rate": 1.8456375838926174e-05,
+      "loss": 0.7257,
+      "step": 275
+    },
+    {
+      "epoch": 0.1881720430107527,
+      "grad_norm": 0.8613709749529885,
+      "learning_rate": 1.8791946308724832e-05,
+      "loss": 0.7315,
+      "step": 280
+    },
+    {
+      "epoch": 0.19153225806451613,
+      "grad_norm": 0.9739010230796492,
+      "learning_rate": 1.9127516778523493e-05,
+      "loss": 0.7272,
+      "step": 285
+    },
+    {
+      "epoch": 0.19489247311827956,
+      "grad_norm": 0.8646191524355271,
+      "learning_rate": 1.946308724832215e-05,
+      "loss": 0.7289,
+      "step": 290
+    },
+    {
+      "epoch": 0.198252688172043,
+      "grad_norm": 0.9363534800514449,
+      "learning_rate": 1.9798657718120807e-05,
+      "loss": 0.7258,
+      "step": 295
+    },
+    {
+      "epoch": 0.20161290322580644,
+      "grad_norm": 0.9701273350687407,
+      "learning_rate": 1.9999972476199807e-05,
+      "loss": 0.7317,
+      "step": 300
+    },
+    {
+      "epoch": 0.2049731182795699,
+      "grad_norm": 0.7773609368388674,
+      "learning_rate": 1.999966283518764e-05,
+      "loss": 0.7227,
+      "step": 305
+    },
+    {
+      "epoch": 0.20833333333333334,
+      "grad_norm": 0.8956665591253581,
+      "learning_rate": 1.999900915910167e-05,
+      "loss": 0.7231,
+      "step": 310
+    },
+    {
+      "epoch": 0.21169354838709678,
+      "grad_norm": 0.6269021652498205,
+      "learning_rate": 1.99980114704314e-05,
+      "loss": 0.7216,
+      "step": 315
+    },
+    {
+      "epoch": 0.21505376344086022,
+      "grad_norm": 1.011329580641999,
+      "learning_rate": 1.9996669803501982e-05,
+      "loss": 0.7276,
+      "step": 320
+    },
+    {
+      "epoch": 0.21841397849462366,
+      "grad_norm": 1.080429434985579,
+      "learning_rate": 1.999498420447302e-05,
+      "loss": 0.7236,
+      "step": 325
+    },
+    {
+      "epoch": 0.2217741935483871,
+      "grad_norm": 0.8583297562008422,
+      "learning_rate": 1.9992954731336992e-05,
+      "loss": 0.7225,
+      "step": 330
+    },
+    {
+      "epoch": 0.22513440860215053,
+      "grad_norm": 0.8080172656804933,
+      "learning_rate": 1.9990581453917234e-05,
+      "loss": 0.7209,
+      "step": 335
+    },
+    {
+      "epoch": 0.22849462365591397,
+      "grad_norm": 0.8049987876429506,
+      "learning_rate": 1.998786445386558e-05,
+      "loss": 0.7178,
+      "step": 340
+    },
+    {
+      "epoch": 0.2318548387096774,
+      "grad_norm": 0.7233007018856564,
+      "learning_rate": 1.9984803824659504e-05,
+      "loss": 0.7181,
+      "step": 345
+    },
+    {
+      "epoch": 0.23521505376344087,
+      "grad_norm": 0.6435752626797113,
+      "learning_rate": 1.998139967159894e-05,
+      "loss": 0.7124,
+      "step": 350
+    },
+    {
+      "epoch": 0.2385752688172043,
+      "grad_norm": 0.6988189980904659,
+      "learning_rate": 1.997765211180264e-05,
+      "loss": 0.7138,
+      "step": 355
+    },
+    {
+      "epoch": 0.24193548387096775,
+      "grad_norm": 0.8449192477747994,
+      "learning_rate": 1.9973561274204153e-05,
+      "loss": 0.7164,
+      "step": 360
+    },
+    {
+      "epoch": 0.2452956989247312,
+      "grad_norm": 0.6199401123601755,
+      "learning_rate": 1.9969127299547387e-05,
+      "loss": 0.711,
+      "step": 365
+    },
+    {
+      "epoch": 0.24865591397849462,
+      "grad_norm": 0.6508335116554123,
+      "learning_rate": 1.9964350340381763e-05,
+      "loss": 0.7099,
+      "step": 370
+    },
+    {
+      "epoch": 0.25201612903225806,
+      "grad_norm": 0.7240178041828755,
+      "learning_rate": 1.995923056105697e-05,
+      "loss": 0.7108,
+      "step": 375
+    },
+    {
+      "epoch": 0.2553763440860215,
+      "grad_norm": 0.7146640916604361,
+      "learning_rate": 1.9953768137717323e-05,
+      "loss": 0.714,
+      "step": 380
+    },
+    {
+      "epoch": 0.25873655913978494,
+      "grad_norm": 0.7235799398485985,
+      "learning_rate": 1.9947963258295665e-05,
+      "loss": 0.7141,
+      "step": 385
+    },
+    {
+      "epoch": 0.2620967741935484,
+      "grad_norm": 0.7085278247346875,
+      "learning_rate": 1.9941816122506958e-05,
+      "loss": 0.7123,
+      "step": 390
+    },
+    {
+      "epoch": 0.2654569892473118,
+      "grad_norm": 0.8336879205562984,
+      "learning_rate": 1.9935326941841348e-05,
+      "loss": 0.7118,
+      "step": 395
+    },
+    {
+      "epoch": 0.26881720430107525,
+      "grad_norm": 0.6472707436647601,
+      "learning_rate": 1.9928495939556952e-05,
+      "loss": 0.7112,
+      "step": 400
+    },
+    {
+      "epoch": 0.2721774193548387,
+      "grad_norm": 0.5404676396301624,
+      "learning_rate": 1.9921323350672123e-05,
+      "loss": 0.7023,
+      "step": 405
+    },
+    {
+      "epoch": 0.27553763440860213,
+      "grad_norm": 0.5867567569524356,
+      "learning_rate": 1.9913809421957395e-05,
+      "loss": 0.7061,
+      "step": 410
+    },
+    {
+      "epoch": 0.27889784946236557,
+      "grad_norm": 0.5310275193870236,
+      "learning_rate": 1.9905954411926992e-05,
+      "loss": 0.7097,
+      "step": 415
+    },
+    {
+      "epoch": 0.28225806451612906,
+      "grad_norm": 0.5428907392759953,
+      "learning_rate": 1.9897758590829915e-05,
+      "loss": 0.7046,
+      "step": 420
+    },
+    {
+      "epoch": 0.2856182795698925,
+      "grad_norm": 0.6360587388135657,
+      "learning_rate": 1.988922224064067e-05,
+      "loss": 0.7059,
+      "step": 425
+    },
+    {
+      "epoch": 0.28897849462365593,
+      "grad_norm": 0.5538060039668187,
+      "learning_rate": 1.988034565504954e-05,
+      "loss": 0.7055,
+      "step": 430
+    },
+    {
+      "epoch": 0.2923387096774194,
+      "grad_norm": 0.7074777520721658,
+      "learning_rate": 1.9871129139452502e-05,
+      "loss": 0.7023,
+      "step": 435
+    },
+    {
+      "epoch": 0.2956989247311828,
+      "grad_norm": 0.6033039358656708,
+      "learning_rate": 1.986157301094071e-05,
+      "loss": 0.702,
+      "step": 440
+    },
+    {
+      "epoch": 0.29905913978494625,
+      "grad_norm": 0.7020301366143326,
+      "learning_rate": 1.9851677598289588e-05,
+      "loss": 0.7072,
+      "step": 445
+    },
+    {
+      "epoch": 0.3024193548387097,
+      "grad_norm": 0.6735214131146494,
+      "learning_rate": 1.9841443241947515e-05,
+      "loss": 0.7046,
+      "step": 450
+    },
+    {
+      "epoch": 0.3057795698924731,
+      "grad_norm": 0.5846703075665397,
+      "learning_rate": 1.983087029402411e-05,
+      "loss": 0.7004,
+      "step": 455
+    },
+    {
+      "epoch": 0.30913978494623656,
+      "grad_norm": 0.558983961781733,
+      "learning_rate": 1.9819959118278144e-05,
+      "loss": 0.7046,
+      "step": 460
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.6880041036412101,
+      "learning_rate": 1.9808710090104983e-05,
+      "loss": 0.7037,
+      "step": 465
+    },
+    {
+      "epoch": 0.31586021505376344,
+      "grad_norm": 0.622425101525314,
+      "learning_rate": 1.9797123596523692e-05,
+      "loss": 0.6984,
+      "step": 470
+    },
+    {
+      "epoch": 0.3192204301075269,
+      "grad_norm": 0.6341081550172605,
+      "learning_rate": 1.978520003616374e-05,
+      "loss": 0.6985,
+      "step": 475
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 0.659287705673367,
+      "learning_rate": 1.9772939819251247e-05,
+      "loss": 0.6954,
+      "step": 480
+    },
+    {
+      "epoch": 0.32594086021505375,
+      "grad_norm": 0.5715874934602246,
+      "learning_rate": 1.976034336759491e-05,
+      "loss": 0.6988,
+      "step": 485
+    },
+    {
+      "epoch": 0.3293010752688172,
+      "grad_norm": 0.8050941273557731,
+      "learning_rate": 1.9747411114571445e-05,
+      "loss": 0.6993,
+      "step": 490
+    },
+    {
+      "epoch": 0.3326612903225806,
+      "grad_norm": 0.6666660914014055,
+      "learning_rate": 1.973414350511072e-05,
+      "loss": 0.7018,
+      "step": 495
+    },
+    {
+      "epoch": 0.33602150537634407,
+      "grad_norm": 0.5898448667662453,
+      "learning_rate": 1.9720540995680428e-05,
+      "loss": 0.697,
+      "step": 500
+    },
+    {
+      "epoch": 0.3393817204301075,
+      "grad_norm": 0.7204191081412994,
+      "learning_rate": 1.9706604054270376e-05,
+      "loss": 0.6966,
+      "step": 505
+    },
+    {
+      "epoch": 0.34274193548387094,
+      "grad_norm": 0.6353645462526023,
+      "learning_rate": 1.9692333160376407e-05,
+      "loss": 0.693,
+      "step": 510
+    },
+    {
+      "epoch": 0.34610215053763443,
+      "grad_norm": 0.5663739741433981,
+      "learning_rate": 1.967772880498387e-05,
+      "loss": 0.6985,
+      "step": 515
+    },
+    {
+      "epoch": 0.34946236559139787,
+      "grad_norm": 0.541549492600514,
+      "learning_rate": 1.9662791490550755e-05,
+      "loss": 0.6933,
+      "step": 520
+    },
+    {
+      "epoch": 0.3528225806451613,
+      "grad_norm": 0.5766179836769963,
+      "learning_rate": 1.9647521730990408e-05,
+      "loss": 0.6996,
+      "step": 525
+    },
+    {
+      "epoch": 0.35618279569892475,
+      "grad_norm": 0.528070730439876,
+      "learning_rate": 1.9631920051653813e-05,
+      "loss": 0.6967,
+      "step": 530
+    },
+    {
+      "epoch": 0.3595430107526882,
+      "grad_norm": 0.590209921037558,
+      "learning_rate": 1.9615986989311567e-05,
+      "loss": 0.6956,
+      "step": 535
+    },
+    {
+      "epoch": 0.3629032258064516,
+      "grad_norm": 0.5291818662889545,
+      "learning_rate": 1.9599723092135376e-05,
+      "loss": 0.695,
+      "step": 540
+    },
+    {
+      "epoch": 0.36626344086021506,
+      "grad_norm": 0.4970212649639191,
+      "learning_rate": 1.9583128919679218e-05,
+      "loss": 0.6917,
+      "step": 545
+    },
+    {
+      "epoch": 0.3696236559139785,
+      "grad_norm": 0.6225127520386264,
+      "learning_rate": 1.956620504286007e-05,
+      "loss": 0.6935,
+      "step": 550
+    },
+    {
+      "epoch": 0.37298387096774194,
+      "grad_norm": 0.6214865530844041,
+      "learning_rate": 1.9548952043938286e-05,
+      "loss": 0.6894,
+      "step": 555
+    },
+    {
+      "epoch": 0.3763440860215054,
+      "grad_norm": 0.6009400925112226,
+      "learning_rate": 1.9531370516497562e-05,
+      "loss": 0.6888,
+      "step": 560
+    },
+    {
+      "epoch": 0.3797043010752688,
+      "grad_norm": 0.6054051937463671,
+      "learning_rate": 1.95134610654245e-05,
+      "loss": 0.6904,
+      "step": 565
+    },
+    {
+      "epoch": 0.38306451612903225,
+      "grad_norm": 0.5850716552121316,
+      "learning_rate": 1.9495224306887797e-05,
+      "loss": 0.6924,
+      "step": 570
+    },
+    {
+      "epoch": 0.3864247311827957,
+      "grad_norm": 0.5796507375628339,
+      "learning_rate": 1.9476660868317076e-05,
+      "loss": 0.69,
+      "step": 575
+    },
+    {
+      "epoch": 0.3897849462365591,
+      "grad_norm": 0.72632335015395,
+      "learning_rate": 1.945777138838126e-05,
+      "loss": 0.6877,
+      "step": 580
+    },
+    {
+      "epoch": 0.39314516129032256,
+      "grad_norm": 0.5476131150629957,
+      "learning_rate": 1.943855651696663e-05,
+      "loss": 0.6902,
+      "step": 585
+    },
+    {
+      "epoch": 0.396505376344086,
+      "grad_norm": 0.5140382069241621,
+      "learning_rate": 1.941901691515444e-05,
+      "loss": 0.6929,
+      "step": 590
+    },
+    {
+      "epoch": 0.39986559139784944,
+      "grad_norm": 0.4839340720292055,
+      "learning_rate": 1.9399153255198193e-05,
+      "loss": 0.694,
+      "step": 595
+    },
+    {
+      "epoch": 0.4032258064516129,
+      "grad_norm": 0.4677177786124492,
+      "learning_rate": 1.9378966220500503e-05,
+      "loss": 0.6952,
+      "step": 600
+    },
+    {
+      "epoch": 0.40658602150537637,
+      "grad_norm": 0.5207299081688349,
+      "learning_rate": 1.9358456505589586e-05,
+      "loss": 0.6932,
+      "step": 605
+    },
+    {
+      "epoch": 0.4099462365591398,
+      "grad_norm": 0.5396909713942541,
+      "learning_rate": 1.933762481609536e-05,
+      "loss": 0.684,
+      "step": 610
+    },
+    {
+      "epoch": 0.41330645161290325,
+      "grad_norm": 0.49142294486557386,
+      "learning_rate": 1.9316471868725167e-05,
+      "loss": 0.6948,
+      "step": 615
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 0.5408353090528122,
+      "learning_rate": 1.9294998391239133e-05,
+      "loss": 0.6858,
+      "step": 620
+    },
+    {
+      "epoch": 0.4200268817204301,
+      "grad_norm": 0.5233730262407599,
+      "learning_rate": 1.9273205122425108e-05,
+      "loss": 0.6887,
+      "step": 625
+    },
+    {
+      "epoch": 0.42338709677419356,
+      "grad_norm": 0.5715799297785072,
+      "learning_rate": 1.925109281207326e-05,
+      "loss": 0.6872,
+      "step": 630
+    },
+    {
+      "epoch": 0.426747311827957,
+      "grad_norm": 0.5328961169706027,
+      "learning_rate": 1.922866222095026e-05,
+      "loss": 0.6899,
+      "step": 635
+    },
+    {
+      "epoch": 0.43010752688172044,
+      "grad_norm": 0.4887114084575751,
+      "learning_rate": 1.9205914120773146e-05,
+      "loss": 0.6906,
+      "step": 640
+    },
+    {
+      "epoch": 0.4334677419354839,
+      "grad_norm": 0.5209483295062493,
+      "learning_rate": 1.9182849294182734e-05,
+      "loss": 0.6879,
+      "step": 645
+    },
+    {
+      "epoch": 0.4368279569892473,
+      "grad_norm": 0.5695163664455944,
+      "learning_rate": 1.915946853471671e-05,
+      "loss": 0.6888,
+      "step": 650
+    },
+    {
+      "epoch": 0.44018817204301075,
+      "grad_norm": 0.5383685605458758,
+      "learning_rate": 1.913577264678233e-05,
+      "loss": 0.6866,
+      "step": 655
+    },
+    {
+      "epoch": 0.4435483870967742,
+      "grad_norm": 0.5016268706035367,
+      "learning_rate": 1.9111762445628738e-05,
+      "loss": 0.6828,
+      "step": 660
+    },
+    {
+      "epoch": 0.4469086021505376,
+      "grad_norm": 0.49415639069338374,
+      "learning_rate": 1.908743875731891e-05,
+      "loss": 0.6902,
+      "step": 665
+    },
+    {
+      "epoch": 0.45026881720430106,
+      "grad_norm": 0.5339722588835791,
+      "learning_rate": 1.906280241870126e-05,
+      "loss": 0.685,
+      "step": 670
+    },
+    {
+      "epoch": 0.4536290322580645,
+      "grad_norm": 0.6037780346276747,
+      "learning_rate": 1.903785427738082e-05,
+      "loss": 0.6837,
+      "step": 675
+    },
+    {
+      "epoch": 0.45698924731182794,
+      "grad_norm": 0.5163759102323392,
+      "learning_rate": 1.9012595191690096e-05,
+      "loss": 0.6864,
+      "step": 680
+    },
+    {
+      "epoch": 0.4603494623655914,
+      "grad_norm": 0.5540520547432817,
+      "learning_rate": 1.8987026030659527e-05,
+      "loss": 0.6868,
+      "step": 685
+    },
+    {
+      "epoch": 0.4637096774193548,
+      "grad_norm": 0.5141253497235312,
+      "learning_rate": 1.8961147673987598e-05,
+      "loss": 0.685,
+      "step": 690
+    },
+    {
+      "epoch": 0.46706989247311825,
+      "grad_norm": 0.47087600726350604,
+      "learning_rate": 1.893496101201056e-05,
+      "loss": 0.6879,
+      "step": 695
+    },
+    {
+      "epoch": 0.47043010752688175,
+      "grad_norm": 0.4745813202969137,
+      "learning_rate": 1.8908466945671805e-05,
+      "loss": 0.6873,
+      "step": 700
+    },
+    {
+      "epoch": 0.4737903225806452,
+      "grad_norm": 0.5512208302577559,
+      "learning_rate": 1.8881666386490874e-05,
+      "loss": 0.6816,
+      "step": 705
+    },
+    {
+      "epoch": 0.4771505376344086,
+      "grad_norm": 0.49728239853688544,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.6839,
+      "step": 710
+    },
+    {
+      "epoch": 0.48051075268817206,
+      "grad_norm": 0.529373520548778,
+      "learning_rate": 1.882714948837286e-05,
+      "loss": 0.6824,
+      "step": 715
+    },
+    {
+      "epoch": 0.4838709677419355,
+      "grad_norm": 0.5220674061051042,
+      "learning_rate": 1.8799435025071515e-05,
+      "loss": 0.6786,
+      "step": 720
+    },
+    {
+      "epoch": 0.48723118279569894,
+      "grad_norm": 0.5130522433448756,
+      "learning_rate": 1.8771417820134963e-05,
+      "loss": 0.6822,
+      "step": 725
+    },
+    {
+      "epoch": 0.4905913978494624,
+      "grad_norm": 0.5290689359475228,
+      "learning_rate": 1.8743098837485813e-05,
+      "loss": 0.6865,
+      "step": 730
+    },
+    {
+      "epoch": 0.4939516129032258,
+      "grad_norm": 0.6064255629442339,
+      "learning_rate": 1.871447905142925e-05,
+      "loss": 0.6806,
+      "step": 735
+    },
+    {
+      "epoch": 0.49731182795698925,
+      "grad_norm": 0.5012647709893052,
+      "learning_rate": 1.868555944661949e-05,
+      "loss": 0.6835,
+      "step": 740
+    },
+    {
+      "epoch": 0.5006720430107527,
+      "grad_norm": 0.644030287593773,
+      "learning_rate": 1.865634101802592e-05,
+      "loss": 0.6833,
+      "step": 745
+    },
+    {
+      "epoch": 0.5040322580645161,
+      "grad_norm": 0.5147725717983542,
+      "learning_rate": 1.8626824770898856e-05,
+      "loss": 0.6869,
+      "step": 750
+    },
+    {
+      "epoch": 0.5073924731182796,
+      "grad_norm": 0.6476710537127688,
+      "learning_rate": 1.859701172073496e-05,
+      "loss": 0.6798,
+      "step": 755
+    },
+    {
+      "epoch": 0.510752688172043,
+      "grad_norm": 0.5471024984874042,
+      "learning_rate": 1.856690289324231e-05,
+      "loss": 0.6809,
+      "step": 760
+    },
+    {
+      "epoch": 0.5141129032258065,
+      "grad_norm": 0.5523881151273712,
+      "learning_rate": 1.8536499324305102e-05,
+      "loss": 0.6818,
+      "step": 765
+    },
+    {
+      "epoch": 0.5174731182795699,
+      "grad_norm": 0.5050686616013327,
+      "learning_rate": 1.8505802059948012e-05,
+      "loss": 0.6843,
+      "step": 770
+    },
+    {
+      "epoch": 0.5208333333333334,
+      "grad_norm": 0.5438632142417479,
+      "learning_rate": 1.847481215630021e-05,
+      "loss": 0.685,
+      "step": 775
+    },
+    {
+      "epoch": 0.5241935483870968,
+      "grad_norm": 0.4941593674742346,
+      "learning_rate": 1.844353067955902e-05,
+      "loss": 0.6808,
+      "step": 780
+    },
+    {
+      "epoch": 0.5275537634408602,
+      "grad_norm": 0.4754694994792113,
+      "learning_rate": 1.8411958705953248e-05,
+      "loss": 0.6816,
+      "step": 785
+    },
+    {
+      "epoch": 0.5309139784946236,
+      "grad_norm": 0.4805660656514537,
+      "learning_rate": 1.838009732170614e-05,
+      "loss": 0.6796,
+      "step": 790
+    },
+    {
+      "epoch": 0.5342741935483871,
+      "grad_norm": 0.44882280902701865,
+      "learning_rate": 1.834794762299803e-05,
+      "loss": 0.6835,
+      "step": 795
+    },
+    {
+      "epoch": 0.5376344086021505,
+      "grad_norm": 0.4955511641250313,
+      "learning_rate": 1.8315510715928607e-05,
+      "loss": 0.6847,
+      "step": 800
+    },
+    {
+      "epoch": 0.540994623655914,
+      "grad_norm": 0.4852311864762166,
+      "learning_rate": 1.828278771647887e-05,
+      "loss": 0.681,
+      "step": 805
+    },
+    {
+      "epoch": 0.5443548387096774,
+      "grad_norm": 0.5005675657439985,
+      "learning_rate": 1.8249779750472725e-05,
+      "loss": 0.6832,
+      "step": 810
+    },
+    {
+      "epoch": 0.5477150537634409,
+      "grad_norm": 0.5102495993533148,
+      "learning_rate": 1.821648795353827e-05,
+      "loss": 0.6817,
+      "step": 815
+    },
+    {
+      "epoch": 0.5510752688172043,
+      "grad_norm": 0.4943223187596057,
+      "learning_rate": 1.81829134710687e-05,
+      "loss": 0.6791,
+      "step": 820
+    },
+    {
+      "epoch": 0.5544354838709677,
+      "grad_norm": 0.542801548624785,
+      "learning_rate": 1.8149057458182924e-05,
+      "loss": 0.6832,
+      "step": 825
+    },
+    {
+      "epoch": 0.5577956989247311,
+      "grad_norm": 0.45148280962815884,
+      "learning_rate": 1.81149210796858e-05,
+      "loss": 0.6784,
+      "step": 830
+    },
+    {
+      "epoch": 0.5611559139784946,
+      "grad_norm": 0.4566100659521107,
+      "learning_rate": 1.8080505510028073e-05,
+      "loss": 0.6811,
+      "step": 835
+    },
+    {
+      "epoch": 0.5645161290322581,
+      "grad_norm": 0.45139497231587655,
+      "learning_rate": 1.8045811933265973e-05,
+      "loss": 0.6782,
+      "step": 840
+    },
+    {
+      "epoch": 0.5678763440860215,
+      "grad_norm": 0.5039358231725258,
+      "learning_rate": 1.8010841543020472e-05,
+      "loss": 0.6776,
+      "step": 845
+    },
+    {
+      "epoch": 0.571236559139785,
+      "grad_norm": 0.44244553735827763,
+      "learning_rate": 1.7975595542436207e-05,
+      "loss": 0.6795,
+      "step": 850
+    },
+    {
+      "epoch": 0.5745967741935484,
+      "grad_norm": 0.513649210570403,
+      "learning_rate": 1.794007514414011e-05,
+      "loss": 0.6777,
+      "step": 855
+    },
+    {
+      "epoch": 0.5779569892473119,
+      "grad_norm": 0.4779895963476636,
+      "learning_rate": 1.790428157019967e-05,
+      "loss": 0.6807,
+      "step": 860
+    },
+    {
+      "epoch": 0.5813172043010753,
+      "grad_norm": 0.5317771601563221,
+      "learning_rate": 1.7868216052080898e-05,
+      "loss": 0.6774,
+      "step": 865
+    },
+    {
+      "epoch": 0.5846774193548387,
+      "grad_norm": 0.4387347378891155,
+      "learning_rate": 1.783187983060594e-05,
+      "loss": 0.6771,
+      "step": 870
+    },
+    {
+      "epoch": 0.5880376344086021,
+      "grad_norm": 0.47208773953042626,
+      "learning_rate": 1.7795274155910408e-05,
+      "loss": 0.6795,
+      "step": 875
+    },
+    {
+      "epoch": 0.5913978494623656,
+      "grad_norm": 0.4919880987398928,
+      "learning_rate": 1.7758400287400372e-05,
+      "loss": 0.6764,
+      "step": 880
+    },
+    {
+      "epoch": 0.594758064516129,
+      "grad_norm": 0.4746143436746516,
+      "learning_rate": 1.772125949370901e-05,
+      "loss": 0.6777,
+      "step": 885
+    },
+    {
+      "epoch": 0.5981182795698925,
+      "grad_norm": 0.4866530917728411,
+      "learning_rate": 1.7683853052652974e-05,
+      "loss": 0.6778,
+      "step": 890
+    },
+    {
+      "epoch": 0.6014784946236559,
+      "grad_norm": 0.45607894178261754,
+      "learning_rate": 1.764618225118843e-05,
+      "loss": 0.675,
+      "step": 895
+    },
+    {
+      "epoch": 0.6048387096774194,
+      "grad_norm": 0.4623134720756107,
+      "learning_rate": 1.7608248385366774e-05,
+      "loss": 0.6724,
+      "step": 900
+    },
+    {
+      "epoch": 0.6081989247311828,
+      "grad_norm": 0.41997040926221346,
+      "learning_rate": 1.7570052760290036e-05,
+      "loss": 0.6749,
+      "step": 905
+    },
+    {
+      "epoch": 0.6115591397849462,
+      "grad_norm": 0.4352461962298591,
+      "learning_rate": 1.7531596690065998e-05,
+      "loss": 0.673,
+      "step": 910
+    },
+    {
+      "epoch": 0.6149193548387096,
+      "grad_norm": 0.47410360513736394,
+      "learning_rate": 1.7492881497762958e-05,
+      "loss": 0.6758,
+      "step": 915
+    },
+    {
+      "epoch": 0.6182795698924731,
+      "grad_norm": 0.45269850953936747,
+      "learning_rate": 1.7453908515364238e-05,
+      "loss": 0.6802,
+      "step": 920
+    },
+    {
+      "epoch": 0.6216397849462365,
+      "grad_norm": 0.46524887071623716,
+      "learning_rate": 1.741467908372233e-05,
+      "loss": 0.674,
+      "step": 925
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.4739090399223664,
+      "learning_rate": 1.7375194552512777e-05,
+      "loss": 0.6743,
+      "step": 930
+    },
+    {
+      "epoch": 0.6283602150537635,
+      "grad_norm": 0.50057753354274,
+      "learning_rate": 1.7335456280187752e-05,
+      "loss": 0.678,
+      "step": 935
+    },
+    {
+      "epoch": 0.6317204301075269,
+      "grad_norm": 0.4342277157697679,
+      "learning_rate": 1.729546563392929e-05,
+      "loss": 0.6766,
+      "step": 940
+    },
+    {
+      "epoch": 0.6350806451612904,
+      "grad_norm": 0.5421833972178155,
+      "learning_rate": 1.7255223989602277e-05,
+      "loss": 0.6725,
+      "step": 945
+    },
+    {
+      "epoch": 0.6384408602150538,
+      "grad_norm": 0.5180088295892716,
+      "learning_rate": 1.72147327317071e-05,
+      "loss": 0.6766,
+      "step": 950
+    },
+    {
+      "epoch": 0.6418010752688172,
+      "grad_norm": 0.4835576817093598,
+      "learning_rate": 1.7173993253332016e-05,
+      "loss": 0.6754,
+      "step": 955
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 0.4566946543115902,
+      "learning_rate": 1.7133006956105237e-05,
+      "loss": 0.6751,
+      "step": 960
+    },
+    {
+      "epoch": 0.6485215053763441,
+      "grad_norm": 0.43744247420203414,
+      "learning_rate": 1.7091775250146678e-05,
+      "loss": 0.6708,
+      "step": 965
+    },
+    {
+      "epoch": 0.6518817204301075,
+      "grad_norm": 0.4237915807278061,
+      "learning_rate": 1.7050299554019466e-05,
+      "loss": 0.6702,
+      "step": 970
+    },
+    {
+      "epoch": 0.655241935483871,
+      "grad_norm": 0.4159411791640258,
+      "learning_rate": 1.700858129468114e-05,
+      "loss": 0.6738,
+      "step": 975
+    },
+    {
+      "epoch": 0.6586021505376344,
+      "grad_norm": 0.47229925238109904,
+      "learning_rate": 1.696662190743453e-05,
+      "loss": 0.6731,
+      "step": 980
+    },
+    {
+      "epoch": 0.6619623655913979,
+      "grad_norm": 0.4357051029703012,
+      "learning_rate": 1.69244228358784e-05,
+      "loss": 0.6732,
+      "step": 985
+    },
+    {
+      "epoch": 0.6653225806451613,
+      "grad_norm": 0.47900841918884185,
+      "learning_rate": 1.688198553185777e-05,
+      "loss": 0.6742,
+      "step": 990
+    },
+    {
+      "epoch": 0.6686827956989247,
+      "grad_norm": 0.529861867014197,
+      "learning_rate": 1.683931145541397e-05,
+      "loss": 0.6753,
+      "step": 995
+    },
+    {
+      "epoch": 0.6720430107526881,
+      "grad_norm": 0.49314598912759483,
+      "learning_rate": 1.6796402074734404e-05,
+      "loss": 0.6694,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6754032258064516,
+      "grad_norm": 0.47658141232504353,
+      "learning_rate": 1.6753258866102047e-05,
+      "loss": 0.6733,
+      "step": 1005
+    },
+    {
+      "epoch": 0.678763440860215,
+      "grad_norm": 0.4555180579717891,
+      "learning_rate": 1.6709883313844634e-05,
+      "loss": 0.6771,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6821236559139785,
+      "grad_norm": 0.4804224902104754,
+      "learning_rate": 1.6666276910283623e-05,
+      "loss": 0.6754,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6854838709677419,
+      "grad_norm": 0.4676965423125691,
+      "learning_rate": 1.662244115568282e-05,
+      "loss": 0.6697,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6888440860215054,
+      "grad_norm": 0.4309666314092742,
+      "learning_rate": 1.657837755819678e-05,
+      "loss": 0.6682,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6922043010752689,
+      "grad_norm": 0.4526083077842183,
+      "learning_rate": 1.6534087633818914e-05,
+      "loss": 0.6738,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6955645161290323,
+      "grad_norm": 0.42731232075306474,
+      "learning_rate": 1.6489572906329345e-05,
+      "loss": 0.6699,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6989247311827957,
+      "grad_norm": 0.4584418364845978,
+      "learning_rate": 1.644483490724247e-05,
+      "loss": 0.671,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7022849462365591,
+      "grad_norm": 0.45883989932174113,
+      "learning_rate": 1.6399875175754258e-05,
+      "loss": 0.6696,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7056451612903226,
+      "grad_norm": 0.44771182993617314,
+      "learning_rate": 1.635469525868932e-05,
+      "loss": 0.6721,
+      "step": 1050
+    },
+    {
+      "epoch": 0.709005376344086,
+      "grad_norm": 0.4588720172063543,
+      "learning_rate": 1.6309296710447674e-05,
+      "loss": 0.6697,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7123655913978495,
+      "grad_norm": 0.45001838732164606,
+      "learning_rate": 1.626368109295128e-05,
+      "loss": 0.667,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7157258064516129,
+      "grad_norm": 0.42641910570356634,
+      "learning_rate": 1.6217849975590275e-05,
+      "loss": 0.6708,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7190860215053764,
+      "grad_norm": 0.422498792435904,
+      "learning_rate": 1.617180493516901e-05,
+      "loss": 0.6711,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7224462365591398,
+      "grad_norm": 0.4415726107421213,
+      "learning_rate": 1.6125547555851787e-05,
+      "loss": 0.6685,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7258064516129032,
+      "grad_norm": 0.4534706255885532,
+      "learning_rate": 1.6079079429108357e-05,
+      "loss": 0.6696,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7291666666666666,
+      "grad_norm": 0.4029176160192046,
+      "learning_rate": 1.6032402153659164e-05,
+      "loss": 0.6726,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7325268817204301,
+      "grad_norm": 0.42621354076663304,
+      "learning_rate": 1.5985517335420346e-05,
+      "loss": 0.6666,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7358870967741935,
+      "grad_norm": 0.44234432560741715,
+      "learning_rate": 1.5938426587448473e-05,
+      "loss": 0.6741,
+      "step": 1095
+    },
+    {
+      "epoch": 0.739247311827957,
+      "grad_norm": 0.4473272678937478,
+      "learning_rate": 1.589113152988507e-05,
+      "loss": 0.6709,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7426075268817204,
+      "grad_norm": 0.44595646225862023,
+      "learning_rate": 1.5843633789900862e-05,
+      "loss": 0.6712,
+      "step": 1105
+    },
+    {
+      "epoch": 0.7459677419354839,
+      "grad_norm": 0.4324555524707172,
+      "learning_rate": 1.5795935001639796e-05,
+      "loss": 0.6658,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7493279569892473,
+      "grad_norm": 0.46394662933500824,
+      "learning_rate": 1.5748036806162816e-05,
+      "loss": 0.6684,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7526881720430108,
+      "grad_norm": 0.4947844886498111,
+      "learning_rate": 1.56999408513914e-05,
+      "loss": 0.6722,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7560483870967742,
+      "grad_norm": 0.4250235043595764,
+      "learning_rate": 1.5651648792050886e-05,
+      "loss": 0.6686,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7594086021505376,
+      "grad_norm": 0.4159591266629867,
+      "learning_rate": 1.5603162289613503e-05,
+      "loss": 0.669,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7627688172043011,
+      "grad_norm": 0.4095733043870104,
+      "learning_rate": 1.555448301224124e-05,
+      "loss": 0.6692,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7661290322580645,
+      "grad_norm": 0.43896150432589165,
+      "learning_rate": 1.550561263472845e-05,
+      "loss": 0.6657,
+      "step": 1140
+    },
+    {
+      "epoch": 0.769489247311828,
+      "grad_norm": 0.47253071219496107,
+      "learning_rate": 1.5456552838444215e-05,
+      "loss": 0.6695,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7728494623655914,
+      "grad_norm": 0.43201524742454317,
+      "learning_rate": 1.5407305311274502e-05,
+      "loss": 0.6639,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7762096774193549,
+      "grad_norm": 0.46875979315386157,
+      "learning_rate": 1.5357871747564108e-05,
+      "loss": 0.6696,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7795698924731183,
+      "grad_norm": 0.4458966308423493,
+      "learning_rate": 1.530825384805835e-05,
+      "loss": 0.6671,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7829301075268817,
+      "grad_norm": 0.4045923602073338,
+      "learning_rate": 1.5258453319844555e-05,
+      "loss": 0.6679,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7862903225806451,
+      "grad_norm": 0.46081178192946914,
+      "learning_rate": 1.5208471876293334e-05,
+      "loss": 0.6653,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7896505376344086,
+      "grad_norm": 0.4412474732597341,
+      "learning_rate": 1.5158311236999622e-05,
+      "loss": 0.6696,
+      "step": 1175
+    },
+    {
+      "epoch": 0.793010752688172,
+      "grad_norm": 0.4561523660741418,
+      "learning_rate": 1.5107973127723537e-05,
+      "loss": 0.6662,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7963709677419355,
+      "grad_norm": 0.45370412836150925,
+      "learning_rate": 1.5057459280330984e-05,
+      "loss": 0.6668,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7997311827956989,
+      "grad_norm": 0.4641379862022367,
+      "learning_rate": 1.500677143273408e-05,
+      "loss": 0.6618,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8030913978494624,
+      "grad_norm": 0.4228354205291783,
+      "learning_rate": 1.4955911328831357e-05,
+      "loss": 0.6689,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8064516129032258,
+      "grad_norm": 0.45194187006117376,
+      "learning_rate": 1.4904880718447781e-05,
+      "loss": 0.6654,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8098118279569892,
+      "grad_norm": 0.41370408468731323,
+      "learning_rate": 1.4853681357274522e-05,
+      "loss": 0.6656,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8131720430107527,
+      "grad_norm": 0.39658836438427153,
+      "learning_rate": 1.4802315006808578e-05,
+      "loss": 0.6642,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8165322580645161,
+      "grad_norm": 0.39180055538376923,
+      "learning_rate": 1.4750783434292147e-05,
+      "loss": 0.6669,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8198924731182796,
+      "grad_norm": 0.4140385959248494,
+      "learning_rate": 1.4699088412651841e-05,
+      "loss": 0.6677,
+      "step": 1220
+    },
+    {
+      "epoch": 0.823252688172043,
+      "grad_norm": 0.4135853126184148,
+      "learning_rate": 1.4647231720437687e-05,
+      "loss": 0.6691,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8266129032258065,
+      "grad_norm": 0.3886901463807677,
+      "learning_rate": 1.4595215141761934e-05,
+      "loss": 0.6612,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8299731182795699,
+      "grad_norm": 0.427261482590677,
+      "learning_rate": 1.4543040466237662e-05,
+      "loss": 0.6635,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.41557585081944015,
+      "learning_rate": 1.4490709488917239e-05,
+      "loss": 0.6642,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8366935483870968,
+      "grad_norm": 0.41396481934719226,
+      "learning_rate": 1.4438224010230526e-05,
+      "loss": 0.6642,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8400537634408602,
+      "grad_norm": 0.4000957409618339,
+      "learning_rate": 1.4385585835922962e-05,
+      "loss": 0.6679,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8434139784946236,
+      "grad_norm": 0.5093831922663081,
+      "learning_rate": 1.4332796776993423e-05,
+      "loss": 0.6665,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8467741935483871,
+      "grad_norm": 0.39777324014388976,
+      "learning_rate": 1.427985864963193e-05,
+      "loss": 0.6608,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8501344086021505,
+      "grad_norm": 0.4123671465949071,
+      "learning_rate": 1.4226773275157142e-05,
+      "loss": 0.6679,
+      "step": 1265
+    },
+    {
+      "epoch": 0.853494623655914,
+      "grad_norm": 0.4178878559263764,
+      "learning_rate": 1.4173542479953712e-05,
+      "loss": 0.6629,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8568548387096774,
+      "grad_norm": 0.4151008943507023,
+      "learning_rate": 1.412016809540944e-05,
+      "loss": 0.6708,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8602150537634409,
+      "grad_norm": 0.39420851761758596,
+      "learning_rate": 1.406665195785228e-05,
+      "loss": 0.66,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8635752688172043,
+      "grad_norm": 0.39995681473664046,
+      "learning_rate": 1.401299590848714e-05,
+      "loss": 0.6631,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8669354838709677,
+      "grad_norm": 0.43024431364313487,
+      "learning_rate": 1.3959201793332554e-05,
+      "loss": 0.6648,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8702956989247311,
+      "grad_norm": 0.3823413783572042,
+      "learning_rate": 1.3905271463157153e-05,
+      "loss": 0.6668,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8736559139784946,
+      "grad_norm": 0.4168656965051171,
+      "learning_rate": 1.385120677341602e-05,
+      "loss": 0.6584,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8770161290322581,
+      "grad_norm": 0.3967459217329594,
+      "learning_rate": 1.3797009584186818e-05,
+      "loss": 0.6627,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8803763440860215,
+      "grad_norm": 0.4090750420770741,
+      "learning_rate": 1.3742681760105814e-05,
+      "loss": 0.6616,
+      "step": 1310
+    },
+    {
+      "epoch": 0.883736559139785,
+      "grad_norm": 0.4175005433442616,
+      "learning_rate": 1.3688225170303727e-05,
+      "loss": 0.6668,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8870967741935484,
+      "grad_norm": 0.425478340178616,
+      "learning_rate": 1.3633641688341413e-05,
+      "loss": 0.6613,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8904569892473119,
+      "grad_norm": 0.39081095331783994,
+      "learning_rate": 1.3578933192145421e-05,
+      "loss": 0.6631,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8938172043010753,
+      "grad_norm": 0.460628551115726,
+      "learning_rate": 1.3524101563943356e-05,
+      "loss": 0.6617,
+      "step": 1330
+    },
+    {
+      "epoch": 0.8971774193548387,
+      "grad_norm": 0.4558498194690748,
+      "learning_rate": 1.3469148690199157e-05,
+      "loss": 0.665,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9005376344086021,
+      "grad_norm": 0.4182583525353485,
+      "learning_rate": 1.3414076461548162e-05,
+      "loss": 0.662,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9038978494623656,
+      "grad_norm": 0.4438115831427305,
+      "learning_rate": 1.3358886772732085e-05,
+      "loss": 0.6638,
+      "step": 1345
+    },
+    {
+      "epoch": 0.907258064516129,
+      "grad_norm": 0.45301099015215857,
+      "learning_rate": 1.3303581522533806e-05,
+      "loss": 0.6602,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9106182795698925,
+      "grad_norm": 0.4353757710607073,
+      "learning_rate": 1.3248162613712066e-05,
+      "loss": 0.6614,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9139784946236559,
+      "grad_norm": 0.3966488654029928,
+      "learning_rate": 1.319263195293599e-05,
+      "loss": 0.661,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9173387096774194,
+      "grad_norm": 0.43451849890336813,
+      "learning_rate": 1.3136991450719493e-05,
+      "loss": 0.6632,
+      "step": 1365
+    },
+    {
+      "epoch": 0.9206989247311828,
+      "grad_norm": 0.4331957813206352,
+      "learning_rate": 1.3081243021355542e-05,
+      "loss": 0.6618,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9240591397849462,
+      "grad_norm": 0.42461868845233375,
+      "learning_rate": 1.3025388582850311e-05,
+      "loss": 0.661,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9274193548387096,
+      "grad_norm": 0.4270005545440327,
+      "learning_rate": 1.2969430056857177e-05,
+      "loss": 0.6619,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9307795698924731,
+      "grad_norm": 0.40199632811605207,
+      "learning_rate": 1.2913369368610618e-05,
+      "loss": 0.6625,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9341397849462365,
+      "grad_norm": 0.40238063432322,
+      "learning_rate": 1.285720844685996e-05,
+      "loss": 0.6619,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.4063919100845298,
+      "learning_rate": 1.2800949223803039e-05,
+      "loss": 0.6593,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9408602150537635,
+      "grad_norm": 0.42658950759070957,
+      "learning_rate": 1.274459363501971e-05,
+      "loss": 0.6663,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9442204301075269,
+      "grad_norm": 0.40879880761812026,
+      "learning_rate": 1.2688143619405266e-05,
+      "loss": 0.6626,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9475806451612904,
+      "grad_norm": 0.4136553833868636,
+      "learning_rate": 1.2631601119103714e-05,
+      "loss": 0.6595,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9509408602150538,
+      "grad_norm": 0.4566408970103592,
+      "learning_rate": 1.2574968079440969e-05,
+      "loss": 0.6612,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9543010752688172,
+      "grad_norm": 0.4707155898476047,
+      "learning_rate": 1.251824644885792e-05,
+      "loss": 0.661,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9576612903225806,
+      "grad_norm": 0.44843086084440054,
+      "learning_rate": 1.2461438178843409e-05,
+      "loss": 0.6603,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9610215053763441,
+      "grad_norm": 0.3887174931118394,
+      "learning_rate": 1.2404545223867067e-05,
+      "loss": 0.6648,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9643817204301075,
+      "grad_norm": 0.3844627071853656,
+      "learning_rate": 1.2347569541312086e-05,
+      "loss": 0.661,
+      "step": 1435
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 0.38712969083554144,
+      "learning_rate": 1.2290513091407871e-05,
+      "loss": 0.6626,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9711021505376344,
+      "grad_norm": 0.3765260995061332,
+      "learning_rate": 1.223337783716261e-05,
+      "loss": 0.658,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9744623655913979,
+      "grad_norm": 0.38880975825103575,
+      "learning_rate": 1.2176165744295716e-05,
+      "loss": 0.6625,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9778225806451613,
+      "grad_norm": 0.38891299522851835,
+      "learning_rate": 1.2118878781170213e-05,
+      "loss": 0.6571,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9811827956989247,
+      "grad_norm": 0.3883707422936783,
+      "learning_rate": 1.2061518918725019e-05,
+      "loss": 0.6554,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9845430107526881,
+      "grad_norm": 0.4326330855587226,
+      "learning_rate": 1.2004088130407114e-05,
+      "loss": 0.6617,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9879032258064516,
+      "grad_norm": 0.40177893100437934,
+      "learning_rate": 1.1946588392103678e-05,
+      "loss": 0.6601,
+      "step": 1470
+    },
+    {
+      "epoch": 0.991263440860215,
+      "grad_norm": 0.4069182709848244,
+      "learning_rate": 1.1889021682074074e-05,
+      "loss": 0.6578,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9946236559139785,
+      "grad_norm": 0.371562924814927,
+      "learning_rate": 1.1831389980881815e-05,
+      "loss": 0.6595,
+      "step": 1480
+    },
+    {
+      "epoch": 0.9979838709677419,
+      "grad_norm": 0.40260684742734143,
+      "learning_rate": 1.1773695271326413e-05,
+      "loss": 0.6598,
+      "step": 1485
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.6596925854682922,
+      "eval_runtime": 38.4699,
+      "eval_samples_per_second": 75.8,
+      "eval_steps_per_second": 1.196,
+      "step": 1488
+    },
+    {
+      "epoch": 1.0013440860215055,
+      "grad_norm": 0.43923520164769597,
+      "learning_rate": 1.1715939538375159e-05,
+      "loss": 0.6515,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0047043010752688,
+      "grad_norm": 0.4093115648975466,
+      "learning_rate": 1.1658124769094834e-05,
+      "loss": 0.6335,
+      "step": 1495
+    },
+    {
+      "epoch": 1.0080645161290323,
+      "grad_norm": 0.44488404164651246,
+      "learning_rate": 1.160025295258335e-05,
+      "loss": 0.6351,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0114247311827957,
+      "grad_norm": 0.41541519893053236,
+      "learning_rate": 1.1542326079901296e-05,
+      "loss": 0.6363,
+      "step": 1505
+    },
+    {
+      "epoch": 1.0147849462365592,
+      "grad_norm": 0.4120083737414266,
+      "learning_rate": 1.1484346144003467e-05,
+      "loss": 0.629,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0181451612903225,
+      "grad_norm": 0.4134775783414065,
+      "learning_rate": 1.1426315139670268e-05,
+      "loss": 0.6358,
+      "step": 1515
+    },
+    {
+      "epoch": 1.021505376344086,
+      "grad_norm": 0.414634455881357,
+      "learning_rate": 1.1368235063439103e-05,
+      "loss": 0.6316,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0248655913978495,
+      "grad_norm": 0.4254446985228589,
+      "learning_rate": 1.1310107913535677e-05,
+      "loss": 0.6323,
+      "step": 1525
+    },
+    {
+      "epoch": 1.028225806451613,
+      "grad_norm": 0.4034820728921977,
+      "learning_rate": 1.1251935689805249e-05,
+      "loss": 0.6348,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0315860215053763,
+      "grad_norm": 0.41791569897419,
+      "learning_rate": 1.1193720393643826e-05,
+      "loss": 0.6319,
+      "step": 1535
+    },
+    {
+      "epoch": 1.0349462365591398,
+      "grad_norm": 0.41814029821287946,
+      "learning_rate": 1.1135464027929306e-05,
+      "loss": 0.6384,
+      "step": 1540
+    },
+    {
+      "epoch": 1.0383064516129032,
+      "grad_norm": 0.42081017954332256,
+      "learning_rate": 1.1077168596952579e-05,
+      "loss": 0.6378,
+      "step": 1545
+    },
+    {
+      "epoch": 1.0416666666666667,
+      "grad_norm": 0.39427585600241666,
+      "learning_rate": 1.1018836106348558e-05,
+      "loss": 0.6308,
+      "step": 1550
+    },
+    {
+      "epoch": 1.04502688172043,
+      "grad_norm": 0.41657992186432496,
+      "learning_rate": 1.0960468563027178e-05,
+      "loss": 0.6359,
+      "step": 1555
+    },
+    {
+      "epoch": 1.0483870967741935,
+      "grad_norm": 0.42812446527859066,
+      "learning_rate": 1.0902067975104354e-05,
+      "loss": 0.6375,
+      "step": 1560
+    },
+    {
+      "epoch": 1.051747311827957,
+      "grad_norm": 0.4250184457384185,
+      "learning_rate": 1.0843636351832888e-05,
+      "loss": 0.6335,
+      "step": 1565
+    },
+    {
+      "epoch": 1.0551075268817205,
+      "grad_norm": 0.4039339824459457,
+      "learning_rate": 1.0785175703533341e-05,
+      "loss": 0.6354,
+      "step": 1570
+    },
+    {
+      "epoch": 1.0584677419354838,
+      "grad_norm": 0.4350937331558841,
+      "learning_rate": 1.072668804152488e-05,
+      "loss": 0.6346,
+      "step": 1575
+    },
+    {
+      "epoch": 1.0618279569892473,
+      "grad_norm": 0.43413177405165077,
+      "learning_rate": 1.0668175378056053e-05,
+      "loss": 0.6328,
+      "step": 1580
+    },
+    {
+      "epoch": 1.0651881720430108,
+      "grad_norm": 0.4543315336438448,
+      "learning_rate": 1.0609639726235592e-05,
+      "loss": 0.6291,
+      "step": 1585
+    },
+    {
+      "epoch": 1.0685483870967742,
+      "grad_norm": 0.4639621657382721,
+      "learning_rate": 1.0551083099963125e-05,
+      "loss": 0.6347,
+      "step": 1590
+    },
+    {
+      "epoch": 1.0719086021505377,
+      "grad_norm": 0.38636625101348104,
+      "learning_rate": 1.0492507513859904e-05,
+      "loss": 0.6329,
+      "step": 1595
+    },
+    {
+      "epoch": 1.075268817204301,
+      "grad_norm": 0.43112374827834826,
+      "learning_rate": 1.043391498319948e-05,
+      "loss": 0.6334,
+      "step": 1600
+    },
+    {
+      "epoch": 1.0786290322580645,
+      "grad_norm": 0.4083117563136851,
+      "learning_rate": 1.037530752383839e-05,
+      "loss": 0.6312,
+      "step": 1605
+    },
+    {
+      "epoch": 1.081989247311828,
+      "grad_norm": 0.4092678104422559,
+      "learning_rate": 1.0316687152146774e-05,
+      "loss": 0.6341,
+      "step": 1610
+    },
+    {
+      "epoch": 1.0853494623655915,
+      "grad_norm": 0.4236890040624566,
+      "learning_rate": 1.0258055884939023e-05,
+      "loss": 0.6342,
+      "step": 1615
+    },
+    {
+      "epoch": 1.0887096774193548,
+      "grad_norm": 0.4129609373942744,
+      "learning_rate": 1.0199415739404381e-05,
+      "loss": 0.6337,
+      "step": 1620
+    },
+    {
+      "epoch": 1.0920698924731183,
+      "grad_norm": 0.45620399681821466,
+      "learning_rate": 1.0140768733037558e-05,
+      "loss": 0.6355,
+      "step": 1625
+    },
+    {
+      "epoch": 1.0954301075268817,
+      "grad_norm": 0.3996898071123565,
+      "learning_rate": 1.0082116883569295e-05,
+      "loss": 0.6353,
+      "step": 1630
+    },
+    {
+      "epoch": 1.0987903225806452,
+      "grad_norm": 0.42181670497621626,
+      "learning_rate": 1.0023462208896967e-05,
+      "loss": 0.6368,
+      "step": 1635
+    },
+    {
+      "epoch": 1.1021505376344085,
+      "grad_norm": 0.4170971985416292,
+      "learning_rate": 9.964806727015144e-06,
+      "loss": 0.6369,
+      "step": 1640
+    },
+    {
+      "epoch": 1.105510752688172,
+      "grad_norm": 0.3850003704964035,
+      "learning_rate": 9.906152455946176e-06,
+      "loss": 0.6316,
+      "step": 1645
+    },
+    {
+      "epoch": 1.1088709677419355,
+      "grad_norm": 0.3959138602164967,
+      "learning_rate": 9.847501413670742e-06,
+      "loss": 0.6329,
+      "step": 1650
+    },
+    {
+      "epoch": 1.112231182795699,
+      "grad_norm": 0.38653649392484757,
+      "learning_rate": 9.788855618058446e-06,
+      "loss": 0.6315,
+      "step": 1655
+    },
+    {
+      "epoch": 1.1155913978494623,
+      "grad_norm": 0.434626659914302,
+      "learning_rate": 9.730217086798387e-06,
+      "loss": 0.6329,
+      "step": 1660
+    },
+    {
+      "epoch": 1.1189516129032258,
+      "grad_norm": 0.4043062586269176,
+      "learning_rate": 9.671587837329717e-06,
+      "loss": 0.6314,
+      "step": 1665
+    },
+    {
+      "epoch": 1.1223118279569892,
+      "grad_norm": 0.3912022835272916,
+      "learning_rate": 9.612969886772272e-06,
+      "loss": 0.6347,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1256720430107527,
+      "grad_norm": 0.3945451917100905,
+      "learning_rate": 9.554365251857141e-06,
+      "loss": 0.6334,
+      "step": 1675
+    },
+    {
+      "epoch": 1.129032258064516,
+      "grad_norm": 0.39115203980879226,
+      "learning_rate": 9.495775948857292e-06,
+      "loss": 0.631,
+      "step": 1680
+    },
+    {
+      "epoch": 1.1323924731182795,
+      "grad_norm": 0.3899923876549679,
+      "learning_rate": 9.437203993518214e-06,
+      "loss": 0.6304,
+      "step": 1685
+    },
+    {
+      "epoch": 1.135752688172043,
+      "grad_norm": 0.4002252410564929,
+      "learning_rate": 9.37865140098854e-06,
+      "loss": 0.6313,
+      "step": 1690
+    },
+    {
+      "epoch": 1.1391129032258065,
+      "grad_norm": 0.43477646973135253,
+      "learning_rate": 9.320120185750745e-06,
+      "loss": 0.6319,
+      "step": 1695
+    },
+    {
+      "epoch": 1.14247311827957,
+      "grad_norm": 0.4090247302348437,
+      "learning_rate": 9.261612361551827e-06,
+      "loss": 0.6317,
+      "step": 1700
+    },
+    {
+      "epoch": 1.1458333333333333,
+      "grad_norm": 0.45307609961969436,
+      "learning_rate": 9.203129941334004e-06,
+      "loss": 0.6333,
+      "step": 1705
+    },
+    {
+      "epoch": 1.1491935483870968,
+      "grad_norm": 0.41729176695078596,
+      "learning_rate": 9.144674937165503e-06,
+      "loss": 0.6351,
+      "step": 1710
+    },
+    {
+      "epoch": 1.1525537634408602,
+      "grad_norm": 0.40602215785080276,
+      "learning_rate": 9.086249360171291e-06,
+      "loss": 0.6343,
+      "step": 1715
+    },
+    {
+      "epoch": 1.1559139784946237,
+      "grad_norm": 0.45074003009366553,
+      "learning_rate": 9.027855220463915e-06,
+      "loss": 0.6335,
+      "step": 1720
+    },
+    {
+      "epoch": 1.159274193548387,
+      "grad_norm": 0.4134807886527817,
+      "learning_rate": 8.969494527074333e-06,
+      "loss": 0.6324,
+      "step": 1725
+    },
+    {
+      "epoch": 1.1626344086021505,
+      "grad_norm": 0.4342008270842711,
+      "learning_rate": 8.91116928788278e-06,
+      "loss": 0.6272,
+      "step": 1730
+    },
+    {
+      "epoch": 1.165994623655914,
+      "grad_norm": 0.4390573307122215,
+      "learning_rate": 8.852881509549716e-06,
+      "loss": 0.6319,
+      "step": 1735
+    },
+    {
+      "epoch": 1.1693548387096775,
+      "grad_norm": 0.39093522707469097,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.6339,
+      "step": 1740
+    },
+    {
+      "epoch": 1.172715053763441,
+      "grad_norm": 0.3954417150387851,
+      "learning_rate": 8.73642635558774e-06,
+      "loss": 0.6323,
+      "step": 1745
+    },
+    {
+      "epoch": 1.1760752688172043,
+      "grad_norm": 0.40100373608359324,
+      "learning_rate": 8.678262986559667e-06,
+      "loss": 0.6269,
+      "step": 1750
+    },
+    {
+      "epoch": 1.1794354838709677,
+      "grad_norm": 0.39734268517939386,
+      "learning_rate": 8.620145091453907e-06,
+      "loss": 0.6332,
+      "step": 1755
+    },
+    {
+      "epoch": 1.1827956989247312,
+      "grad_norm": 0.39585129412547304,
+      "learning_rate": 8.562074669797313e-06,
+      "loss": 0.6292,
+      "step": 1760
+    },
+    {
+      "epoch": 1.1861559139784945,
+      "grad_norm": 0.4037432361209901,
+      "learning_rate": 8.504053719483433e-06,
+      "loss": 0.6311,
+      "step": 1765
+    },
+    {
+      "epoch": 1.189516129032258,
+      "grad_norm": 0.3975358417069771,
+      "learning_rate": 8.446084236703758e-06,
+      "loss": 0.6292,
+      "step": 1770
+    },
+    {
+      "epoch": 1.1928763440860215,
+      "grad_norm": 0.4254328638432713,
+      "learning_rate": 8.38816821587906e-06,
+      "loss": 0.6294,
+      "step": 1775
+    },
+    {
+      "epoch": 1.196236559139785,
+      "grad_norm": 0.4130515802670977,
+      "learning_rate": 8.330307649590782e-06,
+      "loss": 0.6304,
+      "step": 1780
+    },
+    {
+      "epoch": 1.1995967741935485,
+      "grad_norm": 0.417130552856278,
+      "learning_rate": 8.272504528512448e-06,
+      "loss": 0.6329,
+      "step": 1785
+    },
+    {
+      "epoch": 1.2029569892473118,
+      "grad_norm": 0.40653243786568827,
+      "learning_rate": 8.214760841341223e-06,
+      "loss": 0.6343,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2063172043010753,
+      "grad_norm": 0.40605875241786166,
+      "learning_rate": 8.157078574729451e-06,
+      "loss": 0.6302,
+      "step": 1795
+    },
+    {
+      "epoch": 1.2096774193548387,
+      "grad_norm": 0.38683973621773765,
+      "learning_rate": 8.099459713216331e-06,
+      "loss": 0.6266,
+      "step": 1800
+    },
+    {
+      "epoch": 1.2130376344086022,
+      "grad_norm": 0.39898560454739485,
+      "learning_rate": 8.041906239159636e-06,
+      "loss": 0.6303,
+      "step": 1805
+    },
+    {
+      "epoch": 1.2163978494623655,
+      "grad_norm": 0.4377517352185253,
+      "learning_rate": 7.984420132667483e-06,
+      "loss": 0.63,
+      "step": 1810
+    },
+    {
+      "epoch": 1.219758064516129,
+      "grad_norm": 0.39945365735612615,
+      "learning_rate": 7.927003371530254e-06,
+      "loss": 0.6291,
+      "step": 1815
+    },
+    {
+      "epoch": 1.2231182795698925,
+      "grad_norm": 0.4030460272109659,
+      "learning_rate": 7.869657931152522e-06,
+      "loss": 0.6262,
+      "step": 1820
+    },
+    {
+      "epoch": 1.226478494623656,
+      "grad_norm": 0.3958516566003825,
+      "learning_rate": 7.812385784485079e-06,
+      "loss": 0.631,
+      "step": 1825
+    },
+    {
+      "epoch": 1.2298387096774193,
+      "grad_norm": 0.4234150471780541,
+      "learning_rate": 7.755188901957093e-06,
+      "loss": 0.6271,
+      "step": 1830
+    },
+    {
+      "epoch": 1.2331989247311828,
+      "grad_norm": 0.43581653936134584,
+      "learning_rate": 7.698069251408271e-06,
+      "loss": 0.6289,
+      "step": 1835
+    },
+    {
+      "epoch": 1.2365591397849462,
+      "grad_norm": 0.871860809323723,
+      "learning_rate": 7.641028798021197e-06,
+      "loss": 0.6339,
+      "step": 1840
+    },
+    {
+      "epoch": 1.2399193548387097,
+      "grad_norm": 0.41654508596227274,
+      "learning_rate": 7.584069504253703e-06,
+      "loss": 0.6316,
+      "step": 1845
+    },
+    {
+      "epoch": 1.243279569892473,
+      "grad_norm": 0.4186537736756235,
+      "learning_rate": 7.527193329771334e-06,
+      "loss": 0.629,
+      "step": 1850
+    },
+    {
+      "epoch": 1.2466397849462365,
+      "grad_norm": 0.40684069462084127,
+      "learning_rate": 7.470402231379961e-06,
+      "loss": 0.6294,
+      "step": 1855
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4491543755620313,
+      "learning_rate": 7.41369816295844e-06,
+      "loss": 0.6283,
+      "step": 1860
+    },
+    {
+      "epoch": 1.2533602150537635,
+      "grad_norm": 0.43910414887550725,
+      "learning_rate": 7.357083075391373e-06,
+      "loss": 0.6297,
+      "step": 1865
+    },
+    {
+      "epoch": 1.256720430107527,
+      "grad_norm": 0.4242078353209094,
+      "learning_rate": 7.300558916502028e-06,
+      "loss": 0.6274,
+      "step": 1870
+    },
+    {
+      "epoch": 1.2600806451612903,
+      "grad_norm": 0.39181603159793815,
+      "learning_rate": 7.24412763098528e-06,
+      "loss": 0.6268,
+      "step": 1875
+    },
+    {
+      "epoch": 1.2634408602150538,
+      "grad_norm": 0.4093720135553839,
+      "learning_rate": 7.1877911603407446e-06,
+      "loss": 0.6314,
+      "step": 1880
+    },
+    {
+      "epoch": 1.2668010752688172,
+      "grad_norm": 0.4542985585511893,
+      "learning_rate": 7.131551442805957e-06,
+      "loss": 0.6257,
+      "step": 1885
+    },
+    {
+      "epoch": 1.2701612903225805,
+      "grad_norm": 0.39320303904121556,
+      "learning_rate": 7.075410413289687e-06,
+      "loss": 0.632,
+      "step": 1890
+    },
+    {
+      "epoch": 1.273521505376344,
+      "grad_norm": 0.43112292294597837,
+      "learning_rate": 7.019370003305383e-06,
+      "loss": 0.6258,
+      "step": 1895
+    },
+    {
+      "epoch": 1.2768817204301075,
+      "grad_norm": 0.4121211812097676,
+      "learning_rate": 6.963432140904718e-06,
+      "loss": 0.6256,
+      "step": 1900
+    },
+    {
+      "epoch": 1.280241935483871,
+      "grad_norm": 0.4366695143108501,
+      "learning_rate": 6.9075987506112305e-06,
+      "loss": 0.6306,
+      "step": 1905
+    },
+    {
+      "epoch": 1.2836021505376345,
+      "grad_norm": 0.420907190197836,
+      "learning_rate": 6.851871753354154e-06,
+      "loss": 0.6314,
+      "step": 1910
+    },
+    {
+      "epoch": 1.286962365591398,
+      "grad_norm": 0.4473858705160112,
+      "learning_rate": 6.796253066402282e-06,
+      "loss": 0.627,
+      "step": 1915
+    },
+    {
+      "epoch": 1.2903225806451613,
+      "grad_norm": 0.38612233770180776,
+      "learning_rate": 6.740744603298046e-06,
+      "loss": 0.629,
+      "step": 1920
+    },
+    {
+      "epoch": 1.2936827956989247,
+      "grad_norm": 0.39974399253863463,
+      "learning_rate": 6.685348273791661e-06,
+      "loss": 0.6287,
+      "step": 1925
+    },
+    {
+      "epoch": 1.2970430107526882,
+      "grad_norm": 0.3898544394591236,
+      "learning_rate": 6.630065983775406e-06,
+      "loss": 0.6292,
+      "step": 1930
+    },
+    {
+      "epoch": 1.3004032258064515,
+      "grad_norm": 0.40401698025117533,
+      "learning_rate": 6.574899635218091e-06,
+      "loss": 0.6304,
+      "step": 1935
+    },
+    {
+      "epoch": 1.303763440860215,
+      "grad_norm": 0.41519450649319134,
+      "learning_rate": 6.519851126099586e-06,
+      "loss": 0.6287,
+      "step": 1940
+    },
+    {
+      "epoch": 1.3071236559139785,
+      "grad_norm": 0.381670315636576,
+      "learning_rate": 6.464922350345534e-06,
+      "loss": 0.6278,
+      "step": 1945
+    },
+    {
+      "epoch": 1.310483870967742,
+      "grad_norm": 0.4365713923302317,
+      "learning_rate": 6.4101151977622015e-06,
+      "loss": 0.6258,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3138440860215055,
+      "grad_norm": 0.41198291276336774,
+      "learning_rate": 6.3554315539714364e-06,
+      "loss": 0.6261,
+      "step": 1955
+    },
+    {
+      "epoch": 1.3172043010752688,
+      "grad_norm": 0.4118460369714954,
+      "learning_rate": 6.300873300345819e-06,
+      "loss": 0.6245,
+      "step": 1960
+    },
+    {
+      "epoch": 1.3205645161290323,
+      "grad_norm": 0.38775383959228366,
+      "learning_rate": 6.246442313943917e-06,
+      "loss": 0.6258,
+      "step": 1965
+    },
+    {
+      "epoch": 1.3239247311827957,
+      "grad_norm": 0.4088425446696767,
+      "learning_rate": 6.192140467445712e-06,
+      "loss": 0.6291,
+      "step": 1970
+    },
+    {
+      "epoch": 1.327284946236559,
+      "grad_norm": 0.368749879948908,
+      "learning_rate": 6.137969629088174e-06,
+      "loss": 0.629,
+      "step": 1975
+    },
+    {
+      "epoch": 1.3306451612903225,
+      "grad_norm": 0.3907700123306008,
+      "learning_rate": 6.083931662600977e-06,
+      "loss": 0.6287,
+      "step": 1980
+    },
+    {
+      "epoch": 1.334005376344086,
+      "grad_norm": 0.3975024031852529,
+      "learning_rate": 6.0300284271423834e-06,
+      "loss": 0.6227,
+      "step": 1985
+    },
+    {
+      "epoch": 1.3373655913978495,
+      "grad_norm": 0.4075706328227891,
+      "learning_rate": 5.976261777235282e-06,
+      "loss": 0.6248,
+      "step": 1990
+    },
+    {
+      "epoch": 1.340725806451613,
+      "grad_norm": 0.38742274042231606,
+      "learning_rate": 5.922633562703375e-06,
+      "loss": 0.6263,
+      "step": 1995
+    },
+    {
+      "epoch": 1.3440860215053765,
+      "grad_norm": 0.4225129189994307,
+      "learning_rate": 5.869145628607551e-06,
+      "loss": 0.6263,
+      "step": 2000
+    },
+    {
+      "epoch": 1.3474462365591398,
+      "grad_norm": 0.3905401995286742,
+      "learning_rate": 5.815799815182393e-06,
+      "loss": 0.6293,
+      "step": 2005
+    },
+    {
+      "epoch": 1.3508064516129032,
+      "grad_norm": 0.40752390323741944,
+      "learning_rate": 5.762597957772856e-06,
+      "loss": 0.6287,
+      "step": 2010
+    },
+    {
+      "epoch": 1.3541666666666667,
+      "grad_norm": 0.4096029958781252,
+      "learning_rate": 5.709541886771167e-06,
+      "loss": 0.6266,
+      "step": 2015
+    },
+    {
+      "epoch": 1.35752688172043,
+      "grad_norm": 0.38367545152375937,
+      "learning_rate": 5.656633427553784e-06,
+      "loss": 0.6303,
+      "step": 2020
+    },
+    {
+      "epoch": 1.3608870967741935,
+      "grad_norm": 0.3848396403700666,
+      "learning_rate": 5.60387440041866e-06,
+      "loss": 0.6274,
+      "step": 2025
+    },
+    {
+      "epoch": 1.364247311827957,
+      "grad_norm": 0.3871009890350265,
+      "learning_rate": 5.55126662052257e-06,
+      "loss": 0.6224,
+      "step": 2030
+    },
+    {
+      "epoch": 1.3676075268817205,
+      "grad_norm": 0.4065087733067794,
+      "learning_rate": 5.498811897818685e-06,
+      "loss": 0.6268,
+      "step": 2035
+    },
+    {
+      "epoch": 1.370967741935484,
+      "grad_norm": 0.38515450629994064,
+      "learning_rate": 5.446512036994287e-06,
+      "loss": 0.6248,
+      "step": 2040
+    },
+    {
+      "epoch": 1.3743279569892473,
+      "grad_norm": 0.39646022406843767,
+      "learning_rate": 5.394368837408705e-06,
+      "loss": 0.6249,
+      "step": 2045
+    },
+    {
+      "epoch": 1.3776881720430108,
+      "grad_norm": 0.4140801017591075,
+      "learning_rate": 5.342384093031361e-06,
+      "loss": 0.6264,
+      "step": 2050
+    },
+    {
+      "epoch": 1.3810483870967742,
+      "grad_norm": 0.41497129669099586,
+      "learning_rate": 5.290559592380111e-06,
+      "loss": 0.6249,
+      "step": 2055
+    },
+    {
+      "epoch": 1.3844086021505375,
+      "grad_norm": 0.3924624828028891,
+      "learning_rate": 5.238897118459644e-06,
+      "loss": 0.6258,
+      "step": 2060
+    },
+    {
+      "epoch": 1.387768817204301,
+      "grad_norm": 0.3702574417131231,
+      "learning_rate": 5.187398448700205e-06,
+      "loss": 0.6284,
+      "step": 2065
+    },
+    {
+      "epoch": 1.3911290322580645,
+      "grad_norm": 0.41653378611493,
+      "learning_rate": 5.136065354896393e-06,
+      "loss": 0.6263,
+      "step": 2070
+    },
+    {
+      "epoch": 1.394489247311828,
+      "grad_norm": 0.3895372191639466,
+      "learning_rate": 5.084899603146228e-06,
+      "loss": 0.629,
+      "step": 2075
+    },
+    {
+      "epoch": 1.3978494623655915,
+      "grad_norm": 0.3906095795336125,
+      "learning_rate": 5.033902953790375e-06,
+      "loss": 0.6258,
+      "step": 2080
+    },
+    {
+      "epoch": 1.4012096774193548,
+      "grad_norm": 0.39681528851984266,
+      "learning_rate": 4.983077161351606e-06,
+      "loss": 0.6282,
+      "step": 2085
+    },
+    {
+      "epoch": 1.4045698924731183,
+      "grad_norm": 0.40072163095580204,
+      "learning_rate": 4.932423974474386e-06,
+      "loss": 0.6288,
+      "step": 2090
+    },
+    {
+      "epoch": 1.4079301075268817,
+      "grad_norm": 0.41393522692230517,
+      "learning_rate": 4.8819451358647806e-06,
+      "loss": 0.6269,
+      "step": 2095
+    },
+    {
+      "epoch": 1.4112903225806452,
+      "grad_norm": 0.41100607384229376,
+      "learning_rate": 4.831642382230424e-06,
+      "loss": 0.6237,
+      "step": 2100
+    },
+    {
+      "epoch": 1.4146505376344085,
+      "grad_norm": 0.3869191806592057,
+      "learning_rate": 4.781517444220836e-06,
+      "loss": 0.6262,
+      "step": 2105
+    },
+    {
+      "epoch": 1.418010752688172,
+      "grad_norm": 0.40790781380796975,
+      "learning_rate": 4.731572046367832e-06,
+      "loss": 0.629,
+      "step": 2110
+    },
+    {
+      "epoch": 1.4213709677419355,
+      "grad_norm": 0.39391109017587794,
+      "learning_rate": 4.681807907026214e-06,
+      "loss": 0.6241,
+      "step": 2115
+    },
+    {
+      "epoch": 1.424731182795699,
+      "grad_norm": 0.40179170742485937,
+      "learning_rate": 4.632226738314638e-06,
+      "loss": 0.6255,
+      "step": 2120
+    },
+    {
+      "epoch": 1.4280913978494625,
+      "grad_norm": 0.3858060561361162,
+      "learning_rate": 4.582830246056735e-06,
+      "loss": 0.6236,
+      "step": 2125
+    },
+    {
+      "epoch": 1.4314516129032258,
+      "grad_norm": 0.38374047682817425,
+      "learning_rate": 4.533620129722376e-06,
+      "loss": 0.6204,
+      "step": 2130
+    },
+    {
+      "epoch": 1.4348118279569892,
+      "grad_norm": 0.3973135891040321,
+      "learning_rate": 4.4845980823692605e-06,
+      "loss": 0.6238,
+      "step": 2135
+    },
+    {
+      "epoch": 1.4381720430107527,
+      "grad_norm": 0.3883947152151204,
+      "learning_rate": 4.435765790584605e-06,
+      "loss": 0.6245,
+      "step": 2140
+    },
+    {
+      "epoch": 1.441532258064516,
+      "grad_norm": 0.3991192638455965,
+      "learning_rate": 4.387124934427181e-06,
+      "loss": 0.6235,
+      "step": 2145
+    },
+    {
+      "epoch": 1.4448924731182795,
+      "grad_norm": 0.38536161760649323,
+      "learning_rate": 4.338677187369458e-06,
+      "loss": 0.6239,
+      "step": 2150
+    },
+    {
+      "epoch": 1.448252688172043,
+      "grad_norm": 0.38959778197902006,
+      "learning_rate": 4.290424216240062e-06,
+      "loss": 0.624,
+      "step": 2155
+    },
+    {
+      "epoch": 1.4516129032258065,
+      "grad_norm": 0.3932900684900842,
+      "learning_rate": 4.242367681166414e-06,
+      "loss": 0.6261,
+      "step": 2160
+    },
+    {
+      "epoch": 1.45497311827957,
+      "grad_norm": 0.39269029723244625,
+      "learning_rate": 4.19450923551762e-06,
+      "loss": 0.6217,
+      "step": 2165
+    },
+    {
+      "epoch": 1.4583333333333333,
+      "grad_norm": 0.3849808642498751,
+      "learning_rate": 4.1468505258475785e-06,
+      "loss": 0.6205,
+      "step": 2170
+    },
+    {
+      "epoch": 1.4616935483870968,
+      "grad_norm": 0.37433760170956354,
+      "learning_rate": 4.0993931918383556e-06,
+      "loss": 0.6181,
+      "step": 2175
+    },
+    {
+      "epoch": 1.4650537634408602,
+      "grad_norm": 0.38330212314619344,
+      "learning_rate": 4.0521388662437285e-06,
+      "loss": 0.6243,
+      "step": 2180
+    },
+    {
+      "epoch": 1.4684139784946235,
+      "grad_norm": 0.3934565346806046,
+      "learning_rate": 4.0050891748330636e-06,
+      "loss": 0.6253,
+      "step": 2185
+    },
+    {
+      "epoch": 1.471774193548387,
+      "grad_norm": 0.3829851450390039,
+      "learning_rate": 3.958245736335337e-06,
+      "loss": 0.625,
+      "step": 2190
+    },
+    {
+      "epoch": 1.4751344086021505,
+      "grad_norm": 0.38802181960954035,
+      "learning_rate": 3.911610162383475e-06,
+      "loss": 0.6265,
+      "step": 2195
+    },
+    {
+      "epoch": 1.478494623655914,
+      "grad_norm": 0.410711397472092,
+      "learning_rate": 3.865184057458883e-06,
+      "loss": 0.6232,
+      "step": 2200
+    },
+    {
+      "epoch": 1.4818548387096775,
+      "grad_norm": 0.3894505260206867,
+      "learning_rate": 3.8189690188362615e-06,
+      "loss": 0.628,
+      "step": 2205
+    },
+    {
+      "epoch": 1.485215053763441,
+      "grad_norm": 0.385060920155747,
+      "learning_rate": 3.772966636528641e-06,
+      "loss": 0.6186,
+      "step": 2210
+    },
+    {
+      "epoch": 1.4885752688172043,
+      "grad_norm": 0.39063304189411735,
+      "learning_rate": 3.727178493232685e-06,
+      "loss": 0.6236,
+      "step": 2215
+    },
+    {
+      "epoch": 1.4919354838709677,
+      "grad_norm": 0.3690141904300633,
+      "learning_rate": 3.6816061642742294e-06,
+      "loss": 0.6251,
+      "step": 2220
+    },
+    {
+      "epoch": 1.4952956989247312,
+      "grad_norm": 0.3915588795933579,
+      "learning_rate": 3.6362512175541008e-06,
+      "loss": 0.6231,
+      "step": 2225
+    },
+    {
+      "epoch": 1.4986559139784945,
+      "grad_norm": 0.3823928248559548,
+      "learning_rate": 3.591115213494153e-06,
+      "loss": 0.6243,
+      "step": 2230
+    },
+    {
+      "epoch": 1.502016129032258,
+      "grad_norm": 0.375431342074985,
+      "learning_rate": 3.546199704983592e-06,
+      "loss": 0.619,
+      "step": 2235
+    },
+    {
+      "epoch": 1.5053763440860215,
+      "grad_norm": 0.3964171676918981,
+      "learning_rate": 3.501506237325547e-06,
+      "loss": 0.6221,
+      "step": 2240
+    },
+    {
+      "epoch": 1.508736559139785,
+      "grad_norm": 0.3850008299468121,
+      "learning_rate": 3.4570363481839097e-06,
+      "loss": 0.6223,
+      "step": 2245
+    },
+    {
+      "epoch": 1.5120967741935485,
+      "grad_norm": 0.39821907737468476,
+      "learning_rate": 3.4127915675304236e-06,
+      "loss": 0.6251,
+      "step": 2250
+    },
+    {
+      "epoch": 1.515456989247312,
+      "grad_norm": 0.4127390828177173,
+      "learning_rate": 3.3687734175920505e-06,
+      "loss": 0.6235,
+      "step": 2255
+    },
+    {
+      "epoch": 1.5188172043010753,
+      "grad_norm": 0.3900524242402812,
+      "learning_rate": 3.324983412798597e-06,
+      "loss": 0.6228,
+      "step": 2260
+    },
+    {
+      "epoch": 1.5221774193548387,
+      "grad_norm": 0.39215462329521156,
+      "learning_rate": 3.281423059730612e-06,
+      "loss": 0.6272,
+      "step": 2265
+    },
+    {
+      "epoch": 1.525537634408602,
+      "grad_norm": 0.37938495139611367,
+      "learning_rate": 3.238093857067558e-06,
+      "loss": 0.624,
+      "step": 2270
+    },
+    {
+      "epoch": 1.5288978494623655,
+      "grad_norm": 0.38603540777357676,
+      "learning_rate": 3.1949972955362383e-06,
+      "loss": 0.6239,
+      "step": 2275
+    },
+    {
+      "epoch": 1.532258064516129,
+      "grad_norm": 0.3863102533573958,
+      "learning_rate": 3.1521348578595178e-06,
+      "loss": 0.6242,
+      "step": 2280
+    },
+    {
+      "epoch": 1.5356182795698925,
+      "grad_norm": 0.38583036503521356,
+      "learning_rate": 3.1095080187053084e-06,
+      "loss": 0.62,
+      "step": 2285
+    },
+    {
+      "epoch": 1.538978494623656,
+      "grad_norm": 0.38861760389469036,
+      "learning_rate": 3.067118244635833e-06,
+      "loss": 0.6211,
+      "step": 2290
+    },
+    {
+      "epoch": 1.5423387096774195,
+      "grad_norm": 0.3858262856892166,
+      "learning_rate": 3.024966994057168e-06,
+      "loss": 0.6193,
+      "step": 2295
+    },
+    {
+      "epoch": 1.5456989247311828,
+      "grad_norm": 0.39743062676203483,
+      "learning_rate": 2.98305571716907e-06,
+      "loss": 0.6218,
+      "step": 2300
+    },
+    {
+      "epoch": 1.5490591397849462,
+      "grad_norm": 0.3883508569759508,
+      "learning_rate": 2.9413858559150776e-06,
+      "loss": 0.6217,
+      "step": 2305
+    },
+    {
+      "epoch": 1.5524193548387095,
+      "grad_norm": 0.38516305909935,
+      "learning_rate": 2.899958843932915e-06,
+      "loss": 0.6202,
+      "step": 2310
+    },
+    {
+      "epoch": 1.555779569892473,
+      "grad_norm": 0.39366530353726675,
+      "learning_rate": 2.858776106505139e-06,
+      "loss": 0.6194,
+      "step": 2315
+    },
+    {
+      "epoch": 1.5591397849462365,
+      "grad_norm": 0.3773278542481648,
+      "learning_rate": 2.8178390605101414e-06,
+      "loss": 0.6223,
+      "step": 2320
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 0.4009756258267962,
+      "learning_rate": 2.777149114373371e-06,
+      "loss": 0.6195,
+      "step": 2325
+    },
+    {
+      "epoch": 1.5658602150537635,
+      "grad_norm": 0.3858950731460861,
+      "learning_rate": 2.73670766801889e-06,
+      "loss": 0.6217,
+      "step": 2330
+    },
+    {
+      "epoch": 1.569220430107527,
+      "grad_norm": 0.3971612319470807,
+      "learning_rate": 2.696516112821208e-06,
+      "loss": 0.622,
+      "step": 2335
+    },
+    {
+      "epoch": 1.5725806451612905,
+      "grad_norm": 0.38420387771892894,
+      "learning_rate": 2.6565758315574152e-06,
+      "loss": 0.6201,
+      "step": 2340
+    },
+    {
+      "epoch": 1.5759408602150538,
+      "grad_norm": 0.3933992088451066,
+      "learning_rate": 2.6168881983595994e-06,
+      "loss": 0.6226,
+      "step": 2345
+    },
+    {
+      "epoch": 1.5793010752688172,
+      "grad_norm": 0.36479667118510656,
+      "learning_rate": 2.5774545786675887e-06,
+      "loss": 0.6175,
+      "step": 2350
+    },
+    {
+      "epoch": 1.5826612903225805,
+      "grad_norm": 0.37711761279306893,
+      "learning_rate": 2.538276329181942e-06,
+      "loss": 0.6206,
+      "step": 2355
+    },
+    {
+      "epoch": 1.586021505376344,
+      "grad_norm": 0.3816373703635669,
+      "learning_rate": 2.499354797817312e-06,
+      "loss": 0.6205,
+      "step": 2360
+    },
+    {
+      "epoch": 1.5893817204301075,
+      "grad_norm": 0.38654799097580955,
+      "learning_rate": 2.4606913236560283e-06,
+      "loss": 0.6245,
+      "step": 2365
+    },
+    {
+      "epoch": 1.592741935483871,
+      "grad_norm": 0.37621383637459765,
+      "learning_rate": 2.4222872369020676e-06,
+      "loss": 0.6237,
+      "step": 2370
+    },
+    {
+      "epoch": 1.5961021505376345,
+      "grad_norm": 0.38574402970547833,
+      "learning_rate": 2.384143858835258e-06,
+      "loss": 0.6198,
+      "step": 2375
+    },
+    {
+      "epoch": 1.599462365591398,
+      "grad_norm": 0.41266081486593603,
+      "learning_rate": 2.3462625017658356e-06,
+      "loss": 0.6203,
+      "step": 2380
+    },
+    {
+      "epoch": 1.6028225806451613,
+      "grad_norm": 0.39723116620215,
+      "learning_rate": 2.3086444689892872e-06,
+      "loss": 0.6224,
+      "step": 2385
+    },
+    {
+      "epoch": 1.6061827956989247,
+      "grad_norm": 0.39889771375855315,
+      "learning_rate": 2.2712910547415266e-06,
+      "loss": 0.6228,
+      "step": 2390
+    },
+    {
+      "epoch": 1.609543010752688,
+      "grad_norm": 0.38397107249589707,
+      "learning_rate": 2.234203544154335e-06,
+      "loss": 0.6202,
+      "step": 2395
+    },
+    {
+      "epoch": 1.6129032258064515,
+      "grad_norm": 0.3938406367129855,
+      "learning_rate": 2.1973832132111906e-06,
+      "loss": 0.6194,
+      "step": 2400
+    },
+    {
+      "epoch": 1.616263440860215,
+      "grad_norm": 0.37138500127297974,
+      "learning_rate": 2.16083132870332e-06,
+      "loss": 0.6226,
+      "step": 2405
+    },
+    {
+      "epoch": 1.6196236559139785,
+      "grad_norm": 0.386156003115288,
+      "learning_rate": 2.1245491481861615e-06,
+      "loss": 0.6174,
+      "step": 2410
+    },
+    {
+      "epoch": 1.622983870967742,
+      "grad_norm": 0.3747547614429102,
+      "learning_rate": 2.0885379199360646e-06,
+      "loss": 0.6228,
+      "step": 2415
+    },
+    {
+      "epoch": 1.6263440860215055,
+      "grad_norm": 0.3784295571335933,
+      "learning_rate": 2.0527988829073587e-06,
+      "loss": 0.6208,
+      "step": 2420
+    },
+    {
+      "epoch": 1.629704301075269,
+      "grad_norm": 0.3741384831092738,
+      "learning_rate": 2.0173332666897227e-06,
+      "loss": 0.6205,
+      "step": 2425
+    },
+    {
+      "epoch": 1.6330645161290323,
+      "grad_norm": 0.38786521686382086,
+      "learning_rate": 1.982142291465896e-06,
+      "loss": 0.6177,
+      "step": 2430
+    },
+    {
+      "epoch": 1.6364247311827957,
+      "grad_norm": 0.38553674654791287,
+      "learning_rate": 1.947227167969663e-06,
+      "loss": 0.6219,
+      "step": 2435
+    },
+    {
+      "epoch": 1.639784946236559,
+      "grad_norm": 0.37179654564247894,
+      "learning_rate": 1.9125890974442475e-06,
+      "loss": 0.62,
+      "step": 2440
+    },
+    {
+      "epoch": 1.6431451612903225,
+      "grad_norm": 0.3705380721255442,
+      "learning_rate": 1.878229271600931e-06,
+      "loss": 0.6212,
+      "step": 2445
+    },
+    {
+      "epoch": 1.646505376344086,
+      "grad_norm": 0.37577075286504236,
+      "learning_rate": 1.8441488725781043e-06,
+      "loss": 0.6219,
+      "step": 2450
+    },
+    {
+      "epoch": 1.6498655913978495,
+      "grad_norm": 0.37950348342322227,
+      "learning_rate": 1.8103490729005546e-06,
+      "loss": 0.6168,
+      "step": 2455
+    },
+    {
+      "epoch": 1.653225806451613,
+      "grad_norm": 0.3763349869752889,
+      "learning_rate": 1.7768310354391505e-06,
+      "loss": 0.6221,
+      "step": 2460
+    },
+    {
+      "epoch": 1.6565860215053765,
+      "grad_norm": 0.3815949507602912,
+      "learning_rate": 1.7435959133708169e-06,
+      "loss": 0.6175,
+      "step": 2465
+    },
+    {
+      "epoch": 1.6599462365591398,
+      "grad_norm": 0.38010076668823534,
+      "learning_rate": 1.7106448501388827e-06,
+      "loss": 0.6176,
+      "step": 2470
+    },
+    {
+      "epoch": 1.6633064516129032,
+      "grad_norm": 0.3691580309508141,
+      "learning_rate": 1.677978979413708e-06,
+      "loss": 0.6164,
+      "step": 2475
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.38980910582544764,
+      "learning_rate": 1.645599425053721e-06,
+      "loss": 0.6192,
+      "step": 2480
+    },
+    {
+      "epoch": 1.67002688172043,
+      "grad_norm": 0.3915443888876836,
+      "learning_rate": 1.6135073010667091e-06,
+      "loss": 0.6178,
+      "step": 2485
+    },
+    {
+      "epoch": 1.6733870967741935,
+      "grad_norm": 0.38991481296941843,
+      "learning_rate": 1.5817037115715307e-06,
+      "loss": 0.6192,
+      "step": 2490
+    },
+    {
+      "epoch": 1.676747311827957,
+      "grad_norm": 0.3792230038765017,
+      "learning_rate": 1.5501897507601016e-06,
+      "loss": 0.6218,
+      "step": 2495
+    },
+    {
+      "epoch": 1.6801075268817205,
+      "grad_norm": 0.37563408321675024,
+      "learning_rate": 1.5189665028597622e-06,
+      "loss": 0.6201,
+      "step": 2500
+    },
+    {
+      "epoch": 1.683467741935484,
+      "grad_norm": 0.3782912838928693,
+      "learning_rate": 1.4880350420959678e-06,
+      "loss": 0.6148,
+      "step": 2505
+    },
+    {
+      "epoch": 1.6868279569892473,
+      "grad_norm": 0.37676394416952275,
+      "learning_rate": 1.4573964326553447e-06,
+      "loss": 0.6213,
+      "step": 2510
+    },
+    {
+      "epoch": 1.6901881720430108,
+      "grad_norm": 0.4026075731771035,
+      "learning_rate": 1.4270517286490526e-06,
+      "loss": 0.6212,
+      "step": 2515
+    },
+    {
+      "epoch": 1.6935483870967742,
+      "grad_norm": 0.3740041929756673,
+      "learning_rate": 1.397001974076546e-06,
+      "loss": 0.6173,
+      "step": 2520
+    },
+    {
+      "epoch": 1.6969086021505375,
+      "grad_norm": 0.37487781663955083,
+      "learning_rate": 1.3672482027896295e-06,
+      "loss": 0.6157,
+      "step": 2525
+    },
+    {
+      "epoch": 1.700268817204301,
+      "grad_norm": 0.3658110013443917,
+      "learning_rate": 1.3377914384569124e-06,
+      "loss": 0.6197,
+      "step": 2530
+    },
+    {
+      "epoch": 1.7036290322580645,
+      "grad_norm": 0.3698169716895195,
+      "learning_rate": 1.3086326945285721e-06,
+      "loss": 0.6172,
+      "step": 2535
+    },
+    {
+      "epoch": 1.706989247311828,
+      "grad_norm": 0.3726837636042825,
+      "learning_rate": 1.2797729742014952e-06,
+      "loss": 0.6239,
+      "step": 2540
+    },
+    {
+      "epoch": 1.7103494623655915,
+      "grad_norm": 0.3776863339195105,
+      "learning_rate": 1.251213270384759e-06,
+      "loss": 0.6166,
+      "step": 2545
+    },
+    {
+      "epoch": 1.713709677419355,
+      "grad_norm": 0.3701247417408879,
+      "learning_rate": 1.2229545656654784e-06,
+      "loss": 0.6195,
+      "step": 2550
+    },
+    {
+      "epoch": 1.7170698924731183,
+      "grad_norm": 0.37292321840424403,
+      "learning_rate": 1.1949978322749833e-06,
+      "loss": 0.62,
+      "step": 2555
+    },
+    {
+      "epoch": 1.7204301075268817,
+      "grad_norm": 0.36380895900137356,
+      "learning_rate": 1.1673440320553941e-06,
+      "loss": 0.6208,
+      "step": 2560
+    },
+    {
+      "epoch": 1.723790322580645,
+      "grad_norm": 0.3753481673103304,
+      "learning_rate": 1.1399941164265016e-06,
+      "loss": 0.6197,
+      "step": 2565
+    },
+    {
+      "epoch": 1.7271505376344085,
+      "grad_norm": 0.38841007090968127,
+      "learning_rate": 1.112949026353063e-06,
+      "loss": 0.6197,
+      "step": 2570
+    },
+    {
+      "epoch": 1.730510752688172,
+      "grad_norm": 0.4024753571663661,
+      "learning_rate": 1.0862096923124032e-06,
+      "loss": 0.6195,
+      "step": 2575
+    },
+    {
+      "epoch": 1.7338709677419355,
+      "grad_norm": 0.40698102024501104,
+      "learning_rate": 1.0597770342624169e-06,
+      "loss": 0.6212,
+      "step": 2580
+    },
+    {
+      "epoch": 1.737231182795699,
+      "grad_norm": 0.38275795281843294,
+      "learning_rate": 1.0336519616099127e-06,
+      "loss": 0.6164,
+      "step": 2585
+    },
+    {
+      "epoch": 1.7405913978494625,
+      "grad_norm": 0.3849617861720321,
+      "learning_rate": 1.0078353731793245e-06,
+      "loss": 0.621,
+      "step": 2590
+    },
+    {
+      "epoch": 1.7439516129032258,
+      "grad_norm": 0.37357311882407007,
+      "learning_rate": 9.823281571817888e-07,
+      "loss": 0.6154,
+      "step": 2595
+    },
+    {
+      "epoch": 1.7473118279569892,
+      "grad_norm": 0.3700469086083545,
+      "learning_rate": 9.571311911845938e-07,
+      "loss": 0.6168,
+      "step": 2600
+    },
+    {
+      "epoch": 1.7506720430107527,
+      "grad_norm": 0.3720192733243154,
+      "learning_rate": 9.322453420809663e-07,
+      "loss": 0.6188,
+      "step": 2605
+    },
+    {
+      "epoch": 1.754032258064516,
+      "grad_norm": 0.3759460278402678,
+      "learning_rate": 9.076714660602726e-07,
+      "loss": 0.6188,
+      "step": 2610
+    },
+    {
+      "epoch": 1.7573924731182795,
+      "grad_norm": 0.37780059312461417,
+      "learning_rate": 8.834104085785411e-07,
+      "loss": 0.619,
+      "step": 2615
+    },
+    {
+      "epoch": 1.760752688172043,
+      "grad_norm": 0.3644735938402259,
+      "learning_rate": 8.594630043293862e-07,
+      "loss": 0.6166,
+      "step": 2620
+    },
+    {
+      "epoch": 1.7641129032258065,
+      "grad_norm": 0.40102016373518057,
+      "learning_rate": 8.35830077215285e-07,
+      "loss": 0.6205,
+      "step": 2625
+    },
+    {
+      "epoch": 1.76747311827957,
+      "grad_norm": 0.36805712689178754,
+      "learning_rate": 8.125124403192353e-07,
+      "loss": 0.6178,
+      "step": 2630
+    },
+    {
+      "epoch": 1.7708333333333335,
+      "grad_norm": 0.3855167813689446,
+      "learning_rate": 7.89510895876775e-07,
+      "loss": 0.6187,
+      "step": 2635
+    },
+    {
+      "epoch": 1.7741935483870968,
+      "grad_norm": 0.37050943347616344,
+      "learning_rate": 7.668262352483957e-07,
+      "loss": 0.6175,
+      "step": 2640
+    },
+    {
+      "epoch": 1.7775537634408602,
+      "grad_norm": 0.37029783353120194,
+      "learning_rate": 7.444592388922955e-07,
+      "loss": 0.6184,
+      "step": 2645
+    },
+    {
+      "epoch": 1.7809139784946235,
+      "grad_norm": 0.3842300371858157,
+      "learning_rate": 7.224106763375493e-07,
+      "loss": 0.6147,
+      "step": 2650
+    },
+    {
+      "epoch": 1.784274193548387,
+      "grad_norm": 0.3700714291963043,
+      "learning_rate": 7.006813061576145e-07,
+      "loss": 0.6124,
+      "step": 2655
+    },
+    {
+      "epoch": 1.7876344086021505,
+      "grad_norm": 0.3760694332290439,
+      "learning_rate": 6.792718759442474e-07,
+      "loss": 0.6194,
+      "step": 2660
+    },
+    {
+      "epoch": 1.790994623655914,
+      "grad_norm": 0.3671308826023182,
+      "learning_rate": 6.581831222817714e-07,
+      "loss": 0.6149,
+      "step": 2665
+    },
+    {
+      "epoch": 1.7943548387096775,
+      "grad_norm": 0.3746985320397756,
+      "learning_rate": 6.374157707217421e-07,
+      "loss": 0.6152,
+      "step": 2670
+    },
+    {
+      "epoch": 1.797715053763441,
+      "grad_norm": 0.38527941649428066,
+      "learning_rate": 6.169705357579813e-07,
+      "loss": 0.6206,
+      "step": 2675
+    },
+    {
+      "epoch": 1.8010752688172043,
+      "grad_norm": 0.3703868711176167,
+      "learning_rate": 5.968481208020016e-07,
+      "loss": 0.6185,
+      "step": 2680
+    },
+    {
+      "epoch": 1.8044354838709677,
+      "grad_norm": 0.37853482566243507,
+      "learning_rate": 5.770492181587906e-07,
+      "loss": 0.6148,
+      "step": 2685
+    },
+    {
+      "epoch": 1.807795698924731,
+      "grad_norm": 0.3580208698159626,
+      "learning_rate": 5.575745090030138e-07,
+      "loss": 0.6168,
+      "step": 2690
+    },
+    {
+      "epoch": 1.8111559139784945,
+      "grad_norm": 0.3822688397822238,
+      "learning_rate": 5.38424663355559e-07,
+      "loss": 0.6188,
+      "step": 2695
+    },
+    {
+      "epoch": 1.814516129032258,
+      "grad_norm": 0.37115606254622296,
+      "learning_rate": 5.196003400604977e-07,
+      "loss": 0.6159,
+      "step": 2700
+    },
+    {
+      "epoch": 1.8178763440860215,
+      "grad_norm": 0.37498937385752057,
+      "learning_rate": 5.01102186762411e-07,
+      "loss": 0.6193,
+      "step": 2705
+    },
+    {
+      "epoch": 1.821236559139785,
+      "grad_norm": 0.36638337590912584,
+      "learning_rate": 4.829308398841104e-07,
+      "loss": 0.6177,
+      "step": 2710
+    },
+    {
+      "epoch": 1.8245967741935485,
+      "grad_norm": 0.3720335399527678,
+      "learning_rate": 4.650869246047407e-07,
+      "loss": 0.616,
+      "step": 2715
+    },
+    {
+      "epoch": 1.827956989247312,
+      "grad_norm": 0.3883742285519788,
+      "learning_rate": 4.475710548382717e-07,
+      "loss": 0.6183,
+      "step": 2720
+    },
+    {
+      "epoch": 1.8313172043010753,
+      "grad_norm": 0.3660996354157979,
+      "learning_rate": 4.303838332123766e-07,
+      "loss": 0.6181,
+      "step": 2725
+    },
+    {
+      "epoch": 1.8346774193548387,
+      "grad_norm": 0.3832317742511122,
+      "learning_rate": 4.1352585104770136e-07,
+      "loss": 0.6173,
+      "step": 2730
+    },
+    {
+      "epoch": 1.838037634408602,
+      "grad_norm": 0.3733758015066586,
+      "learning_rate": 3.969976883375126e-07,
+      "loss": 0.6212,
+      "step": 2735
+    },
+    {
+      "epoch": 1.8413978494623655,
+      "grad_norm": 0.371152480421909,
+      "learning_rate": 3.807999137277507e-07,
+      "loss": 0.6154,
+      "step": 2740
+    },
+    {
+      "epoch": 1.844758064516129,
+      "grad_norm": 0.38345844939628254,
+      "learning_rate": 3.6493308449746525e-07,
+      "loss": 0.6189,
+      "step": 2745
+    },
+    {
+      "epoch": 1.8481182795698925,
+      "grad_norm": 0.36762462712757643,
+      "learning_rate": 3.4939774653963587e-07,
+      "loss": 0.6203,
+      "step": 2750
+    },
+    {
+      "epoch": 1.851478494623656,
+      "grad_norm": 0.375536271265564,
+      "learning_rate": 3.3419443434240083e-07,
+      "loss": 0.6172,
+      "step": 2755
+    },
+    {
+      "epoch": 1.8548387096774195,
+      "grad_norm": 0.3660221694767357,
+      "learning_rate": 3.19323670970656e-07,
+      "loss": 0.6181,
+      "step": 2760
+    },
+    {
+      "epoch": 1.8581989247311828,
+      "grad_norm": 0.3599790898532119,
+      "learning_rate": 3.0478596804807246e-07,
+      "loss": 0.62,
+      "step": 2765
+    },
+    {
+      "epoch": 1.8615591397849462,
+      "grad_norm": 0.3709615221652835,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.6159,
+      "step": 2770
+    },
+    {
+      "epoch": 1.8649193548387095,
+      "grad_norm": 0.3726522140490793,
+      "learning_rate": 2.7671173273367323e-07,
+      "loss": 0.6192,
+      "step": 2775
+    },
+    {
+      "epoch": 1.868279569892473,
+      "grad_norm": 0.3677308860539174,
+      "learning_rate": 2.631761662265875e-07,
+      "loss": 0.6188,
+      "step": 2780
+    },
+    {
+      "epoch": 1.8716397849462365,
+      "grad_norm": 0.37507182923605303,
+      "learning_rate": 2.499755919048863e-07,
+      "loss": 0.6236,
+      "step": 2785
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.36561243105086355,
+      "learning_rate": 2.371104639299393e-07,
+      "loss": 0.616,
+      "step": 2790
+    },
+    {
+      "epoch": 1.8783602150537635,
+      "grad_norm": 0.3727929463786164,
+      "learning_rate": 2.2458122492219458e-07,
+      "loss": 0.6186,
+      "step": 2795
+    },
+    {
+      "epoch": 1.881720430107527,
+      "grad_norm": 0.3715493910859094,
+      "learning_rate": 2.1238830594595195e-07,
+      "loss": 0.619,
+      "step": 2800
+    },
+    {
+      "epoch": 1.8850806451612905,
+      "grad_norm": 0.3701421554724482,
+      "learning_rate": 2.005321264945348e-07,
+      "loss": 0.615,
+      "step": 2805
+    },
+    {
+      "epoch": 1.8884408602150538,
+      "grad_norm": 0.38162958713764294,
+      "learning_rate": 1.890130944758528e-07,
+      "loss": 0.6189,
+      "step": 2810
+    },
+    {
+      "epoch": 1.8918010752688172,
+      "grad_norm": 0.36878364983672696,
+      "learning_rate": 1.7783160619837202e-07,
+      "loss": 0.6157,
+      "step": 2815
+    },
+    {
+      "epoch": 1.8951612903225805,
+      "grad_norm": 0.3702088013665396,
+      "learning_rate": 1.669880463574758e-07,
+      "loss": 0.6136,
+      "step": 2820
+    },
+    {
+      "epoch": 1.898521505376344,
+      "grad_norm": 0.37799049217494324,
+      "learning_rate": 1.5648278802223526e-07,
+      "loss": 0.6145,
+      "step": 2825
+    },
+    {
+      "epoch": 1.9018817204301075,
+      "grad_norm": 0.37016035977530204,
+      "learning_rate": 1.463161926225687e-07,
+      "loss": 0.6202,
+      "step": 2830
+    },
+    {
+      "epoch": 1.905241935483871,
+      "grad_norm": 0.3683670258810398,
+      "learning_rate": 1.3648860993680903e-07,
+      "loss": 0.6146,
+      "step": 2835
+    },
+    {
+      "epoch": 1.9086021505376345,
+      "grad_norm": 0.3645705769628613,
+      "learning_rate": 1.2700037807967026e-07,
+      "loss": 0.6158,
+      "step": 2840
+    },
+    {
+      "epoch": 1.911962365591398,
+      "grad_norm": 0.3646260231379924,
+      "learning_rate": 1.1785182349061342e-07,
+      "loss": 0.6151,
+      "step": 2845
+    },
+    {
+      "epoch": 1.9153225806451613,
+      "grad_norm": 0.36357216498889616,
+      "learning_rate": 1.0904326092261441e-07,
+      "loss": 0.6173,
+      "step": 2850
+    },
+    {
+      "epoch": 1.9186827956989247,
+      "grad_norm": 0.36891941735522826,
+      "learning_rate": 1.0057499343134269e-07,
+      "loss": 0.617,
+      "step": 2855
+    },
+    {
+      "epoch": 1.922043010752688,
+      "grad_norm": 0.3618844540766792,
+      "learning_rate": 9.24473123647196e-08,
+      "loss": 0.6187,
+      "step": 2860
+    },
+    {
+      "epoch": 1.9254032258064515,
+      "grad_norm": 0.3829664112005493,
+      "learning_rate": 8.466049735291415e-08,
+      "loss": 0.6194,
+      "step": 2865
+    },
+    {
+      "epoch": 1.928763440860215,
+      "grad_norm": 0.36781144520431924,
+      "learning_rate": 7.721481629870076e-08,
+      "loss": 0.6169,
+      "step": 2870
+    },
+    {
+      "epoch": 1.9321236559139785,
+      "grad_norm": 0.3853416687052536,
+      "learning_rate": 7.011052536826435e-08,
+      "loss": 0.6205,
+      "step": 2875
+    },
+    {
+      "epoch": 1.935483870967742,
+      "grad_norm": 0.37397490558359964,
+      "learning_rate": 6.334786898237078e-08,
+      "loss": 0.6154,
+      "step": 2880
+    },
+    {
+      "epoch": 1.9388440860215055,
+      "grad_norm": 0.36691620687136894,
+      "learning_rate": 5.69270798079613e-08,
+      "loss": 0.6196,
+      "step": 2885
+    },
+    {
+      "epoch": 1.942204301075269,
+      "grad_norm": 0.3658247784366582,
+      "learning_rate": 5.084837875015347e-08,
+      "loss": 0.6132,
+      "step": 2890
+    },
+    {
+      "epoch": 1.9455645161290323,
+      "grad_norm": 0.36412582119114073,
+      "learning_rate": 4.511197494463493e-08,
+      "loss": 0.6172,
+      "step": 2895
+    },
+    {
+      "epoch": 1.9489247311827957,
+      "grad_norm": 0.3663620362729526,
+      "learning_rate": 3.971806575047033e-08,
+      "loss": 0.618,
+      "step": 2900
+    },
+    {
+      "epoch": 1.952284946236559,
+      "grad_norm": 0.3675765196570775,
+      "learning_rate": 3.466683674331228e-08,
+      "loss": 0.6166,
+      "step": 2905
+    },
+    {
+      "epoch": 1.9556451612903225,
+      "grad_norm": 0.3591512141413379,
+      "learning_rate": 2.995846170901428e-08,
+      "loss": 0.6156,
+      "step": 2910
+    },
+    {
+      "epoch": 1.959005376344086,
+      "grad_norm": 0.36646296059037736,
+      "learning_rate": 2.5593102637652136e-08,
+      "loss": 0.6145,
+      "step": 2915
+    },
+    {
+      "epoch": 1.9623655913978495,
+      "grad_norm": 0.36873568843189847,
+      "learning_rate": 2.1570909717955058e-08,
+      "loss": 0.6183,
+      "step": 2920
+    },
+    {
+      "epoch": 1.965725806451613,
+      "grad_norm": 0.3643607047509,
+      "learning_rate": 1.789202133212986e-08,
+      "loss": 0.6144,
+      "step": 2925
+    },
+    {
+      "epoch": 1.9690860215053765,
+      "grad_norm": 0.3696768969448687,
+      "learning_rate": 1.4556564051110278e-08,
+      "loss": 0.6221,
+      "step": 2930
+    },
+    {
+      "epoch": 1.9724462365591398,
+      "grad_norm": 0.3739771169389071,
+      "learning_rate": 1.1564652630192686e-08,
+      "loss": 0.6171,
+      "step": 2935
+    },
+    {
+      "epoch": 1.9758064516129032,
+      "grad_norm": 0.36410850268700906,
+      "learning_rate": 8.916390005095921e-09,
+      "loss": 0.619,
+      "step": 2940
+    },
+    {
+      "epoch": 1.9791666666666665,
+      "grad_norm": 0.3635016376462389,
+      "learning_rate": 6.61186728841412e-09,
+      "loss": 0.6121,
+      "step": 2945
+    },
+    {
+      "epoch": 1.98252688172043,
+      "grad_norm": 0.36237306270652886,
+      "learning_rate": 4.651163766484779e-09,
+      "loss": 0.6164,
+      "step": 2950
+    },
+    {
+      "epoch": 1.9858870967741935,
+      "grad_norm": 0.3732498548310209,
+      "learning_rate": 3.0343468966598234e-09,
+      "loss": 0.6169,
+      "step": 2955
+    },
+    {
+      "epoch": 1.989247311827957,
+      "grad_norm": 0.3679867274420493,
+      "learning_rate": 1.761472304987466e-09,
+      "loss": 0.618,
+      "step": 2960
+    },
+    {
+      "epoch": 1.9926075268817205,
+      "grad_norm": 0.37009481264365335,
+      "learning_rate": 8.325837842926288e-10,
+      "loss": 0.6203,
+      "step": 2965
+    },
+    {
+      "epoch": 1.995967741935484,
+      "grad_norm": 0.3760102338533851,
+      "learning_rate": 2.4771329267703206e-10,
+      "loss": 0.621,
+      "step": 2970
+    },
+    {
+      "epoch": 1.9993279569892473,
+      "grad_norm": 0.3689799921255622,
+      "learning_rate": 6.880952415633246e-12,
+      "loss": 0.6204,
+      "step": 2975
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.642189085483551,
+      "eval_runtime": 38.454,
+      "eval_samples_per_second": 75.831,
+      "eval_steps_per_second": 1.196,
+      "step": 2976
+    },
+    {
+      "epoch": 2.0,
+      "step": 2976,
+      "total_flos": 623113855303680.0,
+      "train_loss": 0.6763581057950374,
+      "train_runtime": 8673.6172,
+      "train_samples_per_second": 21.958,
+      "train_steps_per_second": 0.343
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 2976,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 623113855303680.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}