diff --git "a/checkpoint-17632/trainer_state.json" "b/checkpoint-17632/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-17632/trainer_state.json"
@@ -0,0 +1,12410 @@
+{
+  "best_metric": 0.1423337161540985,
+  "best_model_checkpoint": "./models/toxicity-scorer-qwen/checkpoint-8816",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 17632,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval_accuracy": 0.5782279248332997,
+      "eval_f1": 0.6486677195396867,
+      "eval_loss": 2.1836156845092773,
+      "eval_precision": 0.8024809024681206,
+      "eval_recall": 0.5782279248332997,
+      "eval_runtime": 121.6957,
+      "eval_samples_per_second": 488.004,
+      "eval_steps_per_second": 3.813,
+      "step": 0
+    },
+    {
+      "epoch": 0.001134301270417423,
+      "grad_norm": 75.0,
+      "learning_rate": 1.1342155009451796e-07,
+      "loss": 2.1801,
+      "step": 10
+    },
+    {
+      "epoch": 0.002268602540834846,
+      "grad_norm": 72.5,
+      "learning_rate": 2.2684310018903592e-07,
+      "loss": 2.2173,
+      "step": 20
+    },
+    {
+      "epoch": 0.0034029038112522686,
+      "grad_norm": 70.5,
+      "learning_rate": 3.402646502835539e-07,
+      "loss": 2.1489,
+      "step": 30
+    },
+    {
+      "epoch": 0.004537205081669692,
+      "grad_norm": 71.5,
+      "learning_rate": 4.5368620037807185e-07,
+      "loss": 2.1703,
+      "step": 40
+    },
+    {
+      "epoch": 0.0056715063520871144,
+      "grad_norm": 70.5,
+      "learning_rate": 5.671077504725898e-07,
+      "loss": 2.2116,
+      "step": 50
+    },
+    {
+      "epoch": 0.006805807622504537,
+      "grad_norm": 73.5,
+      "learning_rate": 6.805293005671078e-07,
+      "loss": 2.1757,
+      "step": 60
+    },
+    {
+      "epoch": 0.00794010889292196,
+      "grad_norm": 71.0,
+      "learning_rate": 7.939508506616257e-07,
+      "loss": 2.1018,
+      "step": 70
+    },
+    {
+      "epoch": 0.009074410163339383,
+      "grad_norm": 70.0,
+      "learning_rate": 9.073724007561437e-07,
+      "loss": 2.1104,
+      "step": 80
+    },
+    {
+      "epoch": 0.010208711433756805,
+      "grad_norm": 70.5,
+      "learning_rate": 1.0207939508506615e-06,
+      "loss": 2.0932,
+      "step": 90
+    },
+    {
+      "epoch": 0.011343012704174229,
+      "grad_norm": 74.5,
+      "learning_rate": 1.1342155009451797e-06,
+      "loss": 2.0352,
+      "step": 100
+    },
+    {
+      "epoch": 0.01247731397459165,
+      "grad_norm": 68.5,
+      "learning_rate": 1.2476370510396976e-06,
+      "loss": 1.946,
+      "step": 110
+    },
+    {
+      "epoch": 0.013611615245009074,
+      "grad_norm": 74.0,
+      "learning_rate": 1.3610586011342156e-06,
+      "loss": 1.843,
+      "step": 120
+    },
+    {
+      "epoch": 0.014745916515426498,
+      "grad_norm": 70.0,
+      "learning_rate": 1.4744801512287333e-06,
+      "loss": 1.8038,
+      "step": 130
+    },
+    {
+      "epoch": 0.01588021778584392,
+      "grad_norm": 72.5,
+      "learning_rate": 1.5879017013232515e-06,
+      "loss": 1.7602,
+      "step": 140
+    },
+    {
+      "epoch": 0.01701451905626134,
+      "grad_norm": 98.0,
+      "learning_rate": 1.7013232514177694e-06,
+      "loss": 1.633,
+      "step": 150
+    },
+    {
+      "epoch": 0.018148820326678767,
+      "grad_norm": 189.0,
+      "learning_rate": 1.8147448015122874e-06,
+      "loss": 1.527,
+      "step": 160
+    },
+    {
+      "epoch": 0.01928312159709619,
+      "grad_norm": 290.0,
+      "learning_rate": 1.9281663516068055e-06,
+      "loss": 1.3534,
+      "step": 170
+    },
+    {
+      "epoch": 0.02041742286751361,
+      "grad_norm": 201.0,
+      "learning_rate": 2.041587901701323e-06,
+      "loss": 0.9993,
+      "step": 180
+    },
+    {
+      "epoch": 0.021551724137931036,
+      "grad_norm": 126.0,
+      "learning_rate": 2.1550094517958414e-06,
+      "loss": 0.7995,
+      "step": 190
+    },
+    {
+      "epoch": 0.022686025408348458,
+      "grad_norm": 72.5,
+      "learning_rate": 2.2684310018903594e-06,
+      "loss": 0.781,
+      "step": 200
+    },
+    {
+      "epoch": 0.02382032667876588,
+      "grad_norm": 56.5,
+      "learning_rate": 2.381852551984877e-06,
+      "loss": 0.683,
+      "step": 210
+    },
+    {
+      "epoch": 0.0249546279491833,
+      "grad_norm": 47.25,
+      "learning_rate": 2.4952741020793953e-06,
+      "loss": 0.6598,
+      "step": 220
+    },
+    {
+      "epoch": 0.026088929219600727,
+      "grad_norm": 47.0,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 0.6226,
+      "step": 230
+    },
+    {
+      "epoch": 0.02722323049001815,
+      "grad_norm": 40.75,
+      "learning_rate": 2.722117202268431e-06,
+      "loss": 0.5713,
+      "step": 240
+    },
+    {
+      "epoch": 0.02835753176043557,
+      "grad_norm": 44.0,
+      "learning_rate": 2.835538752362949e-06,
+      "loss": 0.5529,
+      "step": 250
+    },
+    {
+      "epoch": 0.029491833030852996,
+      "grad_norm": 54.25,
+      "learning_rate": 2.9489603024574667e-06,
+      "loss": 0.4694,
+      "step": 260
+    },
+    {
+      "epoch": 0.030626134301270418,
+      "grad_norm": 34.25,
+      "learning_rate": 3.062381852551985e-06,
+      "loss": 0.4789,
+      "step": 270
+    },
+    {
+      "epoch": 0.03176043557168784,
+      "grad_norm": 35.5,
+      "learning_rate": 3.175803402646503e-06,
+      "loss": 0.4618,
+      "step": 280
+    },
+    {
+      "epoch": 0.03289473684210526,
+      "grad_norm": 37.75,
+      "learning_rate": 3.289224952741021e-06,
+      "loss": 0.4601,
+      "step": 290
+    },
+    {
+      "epoch": 0.03402903811252268,
+      "grad_norm": 33.25,
+      "learning_rate": 3.402646502835539e-06,
+      "loss": 0.4079,
+      "step": 300
+    },
+    {
+      "epoch": 0.03516333938294011,
+      "grad_norm": 53.25,
+      "learning_rate": 3.516068052930057e-06,
+      "loss": 0.4136,
+      "step": 310
+    },
+    {
+      "epoch": 0.036297640653357534,
+      "grad_norm": 34.0,
+      "learning_rate": 3.6294896030245748e-06,
+      "loss": 0.4257,
+      "step": 320
+    },
+    {
+      "epoch": 0.037431941923774956,
+      "grad_norm": 70.0,
+      "learning_rate": 3.7429111531190927e-06,
+      "loss": 0.3797,
+      "step": 330
+    },
+    {
+      "epoch": 0.03856624319419238,
+      "grad_norm": 34.0,
+      "learning_rate": 3.856332703213611e-06,
+      "loss": 0.3806,
+      "step": 340
+    },
+    {
+      "epoch": 0.0397005444646098,
+      "grad_norm": 31.5,
+      "learning_rate": 3.969754253308128e-06,
+      "loss": 0.3845,
+      "step": 350
+    },
+    {
+      "epoch": 0.04083484573502722,
+      "grad_norm": 48.5,
+      "learning_rate": 4.083175803402646e-06,
+      "loss": 0.3884,
+      "step": 360
+    },
+    {
+      "epoch": 0.04196914700544464,
+      "grad_norm": 39.5,
+      "learning_rate": 4.196597353497165e-06,
+      "loss": 0.3725,
+      "step": 370
+    },
+    {
+      "epoch": 0.04310344827586207,
+      "grad_norm": 38.5,
+      "learning_rate": 4.310018903591683e-06,
+      "loss": 0.3517,
+      "step": 380
+    },
+    {
+      "epoch": 0.044237749546279494,
+      "grad_norm": 39.25,
+      "learning_rate": 4.4234404536862e-06,
+      "loss": 0.3605,
+      "step": 390
+    },
+    {
+      "epoch": 0.045372050816696916,
+      "grad_norm": 26.625,
+      "learning_rate": 4.536862003780719e-06,
+      "loss": 0.3566,
+      "step": 400
+    },
+    {
+      "epoch": 0.04650635208711434,
+      "grad_norm": 24.75,
+      "learning_rate": 4.650283553875237e-06,
+      "loss": 0.3615,
+      "step": 410
+    },
+    {
+      "epoch": 0.04764065335753176,
+      "grad_norm": 19.875,
+      "learning_rate": 4.763705103969754e-06,
+      "loss": 0.3364,
+      "step": 420
+    },
+    {
+      "epoch": 0.04877495462794918,
+      "grad_norm": 40.75,
+      "learning_rate": 4.877126654064273e-06,
+      "loss": 0.3319,
+      "step": 430
+    },
+    {
+      "epoch": 0.0499092558983666,
+      "grad_norm": 40.75,
+      "learning_rate": 4.990548204158791e-06,
+      "loss": 0.329,
+      "step": 440
+    },
+    {
+      "epoch": 0.05104355716878403,
+      "grad_norm": 26.875,
+      "learning_rate": 5.103969754253308e-06,
+      "loss": 0.322,
+      "step": 450
+    },
+    {
+      "epoch": 0.052177858439201454,
+      "grad_norm": 32.25,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 0.3282,
+      "step": 460
+    },
+    {
+      "epoch": 0.053312159709618875,
+      "grad_norm": 23.875,
+      "learning_rate": 5.3308128544423444e-06,
+      "loss": 0.3349,
+      "step": 470
+    },
+    {
+      "epoch": 0.0544464609800363,
+      "grad_norm": 20.75,
+      "learning_rate": 5.444234404536862e-06,
+      "loss": 0.3183,
+      "step": 480
+    },
+    {
+      "epoch": 0.05558076225045372,
+      "grad_norm": 33.75,
+      "learning_rate": 5.5576559546313795e-06,
+      "loss": 0.3125,
+      "step": 490
+    },
+    {
+      "epoch": 0.05671506352087114,
+      "grad_norm": 23.875,
+      "learning_rate": 5.671077504725898e-06,
+      "loss": 0.3047,
+      "step": 500
+    },
+    {
+      "epoch": 0.05784936479128856,
+      "grad_norm": 25.875,
+      "learning_rate": 5.784499054820416e-06,
+      "loss": 0.3113,
+      "step": 510
+    },
+    {
+      "epoch": 0.05898366606170599,
+      "grad_norm": 26.75,
+      "learning_rate": 5.897920604914933e-06,
+      "loss": 0.2892,
+      "step": 520
+    },
+    {
+      "epoch": 0.06011796733212341,
+      "grad_norm": 29.5,
+      "learning_rate": 6.011342155009452e-06,
+      "loss": 0.3104,
+      "step": 530
+    },
+    {
+      "epoch": 0.061252268602540835,
+      "grad_norm": 20.5,
+      "learning_rate": 6.12476370510397e-06,
+      "loss": 0.3059,
+      "step": 540
+    },
+    {
+      "epoch": 0.06238656987295826,
+      "grad_norm": 29.375,
+      "learning_rate": 6.238185255198487e-06,
+      "loss": 0.3083,
+      "step": 550
+    },
+    {
+      "epoch": 0.06352087114337568,
+      "grad_norm": 30.625,
+      "learning_rate": 6.351606805293006e-06,
+      "loss": 0.294,
+      "step": 560
+    },
+    {
+      "epoch": 0.06465517241379311,
+      "grad_norm": 22.0,
+      "learning_rate": 6.465028355387524e-06,
+      "loss": 0.3006,
+      "step": 570
+    },
+    {
+      "epoch": 0.06578947368421052,
+      "grad_norm": 15.4375,
+      "learning_rate": 6.578449905482042e-06,
+      "loss": 0.2832,
+      "step": 580
+    },
+    {
+      "epoch": 0.06692377495462795,
+      "grad_norm": 18.25,
+      "learning_rate": 6.69187145557656e-06,
+      "loss": 0.2919,
+      "step": 590
+    },
+    {
+      "epoch": 0.06805807622504537,
+      "grad_norm": 17.75,
+      "learning_rate": 6.805293005671078e-06,
+      "loss": 0.2856,
+      "step": 600
+    },
+    {
+      "epoch": 0.0691923774954628,
+      "grad_norm": 26.625,
+      "learning_rate": 6.918714555765596e-06,
+      "loss": 0.2779,
+      "step": 610
+    },
+    {
+      "epoch": 0.07032667876588022,
+      "grad_norm": 20.5,
+      "learning_rate": 7.032136105860114e-06,
+      "loss": 0.2941,
+      "step": 620
+    },
+    {
+      "epoch": 0.07146098003629764,
+      "grad_norm": 23.125,
+      "learning_rate": 7.145557655954632e-06,
+      "loss": 0.2805,
+      "step": 630
+    },
+    {
+      "epoch": 0.07259528130671507,
+      "grad_norm": 23.625,
+      "learning_rate": 7.2589792060491495e-06,
+      "loss": 0.2676,
+      "step": 640
+    },
+    {
+      "epoch": 0.07372958257713248,
+      "grad_norm": 10.6875,
+      "learning_rate": 7.3724007561436675e-06,
+      "loss": 0.2927,
+      "step": 650
+    },
+    {
+      "epoch": 0.07486388384754991,
+      "grad_norm": 20.75,
+      "learning_rate": 7.4858223062381854e-06,
+      "loss": 0.2588,
+      "step": 660
+    },
+    {
+      "epoch": 0.07599818511796733,
+      "grad_norm": 20.375,
+      "learning_rate": 7.5992438563327025e-06,
+      "loss": 0.2629,
+      "step": 670
+    },
+    {
+      "epoch": 0.07713248638838476,
+      "grad_norm": 22.0,
+      "learning_rate": 7.712665406427222e-06,
+      "loss": 0.2685,
+      "step": 680
+    },
+    {
+      "epoch": 0.07826678765880218,
+      "grad_norm": 20.375,
+      "learning_rate": 7.826086956521738e-06,
+      "loss": 0.2468,
+      "step": 690
+    },
+    {
+      "epoch": 0.0794010889292196,
+      "grad_norm": 21.875,
+      "learning_rate": 7.939508506616256e-06,
+      "loss": 0.2488,
+      "step": 700
+    },
+    {
+      "epoch": 0.08053539019963703,
+      "grad_norm": 26.5,
+      "learning_rate": 8.052930056710776e-06,
+      "loss": 0.2616,
+      "step": 710
+    },
+    {
+      "epoch": 0.08166969147005444,
+      "grad_norm": 14.0625,
+      "learning_rate": 8.166351606805292e-06,
+      "loss": 0.2529,
+      "step": 720
+    },
+    {
+      "epoch": 0.08280399274047187,
+      "grad_norm": 20.0,
+      "learning_rate": 8.279773156899812e-06,
+      "loss": 0.2425,
+      "step": 730
+    },
+    {
+      "epoch": 0.08393829401088929,
+      "grad_norm": 28.375,
+      "learning_rate": 8.39319470699433e-06,
+      "loss": 0.2323,
+      "step": 740
+    },
+    {
+      "epoch": 0.08507259528130671,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.506616257088846e-06,
+      "loss": 0.2394,
+      "step": 750
+    },
+    {
+      "epoch": 0.08620689655172414,
+      "grad_norm": 25.875,
+      "learning_rate": 8.620037807183366e-06,
+      "loss": 0.2264,
+      "step": 760
+    },
+    {
+      "epoch": 0.08734119782214156,
+      "grad_norm": 35.75,
+      "learning_rate": 8.733459357277884e-06,
+      "loss": 0.2234,
+      "step": 770
+    },
+    {
+      "epoch": 0.08847549909255899,
+      "grad_norm": 24.625,
+      "learning_rate": 8.8468809073724e-06,
+      "loss": 0.231,
+      "step": 780
+    },
+    {
+      "epoch": 0.0896098003629764,
+      "grad_norm": 18.75,
+      "learning_rate": 8.96030245746692e-06,
+      "loss": 0.2253,
+      "step": 790
+    },
+    {
+      "epoch": 0.09074410163339383,
+      "grad_norm": 24.125,
+      "learning_rate": 9.073724007561438e-06,
+      "loss": 0.2098,
+      "step": 800
+    },
+    {
+      "epoch": 0.09187840290381125,
+      "grad_norm": 26.875,
+      "learning_rate": 9.187145557655954e-06,
+      "loss": 0.2211,
+      "step": 810
+    },
+    {
+      "epoch": 0.09301270417422867,
+      "grad_norm": 20.0,
+      "learning_rate": 9.300567107750473e-06,
+      "loss": 0.2178,
+      "step": 820
+    },
+    {
+      "epoch": 0.0941470054446461,
+      "grad_norm": 16.5,
+      "learning_rate": 9.413988657844991e-06,
+      "loss": 0.2264,
+      "step": 830
+    },
+    {
+      "epoch": 0.09528130671506352,
+      "grad_norm": 11.5625,
+      "learning_rate": 9.527410207939508e-06,
+      "loss": 0.2138,
+      "step": 840
+    },
+    {
+      "epoch": 0.09641560798548095,
+      "grad_norm": 21.75,
+      "learning_rate": 9.640831758034027e-06,
+      "loss": 0.2016,
+      "step": 850
+    },
+    {
+      "epoch": 0.09754990925589836,
+      "grad_norm": 17.0,
+      "learning_rate": 9.754253308128545e-06,
+      "loss": 0.2108,
+      "step": 860
+    },
+    {
+      "epoch": 0.09868421052631579,
+      "grad_norm": 11.4375,
+      "learning_rate": 9.867674858223062e-06,
+      "loss": 0.2005,
+      "step": 870
+    },
+    {
+      "epoch": 0.0998185117967332,
+      "grad_norm": 20.25,
+      "learning_rate": 9.981096408317581e-06,
+      "loss": 0.2117,
+      "step": 880
+    },
+    {
+      "epoch": 0.10095281306715063,
+      "grad_norm": 18.625,
+      "learning_rate": 1.0094517958412099e-05,
+      "loss": 0.2037,
+      "step": 890
+    },
+    {
+      "epoch": 0.10208711433756806,
+      "grad_norm": 23.375,
+      "learning_rate": 1.0207939508506615e-05,
+      "loss": 0.1903,
+      "step": 900
+    },
+    {
+      "epoch": 0.10322141560798548,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.0321361058601135e-05,
+      "loss": 0.2131,
+      "step": 910
+    },
+    {
+      "epoch": 0.10435571687840291,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.0434782608695653e-05,
+      "loss": 0.2058,
+      "step": 920
+    },
+    {
+      "epoch": 0.10549001814882032,
+      "grad_norm": 17.25,
+      "learning_rate": 1.0548204158790171e-05,
+      "loss": 0.1915,
+      "step": 930
+    },
+    {
+      "epoch": 0.10662431941923775,
+      "grad_norm": 16.625,
+      "learning_rate": 1.0661625708884689e-05,
+      "loss": 0.1945,
+      "step": 940
+    },
+    {
+      "epoch": 0.10775862068965517,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.0775047258979205e-05,
+      "loss": 0.1945,
+      "step": 950
+    },
+    {
+      "epoch": 0.1088929219600726,
+      "grad_norm": 21.25,
+      "learning_rate": 1.0888468809073725e-05,
+      "loss": 0.1886,
+      "step": 960
+    },
+    {
+      "epoch": 0.11002722323049002,
+      "grad_norm": 12.875,
+      "learning_rate": 1.1001890359168243e-05,
+      "loss": 0.1917,
+      "step": 970
+    },
+    {
+      "epoch": 0.11116152450090744,
+      "grad_norm": 16.75,
+      "learning_rate": 1.1115311909262759e-05,
+      "loss": 0.1886,
+      "step": 980
+    },
+    {
+      "epoch": 0.11229582577132487,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.1228733459357279e-05,
+      "loss": 0.2012,
+      "step": 990
+    },
+    {
+      "epoch": 0.11343012704174228,
+      "grad_norm": 11.25,
+      "learning_rate": 1.1342155009451797e-05,
+      "loss": 0.1909,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11456442831215971,
+      "grad_norm": 29.875,
+      "learning_rate": 1.1455576559546313e-05,
+      "loss": 0.192,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11569872958257713,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.1568998109640832e-05,
+      "loss": 0.1837,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11683303085299455,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.168241965973535e-05,
+      "loss": 0.1769,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11796733212341198,
+      "grad_norm": 20.25,
+      "learning_rate": 1.1795841209829867e-05,
+      "loss": 0.188,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1191016333938294,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.1909262759924386e-05,
+      "loss": 0.1758,
+      "step": 1050
+    },
+    {
+      "epoch": 0.12023593466424683,
+      "grad_norm": 26.375,
+      "learning_rate": 1.2022684310018904e-05,
+      "loss": 0.196,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12137023593466424,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.213610586011342e-05,
+      "loss": 0.184,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12250453720508167,
+      "grad_norm": 11.375,
+      "learning_rate": 1.224952741020794e-05,
+      "loss": 0.1791,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1236388384754991,
+      "grad_norm": 18.0,
+      "learning_rate": 1.2362948960302458e-05,
+      "loss": 0.1682,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12477313974591651,
+      "grad_norm": 9.375,
+      "learning_rate": 1.2476370510396974e-05,
+      "loss": 0.1801,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12590744101633394,
+      "grad_norm": 14.75,
+      "learning_rate": 1.2589792060491494e-05,
+      "loss": 0.1845,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12704174228675136,
+      "grad_norm": 10.75,
+      "learning_rate": 1.2703213610586012e-05,
+      "loss": 0.1737,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12817604355716877,
+      "grad_norm": 13.875,
+      "learning_rate": 1.281663516068053e-05,
+      "loss": 0.181,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12931034482758622,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.2930056710775048e-05,
+      "loss": 0.1781,
+      "step": 1140
+    },
+    {
+      "epoch": 0.13044464609800363,
+      "grad_norm": 12.375,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.1908,
+      "step": 1150
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.3156899810964084e-05,
+      "loss": 0.1736,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1327132486388385,
+      "grad_norm": 14.0,
+      "learning_rate": 1.3270321361058602e-05,
+      "loss": 0.1767,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1338475499092559,
+      "grad_norm": 10.25,
+      "learning_rate": 1.338374291115312e-05,
+      "loss": 0.1693,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13498185117967332,
+      "grad_norm": 10.125,
+      "learning_rate": 1.3497164461247638e-05,
+      "loss": 0.166,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13611615245009073,
+      "grad_norm": 10.625,
+      "learning_rate": 1.3610586011342156e-05,
+      "loss": 0.1616,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13725045372050818,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.3724007561436673e-05,
+      "loss": 0.1792,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1383847549909256,
+      "grad_norm": 13.375,
+      "learning_rate": 1.3837429111531191e-05,
+      "loss": 0.1737,
+      "step": 1220
+    },
+    {
+      "epoch": 0.139519056261343,
+      "grad_norm": 17.125,
+      "learning_rate": 1.395085066162571e-05,
+      "loss": 0.174,
+      "step": 1230
+    },
+    {
+      "epoch": 0.14065335753176045,
+      "grad_norm": 22.75,
+      "learning_rate": 1.4064272211720227e-05,
+      "loss": 0.177,
+      "step": 1240
+    },
+    {
+      "epoch": 0.14178765880217786,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.4177693761814745e-05,
+      "loss": 0.1709,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14292196007259528,
+      "grad_norm": 21.625,
+      "learning_rate": 1.4291115311909263e-05,
+      "loss": 0.172,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1440562613430127,
+      "grad_norm": 13.25,
+      "learning_rate": 1.4404536862003781e-05,
+      "loss": 0.1795,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14519056261343014,
+      "grad_norm": 13.25,
+      "learning_rate": 1.4517958412098299e-05,
+      "loss": 0.1685,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14632486388384755,
+      "grad_norm": 24.625,
+      "learning_rate": 1.4631379962192817e-05,
+      "loss": 0.1795,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14745916515426497,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.4744801512287335e-05,
+      "loss": 0.1653,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1485934664246824,
+      "grad_norm": 12.125,
+      "learning_rate": 1.4858223062381853e-05,
+      "loss": 0.1734,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14972776769509982,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4971644612476371e-05,
+      "loss": 0.1628,
+      "step": 1320
+    },
+    {
+      "epoch": 0.15086206896551724,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.5085066162570889e-05,
+      "loss": 0.1617,
+      "step": 1330
+    },
+    {
+      "epoch": 0.15199637023593465,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.5198487712665405e-05,
+      "loss": 0.171,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1531306715063521,
+      "grad_norm": 21.0,
+      "learning_rate": 1.5311909262759923e-05,
+      "loss": 0.165,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1542649727767695,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.5425330812854444e-05,
+      "loss": 0.1735,
+      "step": 1360
+    },
+    {
+      "epoch": 0.15539927404718692,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.553875236294896e-05,
+      "loss": 0.164,
+      "step": 1370
+    },
+    {
+      "epoch": 0.15653357531760437,
+      "grad_norm": 11.625,
+      "learning_rate": 1.5652173913043477e-05,
+      "loss": 0.1642,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15766787658802178,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.5765595463137998e-05,
+      "loss": 0.1714,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1588021778584392,
+      "grad_norm": 13.5,
+      "learning_rate": 1.5879017013232513e-05,
+      "loss": 0.168,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1599364791288566,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.599243856332703e-05,
+      "loss": 0.1739,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16107078039927405,
+      "grad_norm": 8.75,
+      "learning_rate": 1.6105860113421552e-05,
+      "loss": 0.1753,
+      "step": 1420
+    },
+    {
+      "epoch": 0.16220508166969147,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.621928166351607e-05,
+      "loss": 0.1808,
+      "step": 1430
+    },
+    {
+      "epoch": 0.16333938294010888,
+      "grad_norm": 5.71875,
+      "learning_rate": 1.6332703213610585e-05,
+      "loss": 0.1583,
+      "step": 1440
+    },
+    {
+      "epoch": 0.16447368421052633,
+      "grad_norm": 12.0,
+      "learning_rate": 1.6446124763705106e-05,
+      "loss": 0.1649,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16560798548094374,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.6559546313799624e-05,
+      "loss": 0.1713,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16674228675136116,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.667296786389414e-05,
+      "loss": 0.1685,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16787658802177857,
+      "grad_norm": 9.75,
+      "learning_rate": 1.678638941398866e-05,
+      "loss": 0.1628,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16901088929219601,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6899810964083178e-05,
+      "loss": 0.1636,
+      "step": 1490
+    },
+    {
+      "epoch": 0.17014519056261343,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.7013232514177692e-05,
+      "loss": 0.1639,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17127949183303084,
+      "grad_norm": 9.5,
+      "learning_rate": 1.7126654064272214e-05,
+      "loss": 0.1772,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 17.5,
+      "learning_rate": 1.724007561436673e-05,
+      "loss": 0.1622,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1735480943738657,
+      "grad_norm": 16.625,
+      "learning_rate": 1.7353497164461246e-05,
+      "loss": 0.163,
+      "step": 1530
+    },
+    {
+      "epoch": 0.17468239564428312,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7466918714555767e-05,
+      "loss": 0.1564,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17581669691470053,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.7580340264650285e-05,
+      "loss": 0.1662,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17695099818511797,
+      "grad_norm": 9.125,
+      "learning_rate": 1.76937618147448e-05,
+      "loss": 0.1593,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1780852994555354,
+      "grad_norm": 6.5,
+      "learning_rate": 1.780718336483932e-05,
+      "loss": 0.1633,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1792196007259528,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.792060491493384e-05,
+      "loss": 0.1578,
+      "step": 1580
+    },
+    {
+      "epoch": 0.18035390199637025,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.8034026465028354e-05,
+      "loss": 0.1624,
+      "step": 1590
+    },
+    {
+      "epoch": 0.18148820326678766,
+      "grad_norm": 11.375,
+      "learning_rate": 1.8147448015122875e-05,
+      "loss": 0.1614,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18262250453720508,
+      "grad_norm": 11.75,
+      "learning_rate": 1.8260869565217393e-05,
+      "loss": 0.1594,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1837568058076225,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.8374291115311908e-05,
+      "loss": 0.1574,
+      "step": 1620
+    },
+    {
+      "epoch": 0.18489110707803993,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.848771266540643e-05,
+      "loss": 0.1721,
+      "step": 1630
+    },
+    {
+      "epoch": 0.18602540834845735,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8601134215500947e-05,
+      "loss": 0.1525,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18715970961887476,
+      "grad_norm": 9.75,
+      "learning_rate": 1.871455576559546e-05,
+      "loss": 0.1621,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1882940108892922,
+      "grad_norm": 8.375,
+      "learning_rate": 1.8827977315689983e-05,
+      "loss": 0.1791,
+      "step": 1660
+    },
+    {
+      "epoch": 0.18942831215970962,
+      "grad_norm": 7.25,
+      "learning_rate": 1.89413988657845e-05,
+      "loss": 0.1643,
+      "step": 1670
+    },
+    {
+      "epoch": 0.19056261343012704,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.9054820415879015e-05,
+      "loss": 0.1617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.19169691470054445,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.9168241965973537e-05,
+      "loss": 0.1598,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1928312159709619,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.9281663516068055e-05,
+      "loss": 0.1591,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1939655172413793,
+      "grad_norm": 12.25,
+      "learning_rate": 1.939508506616257e-05,
+      "loss": 0.1566,
+      "step": 1710
+    },
+    {
+      "epoch": 0.19509981851179672,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.950850661625709e-05,
+      "loss": 0.166,
+      "step": 1720
+    },
+    {
+      "epoch": 0.19623411978221417,
+      "grad_norm": 8.75,
+      "learning_rate": 1.962192816635161e-05,
+      "loss": 0.1631,
+      "step": 1730
+    },
+    {
+      "epoch": 0.19736842105263158,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.9735349716446123e-05,
+      "loss": 0.1651,
+      "step": 1740
+    },
+    {
+      "epoch": 0.198502722323049,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9848771266540644e-05,
+      "loss": 0.1647,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1996370235934664,
+      "grad_norm": 14.375,
+      "learning_rate": 1.9962192816635162e-05,
+      "loss": 0.1584,
+      "step": 1760
+    },
+    {
+      "epoch": 0.20077132486388385,
+      "grad_norm": 24.125,
+      "learning_rate": 2.0075614366729677e-05,
+      "loss": 0.1764,
+      "step": 1770
+    },
+    {
+      "epoch": 0.20190562613430127,
+      "grad_norm": 15.25,
+      "learning_rate": 2.0189035916824198e-05,
+      "loss": 0.1751,
+      "step": 1780
+    },
+    {
+      "epoch": 0.20303992740471868,
+      "grad_norm": 19.125,
+      "learning_rate": 2.0302457466918716e-05,
+      "loss": 0.179,
+      "step": 1790
+    },
+    {
+      "epoch": 0.20417422867513613,
+      "grad_norm": 5.5625,
+      "learning_rate": 2.041587901701323e-05,
+      "loss": 0.1565,
+      "step": 1800
+    },
+    {
+      "epoch": 0.20530852994555354,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0529300567107752e-05,
+      "loss": 0.1659,
+      "step": 1810
+    },
+    {
+      "epoch": 0.20644283121597096,
+      "grad_norm": 14.5,
+      "learning_rate": 2.064272211720227e-05,
+      "loss": 0.1686,
+      "step": 1820
+    },
+    {
+      "epoch": 0.20757713248638837,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0756143667296785e-05,
+      "loss": 0.164,
+      "step": 1830
+    },
+    {
+      "epoch": 0.20871143375680581,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0869565217391306e-05,
+      "loss": 0.1626,
+      "step": 1840
+    },
+    {
+      "epoch": 0.20984573502722323,
+      "grad_norm": 11.3125,
+      "learning_rate": 2.0982986767485824e-05,
+      "loss": 0.1633,
+      "step": 1850
+    },
+    {
+      "epoch": 0.21098003629764064,
+      "grad_norm": 7.25,
+      "learning_rate": 2.1096408317580342e-05,
+      "loss": 0.1711,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2121143375680581,
+      "grad_norm": 4.9375,
+      "learning_rate": 2.120982986767486e-05,
+      "loss": 0.1637,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2132486388384755,
+      "grad_norm": 10.8125,
+      "learning_rate": 2.1323251417769378e-05,
+      "loss": 0.1726,
+      "step": 1880
+    },
+    {
+      "epoch": 0.21438294010889292,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1436672967863896e-05,
+      "loss": 0.1721,
+      "step": 1890
+    },
+    {
+      "epoch": 0.21551724137931033,
+      "grad_norm": 8.875,
+      "learning_rate": 2.155009451795841e-05,
+      "loss": 0.1679,
+      "step": 1900
+    },
+    {
+      "epoch": 0.21665154264972777,
+      "grad_norm": 5.03125,
+      "learning_rate": 2.166351606805293e-05,
+      "loss": 0.1545,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2177858439201452,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.177693761814745e-05,
+      "loss": 0.1563,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2189201451905626,
+      "grad_norm": 9.5,
+      "learning_rate": 2.1890359168241964e-05,
+      "loss": 0.1691,
+      "step": 1930
+    },
+    {
+      "epoch": 0.22005444646098005,
+      "grad_norm": 6.375,
+      "learning_rate": 2.2003780718336485e-05,
+      "loss": 0.1596,
+      "step": 1940
+    },
+    {
+      "epoch": 0.22118874773139746,
+      "grad_norm": 9.875,
+      "learning_rate": 2.2117202268431003e-05,
+      "loss": 0.1753,
+      "step": 1950
+    },
+    {
+      "epoch": 0.22232304900181488,
+      "grad_norm": 9.6875,
+      "learning_rate": 2.2230623818525518e-05,
+      "loss": 0.1733,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2234573502722323,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.234404536862004e-05,
+      "loss": 0.1716,
+      "step": 1970
+    },
+    {
+      "epoch": 0.22459165154264973,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2457466918714557e-05,
+      "loss": 0.1626,
+      "step": 1980
+    },
+    {
+      "epoch": 0.22572595281306715,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.2570888468809072e-05,
+      "loss": 0.1563,
+      "step": 1990
+    },
+    {
+      "epoch": 0.22686025408348456,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.2684310018903593e-05,
+      "loss": 0.1603,
+      "step": 2000
+    },
+    {
+      "epoch": 0.227994555353902,
+      "grad_norm": 9.625,
+      "learning_rate": 2.279773156899811e-05,
+      "loss": 0.1667,
+      "step": 2010
+    },
+    {
+      "epoch": 0.22912885662431942,
+      "grad_norm": 17.625,
+      "learning_rate": 2.2911153119092626e-05,
+      "loss": 0.1583,
+      "step": 2020
+    },
+    {
+      "epoch": 0.23026315789473684,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.3024574669187147e-05,
+      "loss": 0.1853,
+      "step": 2030
+    },
+    {
+      "epoch": 0.23139745916515425,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.3137996219281665e-05,
+      "loss": 0.1661,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2325317604355717,
+      "grad_norm": 6.125,
+      "learning_rate": 2.325141776937618e-05,
+      "loss": 0.1555,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2336660617059891,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.33648393194707e-05,
+      "loss": 0.1669,
+      "step": 2060
+    },
+    {
+      "epoch": 0.23480036297640652,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.347826086956522e-05,
+      "loss": 0.153,
+      "step": 2070
+    },
+    {
+      "epoch": 0.23593466424682397,
+      "grad_norm": 5.4375,
+      "learning_rate": 2.3591682419659733e-05,
+      "loss": 0.1509,
+      "step": 2080
+    },
+    {
+      "epoch": 0.23706896551724138,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.3705103969754255e-05,
+      "loss": 0.1562,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2382032667876588,
+      "grad_norm": 5.625,
+      "learning_rate": 2.3818525519848773e-05,
+      "loss": 0.1583,
+      "step": 2100
+    },
+    {
+      "epoch": 0.23933756805807624,
+      "grad_norm": 4.625,
+      "learning_rate": 2.3931947069943287e-05,
+      "loss": 0.1569,
+      "step": 2110
+    },
+    {
+      "epoch": 0.24047186932849365,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.404536862003781e-05,
+      "loss": 0.1492,
+      "step": 2120
+    },
+    {
+      "epoch": 0.24160617059891107,
+      "grad_norm": 5.0,
+      "learning_rate": 2.4158790170132326e-05,
+      "loss": 0.15,
+      "step": 2130
+    },
+    {
+      "epoch": 0.24274047186932848,
+      "grad_norm": 4.75,
+      "learning_rate": 2.427221172022684e-05,
+      "loss": 0.1653,
+      "step": 2140
+    },
+    {
+      "epoch": 0.24387477313974593,
+      "grad_norm": 4.90625,
+      "learning_rate": 2.4385633270321362e-05,
+      "loss": 0.1582,
+      "step": 2150
+    },
+    {
+      "epoch": 0.24500907441016334,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.449905482041588e-05,
+      "loss": 0.1594,
+      "step": 2160
+    },
+    {
+      "epoch": 0.24614337568058076,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.4612476370510395e-05,
+      "loss": 0.1605,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2472776769509982,
+      "grad_norm": 5.1875,
+      "learning_rate": 2.4725897920604916e-05,
+      "loss": 0.1667,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2484119782214156,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.4839319470699434e-05,
+      "loss": 0.165,
+      "step": 2190
+    },
+    {
+      "epoch": 0.24954627949183303,
+      "grad_norm": 6.375,
+      "learning_rate": 2.495274102079395e-05,
+      "loss": 0.1592,
+      "step": 2200
+    },
+    {
+      "epoch": 0.25068058076225047,
+      "grad_norm": 9.5,
+      "learning_rate": 2.506616257088847e-05,
+      "loss": 0.156,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2518148820326679,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.5179584120982988e-05,
+      "loss": 0.1588,
+      "step": 2220
+    },
+    {
+      "epoch": 0.2529491833030853,
+      "grad_norm": 7.125,
+      "learning_rate": 2.5293005671077503e-05,
+      "loss": 0.173,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2540834845735027,
+      "grad_norm": 5.5,
+      "learning_rate": 2.5406427221172024e-05,
+      "loss": 0.1667,
+      "step": 2240
+    },
+    {
+      "epoch": 0.25521778584392013,
+      "grad_norm": 12.125,
+      "learning_rate": 2.5519848771266542e-05,
+      "loss": 0.1624,
+      "step": 2250
+    },
+    {
+      "epoch": 0.25635208711433755,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.563327032136106e-05,
+      "loss": 0.152,
+      "step": 2260
+    },
+    {
+      "epoch": 0.257486388384755,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.5746691871455578e-05,
+      "loss": 0.1643,
+      "step": 2270
+    },
+    {
+      "epoch": 0.25862068965517243,
+      "grad_norm": 7.0,
+      "learning_rate": 2.5860113421550096e-05,
+      "loss": 0.1637,
+      "step": 2280
+    },
+    {
+      "epoch": 0.25975499092558985,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.5973534971644614e-05,
+      "loss": 0.1664,
+      "step": 2290
+    },
+    {
+      "epoch": 0.26088929219600726,
+      "grad_norm": 7.5,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 0.1615,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2620235934664247,
+      "grad_norm": 4.78125,
+      "learning_rate": 2.620037807183365e-05,
+      "loss": 0.1518,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.6313799621928167e-05,
+      "loss": 0.1572,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2642921960072595,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.6427221172022685e-05,
+      "loss": 0.1546,
+      "step": 2330
+    },
+    {
+      "epoch": 0.265426497277677,
+      "grad_norm": 4.96875,
+      "learning_rate": 2.6540642722117203e-05,
+      "loss": 0.1575,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2665607985480944,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.665406427221172e-05,
+      "loss": 0.1622,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2676950998185118,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.676748582230624e-05,
+      "loss": 0.1588,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2688294010889292,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.6880907372400757e-05,
+      "loss": 0.1574,
+      "step": 2370
+    },
+    {
+      "epoch": 0.26996370235934664,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.6994328922495275e-05,
+      "loss": 0.1571,
+      "step": 2380
+    },
+    {
+      "epoch": 0.27109800362976405,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.7107750472589793e-05,
+      "loss": 0.1558,
+      "step": 2390
+    },
+    {
+      "epoch": 0.27223230490018147,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.722117202268431e-05,
+      "loss": 0.1715,
+      "step": 2400
+    },
+    {
+      "epoch": 0.27336660617059894,
+      "grad_norm": 7.25,
+      "learning_rate": 2.733459357277883e-05,
+      "loss": 0.165,
+      "step": 2410
+    },
+    {
+      "epoch": 0.27450090744101635,
+      "grad_norm": 7.75,
+      "learning_rate": 2.7448015122873347e-05,
+      "loss": 0.1572,
+      "step": 2420
+    },
+    {
+      "epoch": 0.27563520871143377,
+      "grad_norm": 5.28125,
+      "learning_rate": 2.7561436672967865e-05,
+      "loss": 0.1647,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2767695099818512,
+      "grad_norm": 7.25,
+      "learning_rate": 2.7674858223062383e-05,
+      "loss": 0.1687,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2779038112522686,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.77882797731569e-05,
+      "loss": 0.1533,
+      "step": 2450
+    },
+    {
+      "epoch": 0.279038112522686,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.790170132325142e-05,
+      "loss": 0.16,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2801724137931034,
+      "grad_norm": 5.1875,
+      "learning_rate": 2.8015122873345937e-05,
+      "loss": 0.1684,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2813067150635209,
+      "grad_norm": 10.0625,
+      "learning_rate": 2.8128544423440455e-05,
+      "loss": 0.1679,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2824410163339383,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.8241965973534973e-05,
+      "loss": 0.1674,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2835753176043557,
+      "grad_norm": 5.03125,
+      "learning_rate": 2.835538752362949e-05,
+      "loss": 0.1676,
+      "step": 2500
+    },
+    {
+      "epoch": 0.28470961887477314,
+      "grad_norm": 4.75,
+      "learning_rate": 2.846880907372401e-05,
+      "loss": 0.1547,
+      "step": 2510
+    },
+    {
+      "epoch": 0.28584392014519056,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.8582230623818526e-05,
+      "loss": 0.1579,
+      "step": 2520
+    },
+    {
+      "epoch": 0.28697822141560797,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8695652173913044e-05,
+      "loss": 0.1559,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2881125226860254,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.8809073724007562e-05,
+      "loss": 0.1673,
+      "step": 2540
+    },
+    {
+      "epoch": 0.28924682395644286,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.892249527410208e-05,
+      "loss": 0.1645,
+      "step": 2550
+    },
+    {
+      "epoch": 0.29038112522686027,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.9035916824196598e-05,
+      "loss": 0.1581,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2915154264972777,
+      "grad_norm": 4.46875,
+      "learning_rate": 2.9149338374291116e-05,
+      "loss": 0.1565,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2926497277676951,
+      "grad_norm": 5.125,
+      "learning_rate": 2.9262759924385634e-05,
+      "loss": 0.1707,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2937840290381125,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.9376181474480152e-05,
+      "loss": 0.1638,
+      "step": 2590
+    },
+    {
+      "epoch": 0.29491833030852993,
+      "grad_norm": 3.25,
+      "learning_rate": 2.948960302457467e-05,
+      "loss": 0.1658,
+      "step": 2600
+    },
+    {
+      "epoch": 0.29605263157894735,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.9603024574669188e-05,
+      "loss": 0.1556,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2971869328493648,
+      "grad_norm": 4.875,
+      "learning_rate": 2.9716446124763706e-05,
+      "loss": 0.1717,
+      "step": 2620
+    },
+    {
+      "epoch": 0.29832123411978223,
+      "grad_norm": 5.09375,
+      "learning_rate": 2.9829867674858224e-05,
+      "loss": 0.1761,
+      "step": 2630
+    },
+    {
+      "epoch": 0.29945553539019965,
+      "grad_norm": 4.25,
+      "learning_rate": 2.9943289224952742e-05,
+      "loss": 0.1669,
+      "step": 2640
+    },
+    {
+      "epoch": 0.30058983666061706,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.9999996733838852e-05,
+      "loss": 0.1562,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3017241379310345,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.99999706045582e-05,
+      "loss": 0.156,
+      "step": 2660
+    },
+    {
+      "epoch": 0.3028584392014519,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.9999918346042412e-05,
+      "loss": 0.1568,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3039927404718693,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.999983995838252e-05,
+      "loss": 0.152,
+      "step": 2680
+    },
+    {
+      "epoch": 0.3051270417422868,
+      "grad_norm": 7.875,
+      "learning_rate": 2.9999735441715067e-05,
+      "loss": 0.159,
+      "step": 2690
+    },
+    {
+      "epoch": 0.3062613430127042,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.999960479622213e-05,
+      "loss": 0.1666,
+      "step": 2700
+    },
+    {
+      "epoch": 0.3073956442831216,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9999448022131267e-05,
+      "loss": 0.166,
+      "step": 2710
+    },
+    {
+      "epoch": 0.308529945553539,
+      "grad_norm": 4.65625,
+      "learning_rate": 2.9999265119715588e-05,
+      "loss": 0.1708,
+      "step": 2720
+    },
+    {
+      "epoch": 0.30966424682395643,
+      "grad_norm": 5.125,
+      "learning_rate": 2.999905608929369e-05,
+      "loss": 0.1618,
+      "step": 2730
+    },
+    {
+      "epoch": 0.31079854809437385,
+      "grad_norm": 3.875,
+      "learning_rate": 2.9998820931229702e-05,
+      "loss": 0.1446,
+      "step": 2740
+    },
+    {
+      "epoch": 0.31193284936479126,
+      "grad_norm": 5.5625,
+      "learning_rate": 2.9998559645933254e-05,
+      "loss": 0.1689,
+      "step": 2750
+    },
+    {
+      "epoch": 0.31306715063520874,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.999827223385949e-05,
+      "loss": 0.1722,
+      "step": 2760
+    },
+    {
+      "epoch": 0.31420145190562615,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9997958695509075e-05,
+      "loss": 0.1568,
+      "step": 2770
+    },
+    {
+      "epoch": 0.31533575317604357,
+      "grad_norm": 5.8125,
+      "learning_rate": 2.9997619031428166e-05,
+      "loss": 0.1615,
+      "step": 2780
+    },
+    {
+      "epoch": 0.316470054446461,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.9997253242208454e-05,
+      "loss": 0.1606,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3176043557168784,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.9996861328487123e-05,
+      "loss": 0.1632,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3187386569872958,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.9996443290946863e-05,
+      "loss": 0.1694,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3198729582577132,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.9995999130315883e-05,
+      "loss": 0.1624,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3210072595281307,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.9995528847367885e-05,
+      "loss": 0.1694,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3221415607985481,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.9995032442922084e-05,
+      "loss": 0.1625,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3232758620689655,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9994509917843184e-05,
+      "loss": 0.1598,
+      "step": 2850
+    },
+    {
+      "epoch": 0.32441016333938294,
+      "grad_norm": 6.125,
+      "learning_rate": 2.9993961273041404e-05,
+      "loss": 0.1481,
+      "step": 2860
+    },
+    {
+      "epoch": 0.32554446460980035,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.9993386509472464e-05,
+      "loss": 0.1578,
+      "step": 2870
+    },
+    {
+      "epoch": 0.32667876588021777,
+      "grad_norm": 3.5,
+      "learning_rate": 2.999278562813756e-05,
+      "loss": 0.158,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3278130671506352,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.999215863008341e-05,
+      "loss": 0.1612,
+      "step": 2890
+    },
+    {
+      "epoch": 0.32894736842105265,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.9991505516402214e-05,
+      "loss": 0.1589,
+      "step": 2900
+    },
+    {
+      "epoch": 0.33008166969147007,
+      "grad_norm": 4.46875,
+      "learning_rate": 2.9990826288231663e-05,
+      "loss": 0.164,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3312159709618875,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.9990120946754937e-05,
+      "loss": 0.1638,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3323502722323049,
+      "grad_norm": 6.125,
+      "learning_rate": 2.998938949320071e-05,
+      "loss": 0.1558,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3334845735027223,
+      "grad_norm": 4.15625,
+      "learning_rate": 2.998863192884314e-05,
+      "loss": 0.1564,
+      "step": 2940
+    },
+    {
+      "epoch": 0.33461887477313973,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.998784825500187e-05,
+      "loss": 0.1573,
+      "step": 2950
+    },
+    {
+      "epoch": 0.33575317604355714,
+      "grad_norm": 5.375,
+      "learning_rate": 2.998703847304202e-05,
+      "loss": 0.1619,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3368874773139746,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.998620258437419e-05,
+      "loss": 0.1536,
+      "step": 2970
+    },
+    {
+      "epoch": 0.33802177858439203,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.998534059045446e-05,
+      "loss": 0.1555,
+      "step": 2980
+    },
+    {
+      "epoch": 0.33915607985480944,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.998445249278438e-05,
+      "loss": 0.1657,
+      "step": 2990
+    },
+    {
+      "epoch": 0.34029038112522686,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.9983538292910982e-05,
+      "loss": 0.156,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3414246823956443,
+      "grad_norm": 5.0625,
+      "learning_rate": 2.9982597992426747e-05,
+      "loss": 0.1722,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3425589836660617,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.998163159296964e-05,
+      "loss": 0.1544,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3436932849364791,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.9980639096223087e-05,
+      "loss": 0.1416,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.9979620503915965e-05,
+      "loss": 0.1604,
+      "step": 3040
+    },
+    {
+      "epoch": 0.345961887477314,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.997857581782261e-05,
+      "loss": 0.1603,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3470961887477314,
+      "grad_norm": 4.625,
+      "learning_rate": 2.997750503976282e-05,
+      "loss": 0.1574,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3482304900181488,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.997640817160184e-05,
+      "loss": 0.1489,
+      "step": 3070
+    },
+    {
+      "epoch": 0.34936479128856623,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.9975285215250364e-05,
+      "loss": 0.1513,
+      "step": 3080
+    },
+    {
+      "epoch": 0.35049909255898365,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.9974136172664526e-05,
+      "loss": 0.168,
+      "step": 3090
+    },
+    {
+      "epoch": 0.35163339382940106,
+      "grad_norm": 5.46875,
+      "learning_rate": 2.9972961045845898e-05,
+      "loss": 0.1561,
+      "step": 3100
+    },
+    {
+      "epoch": 0.35276769509981853,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9971759836841502e-05,
+      "loss": 0.1654,
+      "step": 3110
+    },
+    {
+      "epoch": 0.35390199637023595,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.9970532547743787e-05,
+      "loss": 0.1514,
+      "step": 3120
+    },
+    {
+      "epoch": 0.35503629764065336,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.996927918069063e-05,
+      "loss": 0.1551,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3561705989110708,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.9967999737865338e-05,
+      "loss": 0.1599,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3573049001814882,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9966694221496635e-05,
+      "loss": 0.1585,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3584392014519056,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9965362633858673e-05,
+      "loss": 0.1658,
+      "step": 3160
+    },
+    {
+      "epoch": 0.359573502722323,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.9964004977271013e-05,
+      "loss": 0.1498,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3607078039927405,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.996262125409863e-05,
+      "loss": 0.1654,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3618421052631579,
+      "grad_norm": 3.75,
+      "learning_rate": 2.9961211466751902e-05,
+      "loss": 0.1632,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3629764065335753,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.995977561768661e-05,
+      "loss": 0.142,
+      "step": 3200
+    },
+    {
+      "epoch": 0.36411070780399274,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.995831370940394e-05,
+      "loss": 0.1525,
+      "step": 3210
+    },
+    {
+      "epoch": 0.36524500907441015,
+      "grad_norm": 4.5,
+      "learning_rate": 2.995682574445046e-05,
+      "loss": 0.1604,
+      "step": 3220
+    },
+    {
+      "epoch": 0.36637931034482757,
+      "grad_norm": 4.6875,
+      "learning_rate": 2.9955311725418136e-05,
+      "loss": 0.1599,
+      "step": 3230
+    },
+    {
+      "epoch": 0.367513611615245,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.995377165494432e-05,
+      "loss": 0.1513,
+      "step": 3240
+    },
+    {
+      "epoch": 0.36864791288566245,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.9952205535711742e-05,
+      "loss": 0.1603,
+      "step": 3250
+    },
+    {
+      "epoch": 0.36978221415607987,
+      "grad_norm": 4.75,
+      "learning_rate": 2.9950613370448504e-05,
+      "loss": 0.1651,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3709165154264973,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.9948995161928085e-05,
+      "loss": 0.1591,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3720508166969147,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9947350912969326e-05,
+      "loss": 0.1494,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3731851179673321,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9945680626436432e-05,
+      "loss": 0.1415,
+      "step": 3290
+    },
+    {
+      "epoch": 0.37431941923774953,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9943984305238957e-05,
+      "loss": 0.1593,
+      "step": 3300
+    },
+    {
+      "epoch": 0.37545372050816694,
+      "grad_norm": 5.40625,
+      "learning_rate": 2.9942261952331814e-05,
+      "loss": 0.157,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3765880217785844,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.9940513570715263e-05,
+      "loss": 0.1597,
+      "step": 3320
+    },
+    {
+      "epoch": 0.37772232304900183,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.99387391634349e-05,
+      "loss": 0.1571,
+      "step": 3330
+    },
+    {
+      "epoch": 0.37885662431941924,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.993693873358166e-05,
+      "loss": 0.1503,
+      "step": 3340
+    },
+    {
+      "epoch": 0.37999092558983666,
+      "grad_norm": 4.125,
+      "learning_rate": 2.9935112284291806e-05,
+      "loss": 0.1536,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3811252268602541,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.9933259818746917e-05,
+      "loss": 0.1621,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3822595281306715,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.9931381340173914e-05,
+      "loss": 0.1562,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3833938294010889,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9929476851845005e-05,
+      "loss": 0.1522,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3845281306715064,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.9927546357077727e-05,
+      "loss": 0.1589,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3856624319419238,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.9925589859234904e-05,
+      "loss": 0.153,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3867967332123412,
+      "grad_norm": 4.40625,
+      "learning_rate": 2.9923607361724663e-05,
+      "loss": 0.1477,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3879310344827586,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.9921598868000427e-05,
+      "loss": 0.159,
+      "step": 3420
+    },
+    {
+      "epoch": 0.38906533575317603,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9919564381560884e-05,
+      "loss": 0.1523,
+      "step": 3430
+    },
+    {
+      "epoch": 0.39019963702359345,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.9917503905950022e-05,
+      "loss": 0.1512,
+      "step": 3440
+    },
+    {
+      "epoch": 0.39133393829401086,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.991541744475709e-05,
+      "loss": 0.1487,
+      "step": 3450
+    },
+    {
+      "epoch": 0.39246823956442833,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9913305001616607e-05,
+      "loss": 0.1545,
+      "step": 3460
+    },
+    {
+      "epoch": 0.39360254083484575,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9911166580208346e-05,
+      "loss": 0.1679,
+      "step": 3470
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9909002184257327e-05,
+      "loss": 0.1575,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3958711433756806,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9906811817533835e-05,
+      "loss": 0.1557,
+      "step": 3490
+    },
+    {
+      "epoch": 0.397005444646098,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9904595483853386e-05,
+      "loss": 0.1452,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3981397459165154,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9902353187076715e-05,
+      "loss": 0.1481,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3992740471869328,
+      "grad_norm": 4.375,
+      "learning_rate": 2.9900084931109812e-05,
+      "loss": 0.1578,
+      "step": 3520
+    },
+    {
+      "epoch": 0.4004083484573503,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.989779071990386e-05,
+      "loss": 0.1624,
+      "step": 3530
+    },
+    {
+      "epoch": 0.4015426497277677,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.989547055745527e-05,
+      "loss": 0.1498,
+      "step": 3540
+    },
+    {
+      "epoch": 0.4026769509981851,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.9893124447805654e-05,
+      "loss": 0.1411,
+      "step": 3550
+    },
+    {
+      "epoch": 0.40381125226860254,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9890752395041822e-05,
+      "loss": 0.1577,
+      "step": 3560
+    },
+    {
+      "epoch": 0.40494555353901995,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.9888354403295782e-05,
+      "loss": 0.1748,
+      "step": 3570
+    },
+    {
+      "epoch": 0.40607985480943737,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.9885930476744722e-05,
+      "loss": 0.1546,
+      "step": 3580
+    },
+    {
+      "epoch": 0.4072141560798548,
+      "grad_norm": 4.25,
+      "learning_rate": 2.9883480619610992e-05,
+      "loss": 0.1571,
+      "step": 3590
+    },
+    {
+      "epoch": 0.40834845735027225,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9881004836162145e-05,
+      "loss": 0.1588,
+      "step": 3600
+    },
+    {
+      "epoch": 0.40948275862068967,
+      "grad_norm": 3.0,
+      "learning_rate": 2.987850313071087e-05,
+      "loss": 0.1535,
+      "step": 3610
+    },
+    {
+      "epoch": 0.4106170598911071,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.9875975507615015e-05,
+      "loss": 0.1507,
+      "step": 3620
+    },
+    {
+      "epoch": 0.4117513611615245,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9873421971277588e-05,
+      "loss": 0.1567,
+      "step": 3630
+    },
+    {
+      "epoch": 0.4128856624319419,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9870842526146716e-05,
+      "loss": 0.1465,
+      "step": 3640
+    },
+    {
+      "epoch": 0.41401996370235933,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.9868237176715685e-05,
+      "loss": 0.1575,
+      "step": 3650
+    },
+    {
+      "epoch": 0.41515426497277674,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.9865605927522878e-05,
+      "loss": 0.1605,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4162885662431942,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.9862948783151812e-05,
+      "loss": 0.1392,
+      "step": 3670
+    },
+    {
+      "epoch": 0.41742286751361163,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.98602657482311e-05,
+      "loss": 0.1652,
+      "step": 3680
+    },
+    {
+      "epoch": 0.41855716878402904,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9857556827434468e-05,
+      "loss": 0.1489,
+      "step": 3690
+    },
+    {
+      "epoch": 0.41969147005444646,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.985482202548072e-05,
+      "loss": 0.1605,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4208257713248639,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.985206134713376e-05,
+      "loss": 0.1498,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4219600725952813,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.984927479720255e-05,
+      "loss": 0.1589,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4230943738656987,
+      "grad_norm": 5.6875,
+      "learning_rate": 2.9846462380541135e-05,
+      "loss": 0.1394,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4242286751361162,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9843624102048597e-05,
+      "loss": 0.148,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4253629764065336,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.98407599666691e-05,
+      "loss": 0.149,
+      "step": 3750
+    },
+    {
+      "epoch": 0.426497277676951,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.983786997939182e-05,
+      "loss": 0.1548,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4276315789473684,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.983495414525098e-05,
+      "loss": 0.146,
+      "step": 3770
+    },
+    {
+      "epoch": 0.42876588021778583,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.9832012469325833e-05,
+      "loss": 0.1502,
+      "step": 3780
+    },
+    {
+      "epoch": 0.42990018148820325,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.982904495674062e-05,
+      "loss": 0.156,
+      "step": 3790
+    },
+    {
+      "epoch": 0.43103448275862066,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.982605161266462e-05,
+      "loss": 0.1453,
+      "step": 3800
+    },
+    {
+      "epoch": 0.43216878402903813,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9823032442312097e-05,
+      "loss": 0.1481,
+      "step": 3810
+    },
+    {
+      "epoch": 0.43330308529945555,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9819987450942293e-05,
+      "loss": 0.1392,
+      "step": 3820
+    },
+    {
+      "epoch": 0.43443738656987296,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.9816916643859443e-05,
+      "loss": 0.1522,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4355716878402904,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9813820026412754e-05,
+      "loss": 0.1488,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4367059891107078,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9810697603996373e-05,
+      "loss": 0.1413,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4378402903811252,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.9807549382049423e-05,
+      "loss": 0.1568,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4389745916515426,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.9804375366055948e-05,
+      "loss": 0.1505,
+      "step": 3870
+    },
+    {
+      "epoch": 0.4401088929219601,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9801175561544938e-05,
+      "loss": 0.1591,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4412431941923775,
+      "grad_norm": 3.75,
+      "learning_rate": 2.9797949974090294e-05,
+      "loss": 0.151,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4423774954627949,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.979469860931085e-05,
+      "loss": 0.1593,
+      "step": 3900
+    },
+    {
+      "epoch": 0.44351179673321234,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.979142147287031e-05,
+      "loss": 0.1549,
+      "step": 3910
+    },
+    {
+      "epoch": 0.44464609800362975,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9788118570477303e-05,
+      "loss": 0.145,
+      "step": 3920
+    },
+    {
+      "epoch": 0.44578039927404717,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9784789907885325e-05,
+      "loss": 0.167,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4469147005444646,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.9781435490892745e-05,
+      "loss": 0.1485,
+      "step": 3940
+    },
+    {
+      "epoch": 0.44804900181488205,
+      "grad_norm": 2.125,
+      "learning_rate": 2.97780553253428e-05,
+      "loss": 0.1489,
+      "step": 3950
+    },
+    {
+      "epoch": 0.44918330308529947,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.9774649417123576e-05,
+      "loss": 0.1462,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4503176043557169,
+      "grad_norm": 5.03125,
+      "learning_rate": 2.977121777216801e-05,
+      "loss": 0.1532,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4514519056261343,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.9767760396453853e-05,
+      "loss": 0.1583,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4525862068965517,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9764277296003693e-05,
+      "loss": 0.1598,
+      "step": 3990
+    },
+    {
+      "epoch": 0.4537205081669691,
+      "grad_norm": 3.375,
+      "learning_rate": 2.976076847688493e-05,
+      "loss": 0.146,
+      "step": 4000
+    },
+    {
+      "epoch": 0.45485480943738654,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.975723394520976e-05,
+      "loss": 0.1567,
+      "step": 4010
+    },
+    {
+      "epoch": 0.455989110707804,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.9753673707135162e-05,
+      "loss": 0.1357,
+      "step": 4020
+    },
+    {
+      "epoch": 0.45712341197822143,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.97500877688629e-05,
+      "loss": 0.1553,
+      "step": 4030
+    },
+    {
+      "epoch": 0.45825771324863884,
+      "grad_norm": 2.75,
+      "learning_rate": 2.9746476136639518e-05,
+      "loss": 0.1568,
+      "step": 4040
+    },
+    {
+      "epoch": 0.45939201451905626,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9742838816756303e-05,
+      "loss": 0.1618,
+      "step": 4050
+    },
+    {
+      "epoch": 0.4605263157894737,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.973917581554929e-05,
+      "loss": 0.1559,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4616606170598911,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.973548713939926e-05,
+      "loss": 0.1518,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4627949183303085,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9731772794731704e-05,
+      "loss": 0.1401,
+      "step": 4080
+    },
+    {
+      "epoch": 0.463929219600726,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.9728032788016836e-05,
+      "loss": 0.1428,
+      "step": 4090
+    },
+    {
+      "epoch": 0.4650635208711434,
+      "grad_norm": 2.875,
+      "learning_rate": 2.972426712576957e-05,
+      "loss": 0.1546,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4661978221415608,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.972047581454951e-05,
+      "loss": 0.1524,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4673321234119782,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9716658860960942e-05,
+      "loss": 0.1619,
+      "step": 4120
+    },
+    {
+      "epoch": 0.46846642468239563,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9712816271652816e-05,
+      "loss": 0.1502,
+      "step": 4130
+    },
+    {
+      "epoch": 0.46960072595281305,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.970894805331874e-05,
+      "loss": 0.1528,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4707350272232305,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.9705054212696965e-05,
+      "loss": 0.1445,
+      "step": 4150
+    },
+    {
+      "epoch": 0.47186932849364793,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9701134756570376e-05,
+      "loss": 0.1408,
+      "step": 4160
+    },
+    {
+      "epoch": 0.47300362976406535,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9697189691766483e-05,
+      "loss": 0.1539,
+      "step": 4170
+    },
+    {
+      "epoch": 0.47413793103448276,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.9693219025157398e-05,
+      "loss": 0.1573,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4752722323049002,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9689222763659834e-05,
+      "loss": 0.1519,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4764065335753176,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.9685200914235086e-05,
+      "loss": 0.1489,
+      "step": 4200
+    },
+    {
+      "epoch": 0.477540834845735,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9681153483889026e-05,
+      "loss": 0.162,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4786751361161525,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.967708047967209e-05,
+      "loss": 0.1507,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4798094373865699,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.9672981908679253e-05,
+      "loss": 0.1469,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4809437386569873,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.9668857778050027e-05,
+      "loss": 0.1506,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4820780399274047,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.966470809496846e-05,
+      "loss": 0.1553,
+      "step": 4250
+    },
+    {
+      "epoch": 0.48321234119782214,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.9660532866663094e-05,
+      "loss": 0.1488,
+      "step": 4260
+    },
+    {
+      "epoch": 0.48434664246823955,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.965633210040698e-05,
+      "loss": 0.1396,
+      "step": 4270
+    },
+    {
+      "epoch": 0.48548094373865697,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.965210580351766e-05,
+      "loss": 0.1578,
+      "step": 4280
+    },
+    {
+      "epoch": 0.48661524500907444,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9647853983357133e-05,
+      "loss": 0.1541,
+      "step": 4290
+    },
+    {
+      "epoch": 0.48774954627949185,
+      "grad_norm": 2.375,
+      "learning_rate": 2.964357664733187e-05,
+      "loss": 0.1564,
+      "step": 4300
+    },
+    {
+      "epoch": 0.48888384754990927,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9639273802892792e-05,
+      "loss": 0.149,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4900181488203267,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.9634945457535244e-05,
+      "loss": 0.1523,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4911524500907441,
+      "grad_norm": 4.0,
+      "learning_rate": 2.9630591618798997e-05,
+      "loss": 0.1524,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4922867513611615,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.9626212294268235e-05,
+      "loss": 0.1539,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4934210526315789,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9621807491571533e-05,
+      "loss": 0.1565,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4945553539019964,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9617377218381844e-05,
+      "loss": 0.148,
+      "step": 4360
+    },
+    {
+      "epoch": 0.4956896551724138,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.9612921482416495e-05,
+      "loss": 0.1564,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4968239564428312,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9608440291437168e-05,
+      "loss": 0.1529,
+      "step": 4380
+    },
+    {
+      "epoch": 0.49795825771324864,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.9603933653249878e-05,
+      "loss": 0.1502,
+      "step": 4390
+    },
+    {
+      "epoch": 0.49909255898366606,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.959940157570498e-05,
+      "loss": 0.1499,
+      "step": 4400
+    },
+    {
+      "epoch": 0.5002268602540835,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.9594844066697136e-05,
+      "loss": 0.1458,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5013611615245009,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.959026113416531e-05,
+      "loss": 0.147,
+      "step": 4420
+    },
+    {
+      "epoch": 0.5024954627949183,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9585652786092754e-05,
+      "loss": 0.1517,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5036297640653358,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.958101903050699e-05,
+      "loss": 0.1434,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5047640653357531,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9576359875479798e-05,
+      "loss": 0.1426,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5058983666061706,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.9571675329127203e-05,
+      "loss": 0.1481,
+      "step": 4460
+    },
+    {
+      "epoch": 0.5070326678765881,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.9566965399609468e-05,
+      "loss": 0.151,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5081669691470054,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9562230095131065e-05,
+      "loss": 0.1503,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5093012704174229,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.9557469423940662e-05,
+      "loss": 0.1531,
+      "step": 4490
+    },
+    {
+      "epoch": 0.5104355716878403,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.955268339433113e-05,
+      "loss": 0.1481,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5115698729582577,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.95478720146395e-05,
+      "loss": 0.1454,
+      "step": 4510
+    },
+    {
+      "epoch": 0.5127041742286751,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9543035293246968e-05,
+      "loss": 0.1524,
+      "step": 4520
+    },
+    {
+      "epoch": 0.5138384754990926,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9538173238578872e-05,
+      "loss": 0.1529,
+      "step": 4530
+    },
+    {
+      "epoch": 0.51497277676951,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.953328585910468e-05,
+      "loss": 0.1509,
+      "step": 4540
+    },
+    {
+      "epoch": 0.5161070780399274,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9528373163337977e-05,
+      "loss": 0.1467,
+      "step": 4550
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9523435159836437e-05,
+      "loss": 0.1541,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5183756805807622,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.9518471857201838e-05,
+      "loss": 0.1388,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5195099818511797,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9513483264080016e-05,
+      "loss": 0.1515,
+      "step": 4580
+    },
+    {
+      "epoch": 0.520644283121597,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9508469389160854e-05,
+      "loss": 0.1513,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5217785843920145,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.950343024117829e-05,
+      "loss": 0.1505,
+      "step": 4600
+    },
+    {
+      "epoch": 0.522912885662432,
+      "grad_norm": 4.15625,
+      "learning_rate": 2.949836582891028e-05,
+      "loss": 0.1563,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5240471869328494,
+      "grad_norm": 2.5,
+      "learning_rate": 2.9493276161178788e-05,
+      "loss": 0.1562,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5251814882032668,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9488161246849768e-05,
+      "loss": 0.1507,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 2.5,
+      "learning_rate": 2.948302109483317e-05,
+      "loss": 0.1632,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5274500907441017,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9477855714082884e-05,
+      "loss": 0.1526,
+      "step": 4650
+    },
+    {
+      "epoch": 0.528584392014519,
+      "grad_norm": 2.375,
+      "learning_rate": 2.9472665113596758e-05,
+      "loss": 0.1442,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5297186932849365,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.946744930241658e-05,
+      "loss": 0.1521,
+      "step": 4670
+    },
+    {
+      "epoch": 0.530852994555354,
+      "grad_norm": 2.875,
+      "learning_rate": 2.946220828962803e-05,
+      "loss": 0.1487,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5319872958257713,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.9456942084360715e-05,
+      "loss": 0.1534,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5331215970961888,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.9451650695788114e-05,
+      "loss": 0.1461,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5342558983666061,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.944633413312756e-05,
+      "loss": 0.1524,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5353901996370236,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9440992405640273e-05,
+      "loss": 0.1463,
+      "step": 4720
+    },
+    {
+      "epoch": 0.536524500907441,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.9435625522631277e-05,
+      "loss": 0.1513,
+      "step": 4730
+    },
+    {
+      "epoch": 0.5376588021778584,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.943023349344942e-05,
+      "loss": 0.1538,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5387931034482759,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9424816327487373e-05,
+      "loss": 0.156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5399274047186933,
+      "grad_norm": 2.625,
+      "learning_rate": 2.9419374034181578e-05,
+      "loss": 0.1546,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5410617059891107,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9413906623012244e-05,
+      "loss": 0.1485,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5421960072595281,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.940841410350335e-05,
+      "loss": 0.1475,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5433303085299456,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.94028964852226e-05,
+      "loss": 0.154,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5444646098003629,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9397353777781423e-05,
+      "loss": 0.1464,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5455989110707804,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.939178599083495e-05,
+      "loss": 0.1555,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5467332123411979,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.9386193134082003e-05,
+      "loss": 0.1574,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5478675136116152,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9380575217265072e-05,
+      "loss": 0.1519,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5490018148820327,
+      "grad_norm": 3.125,
+      "learning_rate": 2.93749322501703e-05,
+      "loss": 0.148,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5501361161524501,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9369264242627467e-05,
+      "loss": 0.1508,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5512704174228675,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.936357120450997e-05,
+      "loss": 0.1455,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5524047186932849,
+      "grad_norm": 3.0,
+      "learning_rate": 2.935785314573482e-05,
+      "loss": 0.1408,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5535390199637024,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.9352110076262595e-05,
+      "loss": 0.1588,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5546733212341198,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.934634200609745e-05,
+      "loss": 0.146,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5558076225045372,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.9340548945287088e-05,
+      "loss": 0.1488,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5569419237749547,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.9334730903922747e-05,
+      "loss": 0.1502,
+      "step": 4910
+    },
+    {
+      "epoch": 0.558076225045372,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.932888789213918e-05,
+      "loss": 0.1345,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5592105263157895,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.9323019920114634e-05,
+      "loss": 0.1617,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5603448275862069,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.9317126998070836e-05,
+      "loss": 0.1445,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5614791288566243,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.9311209136272977e-05,
+      "loss": 0.1419,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5626134301270418,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.930526634502969e-05,
+      "loss": 0.1555,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5637477313974592,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9299298634693033e-05,
+      "loss": 0.1436,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5648820326678766,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.9293306015658474e-05,
+      "loss": 0.1554,
+      "step": 4980
+    },
+    {
+      "epoch": 0.566016333938294,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.9287288498364876e-05,
+      "loss": 0.1545,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5671506352087115,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.928124609329446e-05,
+      "loss": 0.158,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5682849364791288,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.9275178810972817e-05,
+      "loss": 0.145,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5694192377495463,
+      "grad_norm": 2.25,
+      "learning_rate": 2.926908666196885e-05,
+      "loss": 0.1479,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5705535390199638,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.926296965689481e-05,
+      "loss": 0.1463,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5716878402903811,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.9256827806406216e-05,
+      "loss": 0.151,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5728221415607986,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9250661121201884e-05,
+      "loss": 0.1503,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5739564428312159,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9244469612023886e-05,
+      "loss": 0.1518,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5750907441016334,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9238253289657535e-05,
+      "loss": 0.1434,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5762250453720508,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9232012164931376e-05,
+      "loss": 0.1548,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5773593466424682,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.9225746248717137e-05,
+      "loss": 0.1472,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5784936479128857,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.921945555192976e-05,
+      "loss": 0.1474,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5796279491833031,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.9213140085527332e-05,
+      "loss": 0.1377,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5807622504537205,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.920679986051109e-05,
+      "loss": 0.1351,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5818965517241379,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.920043488792542e-05,
+      "loss": 0.1317,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5830308529945554,
+      "grad_norm": 2.25,
+      "learning_rate": 2.919404517885779e-05,
+      "loss": 0.1446,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5841651542649727,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9187630744438775e-05,
+      "loss": 0.1412,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5852994555353902,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.918119159584201e-05,
+      "loss": 0.1517,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5864337568058077,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.9174727744284182e-05,
+      "loss": 0.156,
+      "step": 5170
+    },
+    {
+      "epoch": 0.587568058076225,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.916823920102502e-05,
+      "loss": 0.1451,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5887023593466425,
+      "grad_norm": 1.875,
+      "learning_rate": 2.916172597736726e-05,
+      "loss": 0.1433,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5898366606170599,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.9155188084656615e-05,
+      "loss": 0.1564,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5909709618874773,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.91486255342818e-05,
+      "loss": 0.1464,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5921052631578947,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.9142038337674452e-05,
+      "loss": 0.1528,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5932395644283122,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.9135426506309157e-05,
+      "loss": 0.1532,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5943738656987296,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.9128790051703414e-05,
+      "loss": 0.1393,
+      "step": 5240
+    },
+    {
+      "epoch": 0.595508166969147,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.9122128985417612e-05,
+      "loss": 0.151,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5966424682395645,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9115443319055004e-05,
+      "loss": 0.1571,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5977767695099818,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.910873306426171e-05,
+      "loss": 0.142,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5989110707803993,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.910199823272667e-05,
+      "loss": 0.1471,
+      "step": 5280
+    },
+    {
+      "epoch": 0.6000453720508166,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9095238836181648e-05,
+      "loss": 0.1522,
+      "step": 5290
+    },
+    {
+      "epoch": 0.6011796733212341,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.9088454886401184e-05,
+      "loss": 0.14,
+      "step": 5300
+    },
+    {
+      "epoch": 0.6023139745916516,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.90816463952026e-05,
+      "loss": 0.1561,
+      "step": 5310
+    },
+    {
+      "epoch": 0.603448275862069,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9074813374445955e-05,
+      "loss": 0.1443,
+      "step": 5320
+    },
+    {
+      "epoch": 0.6045825771324864,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.9067955836034055e-05,
+      "loss": 0.1519,
+      "step": 5330
+    },
+    {
+      "epoch": 0.6057168784029038,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.9061073791912405e-05,
+      "loss": 0.1456,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6068511796733213,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.9054167254069193e-05,
+      "loss": 0.1507,
+      "step": 5350
+    },
+    {
+      "epoch": 0.6079854809437386,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9047236234535285e-05,
+      "loss": 0.1453,
+      "step": 5360
+    },
+    {
+      "epoch": 0.6091197822141561,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9040280745384188e-05,
+      "loss": 0.1458,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6102540834845736,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.9033300798732024e-05,
+      "loss": 0.1494,
+      "step": 5380
+    },
+    {
+      "epoch": 0.6113883847549909,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.9026296406737536e-05,
+      "loss": 0.1492,
+      "step": 5390
+    },
+    {
+      "epoch": 0.6125226860254084,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.9019267581602042e-05,
+      "loss": 0.1536,
+      "step": 5400
+    },
+    {
+      "epoch": 0.6136569872958257,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.9012214335569417e-05,
+      "loss": 0.1515,
+      "step": 5410
+    },
+    {
+      "epoch": 0.6147912885662432,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.900513668092608e-05,
+      "loss": 0.1413,
+      "step": 5420
+    },
+    {
+      "epoch": 0.6159255898366606,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8998034630000974e-05,
+      "loss": 0.1442,
+      "step": 5430
+    },
+    {
+      "epoch": 0.617059891107078,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.899090819516553e-05,
+      "loss": 0.1499,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6181941923774955,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8983757388833657e-05,
+      "loss": 0.1562,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6193284936479129,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.8976582223461722e-05,
+      "loss": 0.1541,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6204627949183303,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.896938271154852e-05,
+      "loss": 0.1499,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6215970961887477,
+      "grad_norm": 4.84375,
+      "learning_rate": 2.8962158865635258e-05,
+      "loss": 0.1545,
+      "step": 5480
+    },
+    {
+      "epoch": 0.6227313974591652,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.8954910698305532e-05,
+      "loss": 0.1469,
+      "step": 5490
+    },
+    {
+      "epoch": 0.6238656987295825,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.8947638222185312e-05,
+      "loss": 0.1492,
+      "step": 5500
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 2.375,
+      "learning_rate": 2.8940341449942892e-05,
+      "loss": 0.1431,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6261343012704175,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.893302039428891e-05,
+      "loss": 0.1462,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6272686025408348,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.89256750679763e-05,
+      "loss": 0.1319,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6284029038112523,
+      "grad_norm": 2.375,
+      "learning_rate": 2.8918305483800263e-05,
+      "loss": 0.1538,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6295372050816697,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.8910911654598268e-05,
+      "loss": 0.1581,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6306715063520871,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.8903493593250014e-05,
+      "loss": 0.1477,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6318058076225045,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.889605131267741e-05,
+      "loss": 0.1467,
+      "step": 5570
+    },
+    {
+      "epoch": 0.632940108892922,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.8888584825844556e-05,
+      "loss": 0.1407,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6340744101633394,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.8881094145757715e-05,
+      "loss": 0.1553,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6352087114337568,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8873579285465292e-05,
+      "loss": 0.1457,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6363430127041743,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.8866040258057824e-05,
+      "loss": 0.1486,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6374773139745916,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.8858477076667932e-05,
+      "loss": 0.1411,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6386116152450091,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8850889754470324e-05,
+      "loss": 0.1581,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6397459165154264,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8843278304681748e-05,
+      "loss": 0.1584,
+      "step": 5640
+    },
+    {
+      "epoch": 0.6408802177858439,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.883564274056099e-05,
+      "loss": 0.1508,
+      "step": 5650
+    },
+    {
+      "epoch": 0.6420145190562614,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8827983075408842e-05,
+      "loss": 0.1457,
+      "step": 5660
+    },
+    {
+      "epoch": 0.6431488203266787,
+      "grad_norm": 3.625,
+      "learning_rate": 2.882029932256807e-05,
+      "loss": 0.1391,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6442831215970962,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8812591495423413e-05,
+      "loss": 0.1492,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6454174228675136,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8804859607401533e-05,
+      "loss": 0.1622,
+      "step": 5690
+    },
+    {
+      "epoch": 0.646551724137931,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.8797103671971014e-05,
+      "loss": 0.1436,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6476860254083484,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.8789323702642333e-05,
+      "loss": 0.1476,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6488203266787659,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.878151971296781e-05,
+      "loss": 0.1434,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6499546279491834,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.877369171654164e-05,
+      "loss": 0.148,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6510889292196007,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.876583972699981e-05,
+      "loss": 0.1433,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6522232304900182,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.8757963758020122e-05,
+      "loss": 0.1501,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6533575317604355,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.8750063823322127e-05,
+      "loss": 0.1441,
+      "step": 5760
+    },
+    {
+      "epoch": 0.654491833030853,
+      "grad_norm": 2.75,
+      "learning_rate": 2.874213993666714e-05,
+      "loss": 0.1493,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6556261343012704,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.87341921118582e-05,
+      "loss": 0.1527,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6567604355716878,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8726220362740038e-05,
+      "loss": 0.1307,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.8718224703199056e-05,
+      "loss": 0.1534,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6590290381125227,
+      "grad_norm": 2.375,
+      "learning_rate": 2.871020514716332e-05,
+      "loss": 0.1371,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6601633393829401,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8702161708602506e-05,
+      "loss": 0.1483,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6612976406533575,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8694094401527912e-05,
+      "loss": 0.1431,
+      "step": 5830
+    },
+    {
+      "epoch": 0.662431941923775,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.8686003239992394e-05,
+      "loss": 0.1519,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6635662431941923,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.8677888238090373e-05,
+      "loss": 0.1468,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6647005444646098,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.8669749409957802e-05,
+      "loss": 0.1508,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6658348457350273,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8661586769772126e-05,
+      "loss": 0.1492,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6669691470054446,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.8653400331752274e-05,
+      "loss": 0.1418,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6681034482758621,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.8645190110158638e-05,
+      "loss": 0.1381,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6692377495462795,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.863695611929303e-05,
+      "loss": 0.1357,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6703720508166969,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.862869837349867e-05,
+      "loss": 0.1538,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6715063520871143,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.862041688716016e-05,
+      "loss": 0.1522,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6726406533575318,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.861211167470345e-05,
+      "loss": 0.1523,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6737749546279492,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8603782750595833e-05,
+      "loss": 0.1451,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6749092558983666,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8595430129345898e-05,
+      "loss": 0.1399,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6760435571687841,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.8587053825503505e-05,
+      "loss": 0.1362,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6771778584392014,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8578653853659787e-05,
+      "loss": 0.1525,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6783121597096189,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.8570230228447085e-05,
+      "loss": 0.1468,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6794464609800362,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.8561782964538964e-05,
+      "loss": 0.1399,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6805807622504537,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.855331207665015e-05,
+      "loss": 0.1454,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6817150635208712,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8544817579536522e-05,
+      "loss": 0.1481,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6828493647912885,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.85362994879951e-05,
+      "loss": 0.1342,
+      "step": 6020
+    },
+    {
+      "epoch": 0.683983666061706,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8527757816863992e-05,
+      "loss": 0.145,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6851179673321234,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.8519192581022377e-05,
+      "loss": 0.1463,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6862522686025408,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.851060379539049e-05,
+      "loss": 0.1587,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6873865698729582,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.850199147492959e-05,
+      "loss": 0.1414,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6885208711433757,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.849335563464193e-05,
+      "loss": 0.148,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.848469628957073e-05,
+      "loss": 0.1536,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6907894736842105,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8476013454800157e-05,
+      "loss": 0.1517,
+      "step": 6090
+    },
+    {
+      "epoch": 0.691923774954628,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.8467307145455296e-05,
+      "loss": 0.1478,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6930580762250453,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.8458577376702127e-05,
+      "loss": 0.1489,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6941923774954628,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8449824163747487e-05,
+      "loss": 0.1412,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6953266787658802,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.844104752183906e-05,
+      "loss": 0.1533,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6964609800362976,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.843224746626534e-05,
+      "loss": 0.1423,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6975952813067151,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.84234240123556e-05,
+      "loss": 0.1389,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6987295825771325,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.841457717547988e-05,
+      "loss": 0.1415,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6998638838475499,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.8405706971048953e-05,
+      "loss": 0.1395,
+      "step": 6170
+    },
+    {
+      "epoch": 0.7009981851179673,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.8396813414514283e-05,
+      "loss": 0.1342,
+      "step": 6180
+    },
+    {
+      "epoch": 0.7021324863883848,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8387896521368033e-05,
+      "loss": 0.1454,
+      "step": 6190
+    },
+    {
+      "epoch": 0.7032667876588021,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.8378956307143e-05,
+      "loss": 0.1388,
+      "step": 6200
+    },
+    {
+      "epoch": 0.7044010889292196,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.836999278741261e-05,
+      "loss": 0.1429,
+      "step": 6210
+    },
+    {
+      "epoch": 0.7055353901996371,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8361005977790896e-05,
+      "loss": 0.1515,
+      "step": 6220
+    },
+    {
+      "epoch": 0.7066696914700544,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.835199589393245e-05,
+      "loss": 0.1443,
+      "step": 6230
+    },
+    {
+      "epoch": 0.7078039927404719,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8342962551532404e-05,
+      "loss": 0.1458,
+      "step": 6240
+    },
+    {
+      "epoch": 0.7089382940108893,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.8333905966326415e-05,
+      "loss": 0.1483,
+      "step": 6250
+    },
+    {
+      "epoch": 0.7100725952813067,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.8324826154090622e-05,
+      "loss": 0.1413,
+      "step": 6260
+    },
+    {
+      "epoch": 0.7112068965517241,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.8315723130641625e-05,
+      "loss": 0.1392,
+      "step": 6270
+    },
+    {
+      "epoch": 0.7123411978221416,
+      "grad_norm": 3.5,
+      "learning_rate": 2.8306596911836467e-05,
+      "loss": 0.1477,
+      "step": 6280
+    },
+    {
+      "epoch": 0.713475499092559,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.8297447513572572e-05,
+      "loss": 0.152,
+      "step": 6290
+    },
+    {
+      "epoch": 0.7146098003629764,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.8288274951787765e-05,
+      "loss": 0.151,
+      "step": 6300
+    },
+    {
+      "epoch": 0.7157441016333939,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.8279079242460217e-05,
+      "loss": 0.1421,
+      "step": 6310
+    },
+    {
+      "epoch": 0.7168784029038112,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8269860401608406e-05,
+      "loss": 0.1412,
+      "step": 6320
+    },
+    {
+      "epoch": 0.7180127041742287,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8260618445291117e-05,
+      "loss": 0.1459,
+      "step": 6330
+    },
+    {
+      "epoch": 0.719147005444646,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.8251353389607404e-05,
+      "loss": 0.1455,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7202813067150635,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8242065250696548e-05,
+      "loss": 0.156,
+      "step": 6350
+    },
+    {
+      "epoch": 0.721415607985481,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8232754044738043e-05,
+      "loss": 0.1465,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7225499092558983,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.8223419787951573e-05,
+      "loss": 0.1458,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7236842105263158,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.8214062496596966e-05,
+      "loss": 0.1545,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7248185117967332,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.8204682186974177e-05,
+      "loss": 0.1422,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7259528130671506,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.819527887542326e-05,
+      "loss": 0.1538,
+      "step": 6400
+    },
+    {
+      "epoch": 0.727087114337568,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.8185852578324338e-05,
+      "loss": 0.1389,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7282214156079855,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.817640331209757e-05,
+      "loss": 0.1457,
+      "step": 6420
+    },
+    {
+      "epoch": 0.729355716878403,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.816693109320313e-05,
+      "loss": 0.1479,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7304900181488203,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.8157435938141168e-05,
+      "loss": 0.1373,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7316243194192378,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8147917863451802e-05,
+      "loss": 0.1523,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7327586206896551,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8138376885715057e-05,
+      "loss": 0.1473,
+      "step": 6460
+    },
+    {
+      "epoch": 0.7338929219600726,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8128813021550862e-05,
+      "loss": 0.1509,
+      "step": 6470
+    },
+    {
+      "epoch": 0.73502722323049,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8119226287619017e-05,
+      "loss": 0.1533,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7361615245009074,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.810961670061916e-05,
+      "loss": 0.1383,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7372958257713249,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.809998427729072e-05,
+      "loss": 0.1466,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7384301270417423,
+      "grad_norm": 3.25,
+      "learning_rate": 2.809032903441294e-05,
+      "loss": 0.1509,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7395644283121597,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.8080650988804782e-05,
+      "loss": 0.1404,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7406987295825771,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.807095015732494e-05,
+      "loss": 0.1653,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7418330308529946,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.8061226556871805e-05,
+      "loss": 0.1437,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7429673321234119,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.805148020438342e-05,
+      "loss": 0.1458,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7441016333938294,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.8041711116837478e-05,
+      "loss": 0.1417,
+      "step": 6560
+    },
+    {
+      "epoch": 0.7452359346642469,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.8031919311251256e-05,
+      "loss": 0.1418,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7463702359346642,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.8022104804681617e-05,
+      "loss": 0.1489,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7475045372050817,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8012267614224965e-05,
+      "loss": 0.1462,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7486388384754991,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.800240775701721e-05,
+      "loss": 0.1604,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7497731397459165,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.799252525023377e-05,
+      "loss": 0.1476,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7509074410163339,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.7982620111089486e-05,
+      "loss": 0.1526,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7520417422867514,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.7972692356838652e-05,
+      "loss": 0.1477,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7531760435571688,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.7962742004774935e-05,
+      "loss": 0.1442,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7543103448275862,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.795276907223137e-05,
+      "loss": 0.1446,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7554446460980037,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.7942773576580347e-05,
+      "loss": 0.1422,
+      "step": 6660
+    },
+    {
+      "epoch": 0.756578947368421,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.793275553523353e-05,
+      "loss": 0.1489,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7577132486388385,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.792271496564188e-05,
+      "loss": 0.1477,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7588475499092558,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.791265188529558e-05,
+      "loss": 0.1515,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7599818511796733,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.790256631172405e-05,
+      "loss": 0.1277,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7611161524500908,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.789245826249587e-05,
+      "loss": 0.1442,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7622504537205081,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.7882327755218787e-05,
+      "loss": 0.1472,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7633847549909256,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.787217480753966e-05,
+      "loss": 0.1364,
+      "step": 6730
+    },
+    {
+      "epoch": 0.764519056261343,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.7861999437144433e-05,
+      "loss": 0.143,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7656533575317604,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.785180166175813e-05,
+      "loss": 0.1431,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7667876588021778,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.784158149914479e-05,
+      "loss": 0.1351,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7679219600725953,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.7831338967107436e-05,
+      "loss": 0.1435,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7690562613430127,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.782107408348809e-05,
+      "loss": 0.1498,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7701905626134301,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.7810786866167672e-05,
+      "loss": 0.1455,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7713248638838476,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.7800477333066043e-05,
+      "loss": 0.1488,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7724591651542649,
+      "grad_norm": 3.625,
+      "learning_rate": 2.7790145502141907e-05,
+      "loss": 0.1441,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7735934664246824,
+      "grad_norm": 3.125,
+      "learning_rate": 2.777979139139283e-05,
+      "loss": 0.1442,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7747277676950998,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.7769415018855173e-05,
+      "loss": 0.1453,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7758620689655172,
+      "grad_norm": 2.875,
+      "learning_rate": 2.7759016402604084e-05,
+      "loss": 0.1354,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7769963702359347,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.7748595560753463e-05,
+      "loss": 0.1459,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7781306715063521,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.7738152511455922e-05,
+      "loss": 0.1478,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7792649727767695,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.7727687272902747e-05,
+      "loss": 0.1401,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7803992740471869,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.771719986332389e-05,
+      "loss": 0.1398,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7815335753176044,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.7706690300987926e-05,
+      "loss": 0.1389,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7826678765880217,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.7696158604202e-05,
+      "loss": 0.1462,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7838021778584392,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.7685604791311832e-05,
+      "loss": 0.1429,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7849364791288567,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.7675028880701675e-05,
+      "loss": 0.1522,
+      "step": 6920
+    },
+    {
+      "epoch": 0.786070780399274,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.766443089079424e-05,
+      "loss": 0.1394,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7872050816696915,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.7653810840050736e-05,
+      "loss": 0.1481,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7883393829401089,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.764316874697078e-05,
+      "loss": 0.1453,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.7632504630092395e-05,
+      "loss": 0.1376,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7906079854809437,
+      "grad_norm": 2.25,
+      "learning_rate": 2.7621818507991964e-05,
+      "loss": 0.1395,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7917422867513612,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.76111103992842e-05,
+      "loss": 0.1401,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7928765880217786,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.7600380322622127e-05,
+      "loss": 0.1417,
+      "step": 6990
+    },
+    {
+      "epoch": 0.794010889292196,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.758962829669702e-05,
+      "loss": 0.1417,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7951451905626135,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.7578854340238402e-05,
+      "loss": 0.1468,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7962794918330308,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.7568058472013985e-05,
+      "loss": 0.1462,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7974137931034483,
+      "grad_norm": 3.0,
+      "learning_rate": 2.755724071082967e-05,
+      "loss": 0.1498,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7985480943738656,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.754640107552947e-05,
+      "loss": 0.1345,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7996823956442831,
+      "grad_norm": 1.6171875,
+      "learning_rate": 2.7535539584995523e-05,
+      "loss": 0.1368,
+      "step": 7050
+    },
+    {
+      "epoch": 0.8008166969147006,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.7524656258148023e-05,
+      "loss": 0.1401,
+      "step": 7060
+    },
+    {
+      "epoch": 0.801950998185118,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7513751113945217e-05,
+      "loss": 0.1326,
+      "step": 7070
+    },
+    {
+      "epoch": 0.8030852994555354,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.750282417138333e-05,
+      "loss": 0.1371,
+      "step": 7080
+    },
+    {
+      "epoch": 0.8042196007259528,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.7491875449496592e-05,
+      "loss": 0.1307,
+      "step": 7090
+    },
+    {
+      "epoch": 0.8053539019963702,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.7480904967357146e-05,
+      "loss": 0.1364,
+      "step": 7100
+    },
+    {
+      "epoch": 0.8064882032667876,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.7469912744075055e-05,
+      "loss": 0.1497,
+      "step": 7110
+    },
+    {
+      "epoch": 0.8076225045372051,
+      "grad_norm": 2.375,
+      "learning_rate": 2.745889879879825e-05,
+      "loss": 0.1511,
+      "step": 7120
+    },
+    {
+      "epoch": 0.8087568058076225,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.744786315071249e-05,
+      "loss": 0.1467,
+      "step": 7130
+    },
+    {
+      "epoch": 0.8098911070780399,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7436805819041366e-05,
+      "loss": 0.1407,
+      "step": 7140
+    },
+    {
+      "epoch": 0.8110254083484574,
+      "grad_norm": 2.625,
+      "learning_rate": 2.742572682304621e-05,
+      "loss": 0.1465,
+      "step": 7150
+    },
+    {
+      "epoch": 0.8121597096188747,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.741462618202611e-05,
+      "loss": 0.1394,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8132940108892922,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.7403503915317854e-05,
+      "loss": 0.148,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8144283121597096,
+      "grad_norm": 2.5,
+      "learning_rate": 2.7392360042295902e-05,
+      "loss": 0.1367,
+      "step": 7180
+    },
+    {
+      "epoch": 0.815562613430127,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.738119458237235e-05,
+      "loss": 0.1427,
+      "step": 7190
+    },
+    {
+      "epoch": 0.8166969147005445,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.7370007554996898e-05,
+      "loss": 0.1427,
+      "step": 7200
+    },
+    {
+      "epoch": 0.8178312159709619,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.735879897965682e-05,
+      "loss": 0.1425,
+      "step": 7210
+    },
+    {
+      "epoch": 0.8189655172413793,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.734756887587691e-05,
+      "loss": 0.1248,
+      "step": 7220
+    },
+    {
+      "epoch": 0.8200998185117967,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7336317263219487e-05,
+      "loss": 0.1459,
+      "step": 7230
+    },
+    {
+      "epoch": 0.8212341197822142,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.732504416128432e-05,
+      "loss": 0.146,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8223684210526315,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.7313749589708607e-05,
+      "loss": 0.1421,
+      "step": 7250
+    },
+    {
+      "epoch": 0.823502722323049,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.7302433568166972e-05,
+      "loss": 0.1401,
+      "step": 7260
+    },
+    {
+      "epoch": 0.8246370235934665,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.7291096116371372e-05,
+      "loss": 0.1538,
+      "step": 7270
+    },
+    {
+      "epoch": 0.8257713248638838,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.7279737254071114e-05,
+      "loss": 0.1331,
+      "step": 7280
+    },
+    {
+      "epoch": 0.8269056261343013,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.7268357001052795e-05,
+      "loss": 0.1491,
+      "step": 7290
+    },
+    {
+      "epoch": 0.8280399274047187,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.7256955377140275e-05,
+      "loss": 0.142,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8291742286751361,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.724553240219464e-05,
+      "loss": 0.1454,
+      "step": 7310
+    },
+    {
+      "epoch": 0.8303085299455535,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.723408809611417e-05,
+      "loss": 0.1432,
+      "step": 7320
+    },
+    {
+      "epoch": 0.831442831215971,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.72226224788343e-05,
+      "loss": 0.1461,
+      "step": 7330
+    },
+    {
+      "epoch": 0.8325771324863884,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.7211135570327592e-05,
+      "loss": 0.143,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8337114337568058,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.719962739060369e-05,
+      "loss": 0.1635,
+      "step": 7350
+    },
+    {
+      "epoch": 0.8348457350272233,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.71880979597093e-05,
+      "loss": 0.1393,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8359800362976406,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.717654729772814e-05,
+      "loss": 0.1359,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8371143375680581,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.7164975424780914e-05,
+      "loss": 0.1436,
+      "step": 7380
+    },
+    {
+      "epoch": 0.8382486388384754,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.715338236102527e-05,
+      "loss": 0.1415,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8393829401088929,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.7141768126655777e-05,
+      "loss": 0.1401,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8405172413793104,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.7130132741903887e-05,
+      "loss": 0.1418,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8416515426497277,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.711847622703787e-05,
+      "loss": 0.1589,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8427858439201452,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7106798602362833e-05,
+      "loss": 0.1465,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8439201451905626,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.709509988822064e-05,
+      "loss": 0.1378,
+      "step": 7440
+    },
+    {
+      "epoch": 0.84505444646098,
+      "grad_norm": 1.875,
+      "learning_rate": 2.7083380104989887e-05,
+      "loss": 0.138,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8461887477313974,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.7071639273085892e-05,
+      "loss": 0.1449,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8473230490018149,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.7059877412960616e-05,
+      "loss": 0.1375,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8484573502722323,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.704809454510266e-05,
+      "loss": 0.1462,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8495916515426497,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.7036290690037223e-05,
+      "loss": 0.1371,
+      "step": 7490
+    },
+    {
+      "epoch": 0.8507259528130672,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.7024465868326057e-05,
+      "loss": 0.1327,
+      "step": 7500
+    },
+    {
+      "epoch": 0.8518602540834845,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7012620100567437e-05,
+      "loss": 0.1525,
+      "step": 7510
+    },
+    {
+      "epoch": 0.852994555353902,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.7000753407396134e-05,
+      "loss": 0.1401,
+      "step": 7520
+    },
+    {
+      "epoch": 0.8541288566243194,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.6988865809483353e-05,
+      "loss": 0.1451,
+      "step": 7530
+    },
+    {
+      "epoch": 0.8552631578947368,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.697695732753673e-05,
+      "loss": 0.1294,
+      "step": 7540
+    },
+    {
+      "epoch": 0.8563974591651543,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.6965027982300274e-05,
+      "loss": 0.139,
+      "step": 7550
+    },
+    {
+      "epoch": 0.8575317604355717,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.6953077794554335e-05,
+      "loss": 0.1349,
+      "step": 7560
+    },
+    {
+      "epoch": 0.8586660617059891,
+      "grad_norm": 2.125,
+      "learning_rate": 2.6941106785115576e-05,
+      "loss": 0.1454,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8598003629764065,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.6929114974836918e-05,
+      "loss": 0.1443,
+      "step": 7580
+    },
+    {
+      "epoch": 0.860934664246824,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.6917102384607526e-05,
+      "loss": 0.1397,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.690506903535277e-05,
+      "loss": 0.1517,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8632032667876588,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.6893014948034162e-05,
+      "loss": 0.1355,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8643375680580763,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.6880940143649346e-05,
+      "loss": 0.1473,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8654718693284936,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.6868844643232063e-05,
+      "loss": 0.1478,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8666061705989111,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.6856728467852102e-05,
+      "loss": 0.1491,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8677404718693285,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.684459163861526e-05,
+      "loss": 0.1451,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8688747731397459,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.683243417666331e-05,
+      "loss": 0.1391,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8700090744101633,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.6820256103173988e-05,
+      "loss": 0.1413,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8711433756805808,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.6808057439360903e-05,
+      "loss": 0.1417,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8722776769509982,
+      "grad_norm": 2.625,
+      "learning_rate": 2.6795838206473557e-05,
+      "loss": 0.1477,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8734119782214156,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.678359842579727e-05,
+      "loss": 0.1438,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8745462794918331,
+      "grad_norm": 2.75,
+      "learning_rate": 2.677133811865315e-05,
+      "loss": 0.1361,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8756805807622504,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.6759057306398077e-05,
+      "loss": 0.1394,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8768148820326679,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.6746756010424644e-05,
+      "loss": 0.1382,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8779491833030852,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.673443425216111e-05,
+      "loss": 0.1511,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8790834845735027,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.6722092053071406e-05,
+      "loss": 0.14,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8802177858439202,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.6709729434655054e-05,
+      "loss": 0.1501,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8813520871143375,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.669734641844714e-05,
+      "loss": 0.1454,
+      "step": 7770
+    },
+    {
+      "epoch": 0.882486388384755,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.668494302601829e-05,
+      "loss": 0.1441,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8836206896551724,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.667251927897463e-05,
+      "loss": 0.1409,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8847549909255898,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.666007519895773e-05,
+      "loss": 0.14,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8858892921960072,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.6647610807644587e-05,
+      "loss": 0.1403,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8870235934664247,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.6635126126747577e-05,
+      "loss": 0.1498,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8881578947368421,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.6622621178014422e-05,
+      "loss": 0.139,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8892921960072595,
+      "grad_norm": 3.25,
+      "learning_rate": 2.6610095983228145e-05,
+      "loss": 0.1463,
+      "step": 7840
+    },
+    {
+      "epoch": 0.890426497277677,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.6597550564207048e-05,
+      "loss": 0.1241,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8915607985480943,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.6584984942804642e-05,
+      "loss": 0.1412,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8926950998185118,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.657239914090965e-05,
+      "loss": 0.1374,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8938294010889292,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.6559793180445936e-05,
+      "loss": 0.145,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8949637023593466,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.6547167083372486e-05,
+      "loss": 0.1402,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8960980036297641,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.6534520871683354e-05,
+      "loss": 0.1426,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8972323049001815,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.6521854567407647e-05,
+      "loss": 0.1439,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8983666061705989,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.6509168192609464e-05,
+      "loss": 0.1447,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8995009074410163,
+      "grad_norm": 2.625,
+      "learning_rate": 2.6496461769387862e-05,
+      "loss": 0.144,
+      "step": 7930
+    },
+    {
+      "epoch": 0.9006352087114338,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.6483735319876827e-05,
+      "loss": 0.1482,
+      "step": 7940
+    },
+    {
+      "epoch": 0.9017695099818511,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.6470988866245228e-05,
+      "loss": 0.1417,
+      "step": 7950
+    },
+    {
+      "epoch": 0.9029038112522686,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.6458222430696782e-05,
+      "loss": 0.1501,
+      "step": 7960
+    },
+    {
+      "epoch": 0.9040381125226861,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.6445436035470014e-05,
+      "loss": 0.1418,
+      "step": 7970
+    },
+    {
+      "epoch": 0.9051724137931034,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.6432629702838213e-05,
+      "loss": 0.1517,
+      "step": 7980
+    },
+    {
+      "epoch": 0.9063067150635209,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.6419803455109406e-05,
+      "loss": 0.1488,
+      "step": 7990
+    },
+    {
+      "epoch": 0.9074410163339383,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.6406957314626298e-05,
+      "loss": 0.1348,
+      "step": 8000
+    },
+    {
+      "epoch": 0.9085753176043557,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.639409130376626e-05,
+      "loss": 0.1528,
+      "step": 8010
+    },
+    {
+      "epoch": 0.9097096188747731,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.6381205444941272e-05,
+      "loss": 0.1453,
+      "step": 8020
+    },
+    {
+      "epoch": 0.9108439201451906,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.6368299760597884e-05,
+      "loss": 0.1448,
+      "step": 8030
+    },
+    {
+      "epoch": 0.911978221415608,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.6355374273217186e-05,
+      "loss": 0.1332,
+      "step": 8040
+    },
+    {
+      "epoch": 0.9131125226860254,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.634242900531476e-05,
+      "loss": 0.1447,
+      "step": 8050
+    },
+    {
+      "epoch": 0.9142468239564429,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.6329463979440648e-05,
+      "loss": 0.1339,
+      "step": 8060
+    },
+    {
+      "epoch": 0.9153811252268602,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.6316479218179313e-05,
+      "loss": 0.1477,
+      "step": 8070
+    },
+    {
+      "epoch": 0.9165154264972777,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.630347474414958e-05,
+      "loss": 0.1368,
+      "step": 8080
+    },
+    {
+      "epoch": 0.917649727767695,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.6290450580004633e-05,
+      "loss": 0.1334,
+      "step": 8090
+    },
+    {
+      "epoch": 0.9187840290381125,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.6277406748431946e-05,
+      "loss": 0.1437,
+      "step": 8100
+    },
+    {
+      "epoch": 0.91991833030853,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.6264343272153247e-05,
+      "loss": 0.1466,
+      "step": 8110
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.62512601739245e-05,
+      "loss": 0.141,
+      "step": 8120
+    },
+    {
+      "epoch": 0.9221869328493648,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.6238157476535828e-05,
+      "loss": 0.1484,
+      "step": 8130
+    },
+    {
+      "epoch": 0.9233212341197822,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.6225035202811515e-05,
+      "loss": 0.1469,
+      "step": 8140
+    },
+    {
+      "epoch": 0.9244555353901996,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.6211893375609933e-05,
+      "loss": 0.1369,
+      "step": 8150
+    },
+    {
+      "epoch": 0.925589836660617,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.6198732017823523e-05,
+      "loss": 0.1377,
+      "step": 8160
+    },
+    {
+      "epoch": 0.9267241379310345,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.6185551152378745e-05,
+      "loss": 0.139,
+      "step": 8170
+    },
+    {
+      "epoch": 0.927858439201452,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.617235080223603e-05,
+      "loss": 0.143,
+      "step": 8180
+    },
+    {
+      "epoch": 0.9289927404718693,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.6159130990389774e-05,
+      "loss": 0.1486,
+      "step": 8190
+    },
+    {
+      "epoch": 0.9301270417422868,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.614589173986825e-05,
+      "loss": 0.1434,
+      "step": 8200
+    },
+    {
+      "epoch": 0.9312613430127041,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.6132633073733606e-05,
+      "loss": 0.1388,
+      "step": 8210
+    },
+    {
+      "epoch": 0.9323956442831216,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.6119355015081807e-05,
+      "loss": 0.1378,
+      "step": 8220
+    },
+    {
+      "epoch": 0.933529945553539,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.6106057587042604e-05,
+      "loss": 0.1457,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9346642468239564,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.609274081277947e-05,
+      "loss": 0.1434,
+      "step": 8240
+    },
+    {
+      "epoch": 0.9357985480943739,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.6079404715489616e-05,
+      "loss": 0.1452,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9369328493647913,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.6066049318403864e-05,
+      "loss": 0.1486,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9380671506352087,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.6052674644786695e-05,
+      "loss": 0.133,
+      "step": 8270
+    },
+    {
+      "epoch": 0.9392014519056261,
+      "grad_norm": 2.125,
+      "learning_rate": 2.603928071793615e-05,
+      "loss": 0.1482,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9403357531760436,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.602586756118381e-05,
+      "loss": 0.1464,
+      "step": 8290
+    },
+    {
+      "epoch": 0.941470054446461,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.601243519789476e-05,
+      "loss": 0.1483,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9426043557168784,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.599898365146754e-05,
+      "loss": 0.1439,
+      "step": 8310
+    },
+    {
+      "epoch": 0.9437386569872959,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.59855129453341e-05,
+      "loss": 0.1422,
+      "step": 8320
+    },
+    {
+      "epoch": 0.9448729582577132,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.5972023102959766e-05,
+      "loss": 0.1491,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9460072595281307,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.5958514147843208e-05,
+      "loss": 0.1433,
+      "step": 8340
+    },
+    {
+      "epoch": 0.947141560798548,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.5944986103516376e-05,
+      "loss": 0.1456,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9482758620689655,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.593143899354449e-05,
+      "loss": 0.1367,
+      "step": 8360
+    },
+    {
+      "epoch": 0.949410163339383,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.591787284152596e-05,
+      "loss": 0.1474,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9505444646098004,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.5904287671092383e-05,
+      "loss": 0.1442,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9516787658802178,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.589068350590848e-05,
+      "loss": 0.1487,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9528130671506352,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.5877060369672057e-05,
+      "loss": 0.1431,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9539473684210527,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.586341828611397e-05,
+      "loss": 0.1399,
+      "step": 8410
+    },
+    {
+      "epoch": 0.95508166969147,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.584975727899808e-05,
+      "loss": 0.1495,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9562159709618875,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.5836077372121213e-05,
+      "loss": 0.1373,
+      "step": 8430
+    },
+    {
+      "epoch": 0.957350272232305,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.5822378589313112e-05,
+      "loss": 0.1366,
+      "step": 8440
+    },
+    {
+      "epoch": 0.9584845735027223,
+      "grad_norm": 2.75,
+      "learning_rate": 2.5808660954436413e-05,
+      "loss": 0.1381,
+      "step": 8450
+    },
+    {
+      "epoch": 0.9596188747731398,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.579492449138657e-05,
+      "loss": 0.1449,
+      "step": 8460
+    },
+    {
+      "epoch": 0.9607531760435571,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.5781169224091858e-05,
+      "loss": 0.1442,
+      "step": 8470
+    },
+    {
+      "epoch": 0.9618874773139746,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.57673951765133e-05,
+      "loss": 0.1361,
+      "step": 8480
+    },
+    {
+      "epoch": 0.963021778584392,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.575360237264462e-05,
+      "loss": 0.1525,
+      "step": 8490
+    },
+    {
+      "epoch": 0.9641560798548094,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.573979083651224e-05,
+      "loss": 0.1443,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9652903811252269,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.572596059217519e-05,
+      "loss": 0.139,
+      "step": 8510
+    },
+    {
+      "epoch": 0.9664246823956443,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.57121116637251e-05,
+      "loss": 0.1433,
+      "step": 8520
+    },
+    {
+      "epoch": 0.9675589836660617,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.569824407528614e-05,
+      "loss": 0.1398,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9686932849364791,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.5684357851014998e-05,
+      "loss": 0.1436,
+      "step": 8540
+    },
+    {
+      "epoch": 0.9698275862068966,
+      "grad_norm": 2.5,
+      "learning_rate": 2.567045301510081e-05,
+      "loss": 0.1479,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9709618874773139,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.5656529591765142e-05,
+      "loss": 0.1372,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9720961887477314,
+      "grad_norm": 2.125,
+      "learning_rate": 2.5642587605261933e-05,
+      "loss": 0.1278,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9732304900181489,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.562862707987746e-05,
+      "loss": 0.1412,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9743647912885662,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.5614648039930294e-05,
+      "loss": 0.1403,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9754990925589837,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.560065050977126e-05,
+      "loss": 0.1418,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9766333938294011,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.5586634513783382e-05,
+      "loss": 0.1471,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9777676950998185,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.557260007638187e-05,
+      "loss": 0.1391,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9789019963702359,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.555854722201404e-05,
+      "loss": 0.1447,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9800362976406534,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.5544475975159292e-05,
+      "loss": 0.1356,
+      "step": 8640
+    },
+    {
+      "epoch": 0.9811705989110708,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.553038636032907e-05,
+      "loss": 0.1309,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9823049001814882,
+      "grad_norm": 2.5,
+      "learning_rate": 2.5516278402066823e-05,
+      "loss": 0.1456,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9834392014519057,
+      "grad_norm": 2.25,
+      "learning_rate": 2.5502152124947927e-05,
+      "loss": 0.1422,
+      "step": 8670
+    },
+    {
+      "epoch": 0.984573502722323,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.5488007553579697e-05,
+      "loss": 0.1515,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9857078039927405,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.54738447126013e-05,
+      "loss": 0.1359,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9868421052631579,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.5459663626683728e-05,
+      "loss": 0.1488,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9879764065335753,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.5445464320529756e-05,
+      "loss": 0.1399,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9891107078039928,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.543124681887391e-05,
+      "loss": 0.1339,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9902450090744102,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.541701114648239e-05,
+      "loss": 0.1406,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9913793103448276,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.540275732815307e-05,
+      "loss": 0.1445,
+      "step": 8740
+    },
+    {
+      "epoch": 0.992513611615245,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.5388485388715415e-05,
+      "loss": 0.1467,
+      "step": 8750
+    },
+    {
+      "epoch": 0.9936479128856625,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.5374195353030473e-05,
+      "loss": 0.1421,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9947822141560798,
+      "grad_norm": 2.375,
+      "learning_rate": 2.53598872459908e-05,
+      "loss": 0.1482,
+      "step": 8770
+    },
+    {
+      "epoch": 0.9959165154264973,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.534556109252044e-05,
+      "loss": 0.1485,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9970508166969148,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.5331216917574864e-05,
+      "loss": 0.1352,
+      "step": 8790
+    },
+    {
+      "epoch": 0.9981851179673321,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.5316854746140955e-05,
+      "loss": 0.1428,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9993194192377496,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.5302474603236922e-05,
+      "loss": 0.1413,
+      "step": 8810
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9400585976965043,
+      "eval_f1": 0.9377027870372914,
+      "eval_loss": 0.1423337161540985,
+      "eval_precision": 0.9371459150383395,
+      "eval_recall": 0.9400585976965043,
+      "eval_runtime": 122.6099,
+      "eval_samples_per_second": 484.366,
+      "eval_steps_per_second": 3.784,
+      "step": 8816
+    },
+    {
+      "epoch": 1.000453720508167,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.5288076513912294e-05,
+      "loss": 0.1327,
+      "step": 8820
+    },
+    {
+      "epoch": 1.0015880217785844,
+      "grad_norm": 2.125,
+      "learning_rate": 2.5273660503247857e-05,
+      "loss": 0.1193,
+      "step": 8830
+    },
+    {
+      "epoch": 1.0027223230490019,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.5259226596355615e-05,
+      "loss": 0.1165,
+      "step": 8840
+    },
+    {
+      "epoch": 1.0038566243194191,
+      "grad_norm": 2.125,
+      "learning_rate": 2.524477481837875e-05,
+      "loss": 0.1215,
+      "step": 8850
+    },
+    {
+      "epoch": 1.0049909255898366,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.5230305194491567e-05,
+      "loss": 0.1182,
+      "step": 8860
+    },
+    {
+      "epoch": 1.006125226860254,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.5215817749899464e-05,
+      "loss": 0.1184,
+      "step": 8870
+    },
+    {
+      "epoch": 1.0072595281306715,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.5201312509838884e-05,
+      "loss": 0.1134,
+      "step": 8880
+    },
+    {
+      "epoch": 1.008393829401089,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.5186789499577258e-05,
+      "loss": 0.1226,
+      "step": 8890
+    },
+    {
+      "epoch": 1.0095281306715063,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.5172248744412985e-05,
+      "loss": 0.1226,
+      "step": 8900
+    },
+    {
+      "epoch": 1.0106624319419237,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.5157690269675372e-05,
+      "loss": 0.1137,
+      "step": 8910
+    },
+    {
+      "epoch": 1.0117967332123412,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.514311410072458e-05,
+      "loss": 0.1222,
+      "step": 8920
+    },
+    {
+      "epoch": 1.0129310344827587,
+      "grad_norm": 2.375,
+      "learning_rate": 2.5128520262951603e-05,
+      "loss": 0.1105,
+      "step": 8930
+    },
+    {
+      "epoch": 1.0140653357531761,
+      "grad_norm": 3.125,
+      "learning_rate": 2.5113908781778216e-05,
+      "loss": 0.1255,
+      "step": 8940
+    },
+    {
+      "epoch": 1.0151996370235934,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.509927968265692e-05,
+      "loss": 0.107,
+      "step": 8950
+    },
+    {
+      "epoch": 1.0163339382940109,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.508463299107092e-05,
+      "loss": 0.1135,
+      "step": 8960
+    },
+    {
+      "epoch": 1.0174682395644283,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.5069968732534043e-05,
+      "loss": 0.1173,
+      "step": 8970
+    },
+    {
+      "epoch": 1.0186025408348458,
+      "grad_norm": 2.875,
+      "learning_rate": 2.5055286932590738e-05,
+      "loss": 0.1157,
+      "step": 8980
+    },
+    {
+      "epoch": 1.019736842105263,
+      "grad_norm": 2.625,
+      "learning_rate": 2.5040587616815996e-05,
+      "loss": 0.1069,
+      "step": 8990
+    },
+    {
+      "epoch": 1.0208711433756805,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.502587081081533e-05,
+      "loss": 0.1204,
+      "step": 9000
+    },
+    {
+      "epoch": 1.022005444646098,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.501113654022472e-05,
+      "loss": 0.1145,
+      "step": 9010
+    },
+    {
+      "epoch": 1.0231397459165155,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.499638483071055e-05,
+      "loss": 0.1217,
+      "step": 9020
+    },
+    {
+      "epoch": 1.024274047186933,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.4981615707969608e-05,
+      "loss": 0.1137,
+      "step": 9030
+    },
+    {
+      "epoch": 1.0254083484573502,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.4966829197729002e-05,
+      "loss": 0.1133,
+      "step": 9040
+    },
+    {
+      "epoch": 1.0265426497277677,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.495202532574613e-05,
+      "loss": 0.1122,
+      "step": 9050
+    },
+    {
+      "epoch": 1.0276769509981851,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.493720411780863e-05,
+      "loss": 0.1114,
+      "step": 9060
+    },
+    {
+      "epoch": 1.0288112522686026,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.4922365599734338e-05,
+      "loss": 0.1109,
+      "step": 9070
+    },
+    {
+      "epoch": 1.02994555353902,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.490750979737125e-05,
+      "loss": 0.1119,
+      "step": 9080
+    },
+    {
+      "epoch": 1.0310798548094373,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.4892636736597467e-05,
+      "loss": 0.1149,
+      "step": 9090
+    },
+    {
+      "epoch": 1.0322141560798548,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.487774644332115e-05,
+      "loss": 0.1122,
+      "step": 9100
+    },
+    {
+      "epoch": 1.0333484573502723,
+      "grad_norm": 3.375,
+      "learning_rate": 2.4862838943480492e-05,
+      "loss": 0.122,
+      "step": 9110
+    },
+    {
+      "epoch": 1.0344827586206897,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.4847914263043637e-05,
+      "loss": 0.1129,
+      "step": 9120
+    },
+    {
+      "epoch": 1.035617059891107,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.4832972428008673e-05,
+      "loss": 0.1181,
+      "step": 9130
+    },
+    {
+      "epoch": 1.0367513611615244,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.4818013464403565e-05,
+      "loss": 0.1272,
+      "step": 9140
+    },
+    {
+      "epoch": 1.037885662431942,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.4803037398286114e-05,
+      "loss": 0.1194,
+      "step": 9150
+    },
+    {
+      "epoch": 1.0390199637023594,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.4788044255743917e-05,
+      "loss": 0.1141,
+      "step": 9160
+    },
+    {
+      "epoch": 1.0401542649727769,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.4773034062894315e-05,
+      "loss": 0.1138,
+      "step": 9170
+    },
+    {
+      "epoch": 1.041288566243194,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.4758006845884352e-05,
+      "loss": 0.0993,
+      "step": 9180
+    },
+    {
+      "epoch": 1.0424228675136116,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.4742962630890718e-05,
+      "loss": 0.1084,
+      "step": 9190
+    },
+    {
+      "epoch": 1.043557168784029,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.472790144411973e-05,
+      "loss": 0.1093,
+      "step": 9200
+    },
+    {
+      "epoch": 1.0446914700544465,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.4712823311807246e-05,
+      "loss": 0.115,
+      "step": 9210
+    },
+    {
+      "epoch": 1.045825771324864,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.4697728260218668e-05,
+      "loss": 0.117,
+      "step": 9220
+    },
+    {
+      "epoch": 1.0469600725952812,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.468261631564885e-05,
+      "loss": 0.1178,
+      "step": 9230
+    },
+    {
+      "epoch": 1.0480943738656987,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.466748750442208e-05,
+      "loss": 0.1125,
+      "step": 9240
+    },
+    {
+      "epoch": 1.0492286751361162,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.465234185289203e-05,
+      "loss": 0.1132,
+      "step": 9250
+    },
+    {
+      "epoch": 1.0503629764065336,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.4637179387441707e-05,
+      "loss": 0.1103,
+      "step": 9260
+    },
+    {
+      "epoch": 1.051497277676951,
+      "grad_norm": 3.375,
+      "learning_rate": 2.4622000134483396e-05,
+      "loss": 0.1093,
+      "step": 9270
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.460680412045864e-05,
+      "loss": 0.1134,
+      "step": 9280
+    },
+    {
+      "epoch": 1.0537658802177858,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.4591591371838176e-05,
+      "loss": 0.1176,
+      "step": 9290
+    },
+    {
+      "epoch": 1.0549001814882033,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.457636191512188e-05,
+      "loss": 0.1214,
+      "step": 9300
+    },
+    {
+      "epoch": 1.0560344827586208,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.4561115776838743e-05,
+      "loss": 0.1086,
+      "step": 9310
+    },
+    {
+      "epoch": 1.057168784029038,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.454585298354682e-05,
+      "loss": 0.1071,
+      "step": 9320
+    },
+    {
+      "epoch": 1.0583030852994555,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.4530573561833157e-05,
+      "loss": 0.1196,
+      "step": 9330
+    },
+    {
+      "epoch": 1.059437386569873,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.4515277538313795e-05,
+      "loss": 0.1173,
+      "step": 9340
+    },
+    {
+      "epoch": 1.0605716878402904,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.449996493963367e-05,
+      "loss": 0.1147,
+      "step": 9350
+    },
+    {
+      "epoch": 1.061705989110708,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.44846357924666e-05,
+      "loss": 0.1114,
+      "step": 9360
+    },
+    {
+      "epoch": 1.0628402903811252,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.4469290123515228e-05,
+      "loss": 0.1134,
+      "step": 9370
+    },
+    {
+      "epoch": 1.0639745916515426,
+      "grad_norm": 3.25,
+      "learning_rate": 2.445392795951099e-05,
+      "loss": 0.1099,
+      "step": 9380
+    },
+    {
+      "epoch": 1.06510889292196,
+      "grad_norm": 3.125,
+      "learning_rate": 2.4438549327214036e-05,
+      "loss": 0.1087,
+      "step": 9390
+    },
+    {
+      "epoch": 1.0662431941923776,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.4423154253413218e-05,
+      "loss": 0.1149,
+      "step": 9400
+    },
+    {
+      "epoch": 1.0673774954627948,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.4407742764926017e-05,
+      "loss": 0.1112,
+      "step": 9410
+    },
+    {
+      "epoch": 1.0685117967332123,
+      "grad_norm": 2.875,
+      "learning_rate": 2.439231488859851e-05,
+      "loss": 0.1165,
+      "step": 9420
+    },
+    {
+      "epoch": 1.0696460980036298,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.4376870651305326e-05,
+      "loss": 0.1032,
+      "step": 9430
+    },
+    {
+      "epoch": 1.0707803992740472,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.4361410079949598e-05,
+      "loss": 0.11,
+      "step": 9440
+    },
+    {
+      "epoch": 1.0719147005444647,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.4345933201462894e-05,
+      "loss": 0.1194,
+      "step": 9450
+    },
+    {
+      "epoch": 1.073049001814882,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.4330440042805213e-05,
+      "loss": 0.1193,
+      "step": 9460
+    },
+    {
+      "epoch": 1.0741833030852994,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.4314930630964894e-05,
+      "loss": 0.1156,
+      "step": 9470
+    },
+    {
+      "epoch": 1.0753176043557169,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.429940499295859e-05,
+      "loss": 0.1213,
+      "step": 9480
+    },
+    {
+      "epoch": 1.0764519056261344,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.428386315583123e-05,
+      "loss": 0.1186,
+      "step": 9490
+    },
+    {
+      "epoch": 1.0775862068965518,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.4268305146655955e-05,
+      "loss": 0.1101,
+      "step": 9500
+    },
+    {
+      "epoch": 1.078720508166969,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.4252730992534073e-05,
+      "loss": 0.1036,
+      "step": 9510
+    },
+    {
+      "epoch": 1.0798548094373865,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.4237140720595023e-05,
+      "loss": 0.1117,
+      "step": 9520
+    },
+    {
+      "epoch": 1.080989110707804,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.4221534357996315e-05,
+      "loss": 0.1115,
+      "step": 9530
+    },
+    {
+      "epoch": 1.0821234119782215,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.420591193192349e-05,
+      "loss": 0.1182,
+      "step": 9540
+    },
+    {
+      "epoch": 1.083257713248639,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.4190273469590066e-05,
+      "loss": 0.1124,
+      "step": 9550
+    },
+    {
+      "epoch": 1.0843920145190562,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.417461899823751e-05,
+      "loss": 0.1155,
+      "step": 9560
+    },
+    {
+      "epoch": 1.0855263157894737,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.4158948545135163e-05,
+      "loss": 0.1123,
+      "step": 9570
+    },
+    {
+      "epoch": 1.0866606170598911,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.414326213758021e-05,
+      "loss": 0.115,
+      "step": 9580
+    },
+    {
+      "epoch": 1.0877949183303086,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.4127559802897618e-05,
+      "loss": 0.1122,
+      "step": 9590
+    },
+    {
+      "epoch": 1.0889292196007259,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.411184156844012e-05,
+      "loss": 0.1092,
+      "step": 9600
+    },
+    {
+      "epoch": 1.0900635208711433,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.4096107461588118e-05,
+      "loss": 0.1127,
+      "step": 9610
+    },
+    {
+      "epoch": 1.0911978221415608,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.408035750974969e-05,
+      "loss": 0.119,
+      "step": 9620
+    },
+    {
+      "epoch": 1.0923321234119783,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.40645917403605e-05,
+      "loss": 0.109,
+      "step": 9630
+    },
+    {
+      "epoch": 1.0934664246823957,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.4048810180883764e-05,
+      "loss": 0.1138,
+      "step": 9640
+    },
+    {
+      "epoch": 1.094600725952813,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.4033012858810213e-05,
+      "loss": 0.1119,
+      "step": 9650
+    },
+    {
+      "epoch": 1.0957350272232305,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.4017199801658024e-05,
+      "loss": 0.1175,
+      "step": 9660
+    },
+    {
+      "epoch": 1.096869328493648,
+      "grad_norm": 2.625,
+      "learning_rate": 2.4001371036972796e-05,
+      "loss": 0.1089,
+      "step": 9670
+    },
+    {
+      "epoch": 1.0980036297640654,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.3985526592327484e-05,
+      "loss": 0.1134,
+      "step": 9680
+    },
+    {
+      "epoch": 1.0991379310344827,
+      "grad_norm": 3.25,
+      "learning_rate": 2.3969666495322352e-05,
+      "loss": 0.1093,
+      "step": 9690
+    },
+    {
+      "epoch": 1.1002722323049001,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.395379077358494e-05,
+      "loss": 0.1085,
+      "step": 9700
+    },
+    {
+      "epoch": 1.1014065335753176,
+      "grad_norm": 3.625,
+      "learning_rate": 2.393789945476999e-05,
+      "loss": 0.109,
+      "step": 9710
+    },
+    {
+      "epoch": 1.102540834845735,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.3921992566559434e-05,
+      "loss": 0.1167,
+      "step": 9720
+    },
+    {
+      "epoch": 1.1036751361161525,
+      "grad_norm": 4.5,
+      "learning_rate": 2.390607013666231e-05,
+      "loss": 0.116,
+      "step": 9730
+    },
+    {
+      "epoch": 1.1048094373865698,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.3890132192814736e-05,
+      "loss": 0.1107,
+      "step": 9740
+    },
+    {
+      "epoch": 1.1059437386569873,
+      "grad_norm": 3.375,
+      "learning_rate": 2.387417876277985e-05,
+      "loss": 0.116,
+      "step": 9750
+    },
+    {
+      "epoch": 1.1070780399274047,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3858209874347764e-05,
+      "loss": 0.1064,
+      "step": 9760
+    },
+    {
+      "epoch": 1.1082123411978222,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.384222555533553e-05,
+      "loss": 0.1036,
+      "step": 9770
+    },
+    {
+      "epoch": 1.1093466424682397,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.382622583358707e-05,
+      "loss": 0.1197,
+      "step": 9780
+    },
+    {
+      "epoch": 1.110480943738657,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.3810210736973127e-05,
+      "loss": 0.116,
+      "step": 9790
+    },
+    {
+      "epoch": 1.1116152450090744,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.3794180293391253e-05,
+      "loss": 0.1046,
+      "step": 9800
+    },
+    {
+      "epoch": 1.1127495462794919,
+      "grad_norm": 3.375,
+      "learning_rate": 2.3778134530765707e-05,
+      "loss": 0.1217,
+      "step": 9810
+    },
+    {
+      "epoch": 1.1138838475499093,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.376207347704745e-05,
+      "loss": 0.1053,
+      "step": 9820
+    },
+    {
+      "epoch": 1.1150181488203268,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.3745997160214067e-05,
+      "loss": 0.1044,
+      "step": 9830
+    },
+    {
+      "epoch": 1.116152450090744,
+      "grad_norm": 4.125,
+      "learning_rate": 2.3729905608269747e-05,
+      "loss": 0.1198,
+      "step": 9840
+    },
+    {
+      "epoch": 1.1172867513611615,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.37137988492452e-05,
+      "loss": 0.1161,
+      "step": 9850
+    },
+    {
+      "epoch": 1.118421052631579,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.369767691119763e-05,
+      "loss": 0.1146,
+      "step": 9860
+    },
+    {
+      "epoch": 1.1195553539019965,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.3681539822210694e-05,
+      "loss": 0.1147,
+      "step": 9870
+    },
+    {
+      "epoch": 1.1206896551724137,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.3665387610394435e-05,
+      "loss": 0.1059,
+      "step": 9880
+    },
+    {
+      "epoch": 1.1218239564428312,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.3649220303885235e-05,
+      "loss": 0.1163,
+      "step": 9890
+    },
+    {
+      "epoch": 1.1229582577132486,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.3633037930845773e-05,
+      "loss": 0.1048,
+      "step": 9900
+    },
+    {
+      "epoch": 1.1240925589836661,
+      "grad_norm": 4.125,
+      "learning_rate": 2.3616840519464968e-05,
+      "loss": 0.1172,
+      "step": 9910
+    },
+    {
+      "epoch": 1.1252268602540836,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.360062809795794e-05,
+      "loss": 0.1112,
+      "step": 9920
+    },
+    {
+      "epoch": 1.1263611615245008,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.358440069456596e-05,
+      "loss": 0.1107,
+      "step": 9930
+    },
+    {
+      "epoch": 1.1274954627949183,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.3568158337556396e-05,
+      "loss": 0.1173,
+      "step": 9940
+    },
+    {
+      "epoch": 1.1286297640653358,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.355190105522266e-05,
+      "loss": 0.1113,
+      "step": 9950
+    },
+    {
+      "epoch": 1.1297640653357532,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.3535628875884155e-05,
+      "loss": 0.1138,
+      "step": 9960
+    },
+    {
+      "epoch": 1.1308983666061705,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.351934182788625e-05,
+      "loss": 0.1123,
+      "step": 9970
+    },
+    {
+      "epoch": 1.132032667876588,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.350303993960021e-05,
+      "loss": 0.1111,
+      "step": 9980
+    },
+    {
+      "epoch": 1.1331669691470054,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.348672323942315e-05,
+      "loss": 0.1053,
+      "step": 9990
+    },
+    {
+      "epoch": 1.134301270417423,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.3470391755777978e-05,
+      "loss": 0.1101,
+      "step": 10000
+    },
+    {
+      "epoch": 1.1354355716878404,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.3454045517113363e-05,
+      "loss": 0.109,
+      "step": 10010
+    },
+    {
+      "epoch": 1.1365698729582576,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.343768455190368e-05,
+      "loss": 0.1091,
+      "step": 10020
+    },
+    {
+      "epoch": 1.137704174228675,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.342130888864895e-05,
+      "loss": 0.1047,
+      "step": 10030
+    },
+    {
+      "epoch": 1.1388384754990926,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.3404918555874793e-05,
+      "loss": 0.1128,
+      "step": 10040
+    },
+    {
+      "epoch": 1.13997277676951,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.33885135821324e-05,
+      "loss": 0.1087,
+      "step": 10050
+    },
+    {
+      "epoch": 1.1411070780399273,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.337209399599845e-05,
+      "loss": 0.1098,
+      "step": 10060
+    },
+    {
+      "epoch": 1.1422413793103448,
+      "grad_norm": 3.5,
+      "learning_rate": 2.3355659826075077e-05,
+      "loss": 0.1163,
+      "step": 10070
+    },
+    {
+      "epoch": 1.1433756805807622,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.3339211100989822e-05,
+      "loss": 0.1157,
+      "step": 10080
+    },
+    {
+      "epoch": 1.1445099818511797,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.3322747849395595e-05,
+      "loss": 0.1067,
+      "step": 10090
+    },
+    {
+      "epoch": 1.1456442831215972,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.330627009997058e-05,
+      "loss": 0.1128,
+      "step": 10100
+    },
+    {
+      "epoch": 1.1467785843920146,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.3289777881418242e-05,
+      "loss": 0.114,
+      "step": 10110
+    },
+    {
+      "epoch": 1.1479128856624319,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.327327122246724e-05,
+      "loss": 0.1116,
+      "step": 10120
+    },
+    {
+      "epoch": 1.1490471869328494,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.3256750151871392e-05,
+      "loss": 0.1244,
+      "step": 10130
+    },
+    {
+      "epoch": 1.1501814882032668,
+      "grad_norm": 3.625,
+      "learning_rate": 2.324021469840961e-05,
+      "loss": 0.1123,
+      "step": 10140
+    },
+    {
+      "epoch": 1.1513157894736843,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.3223664890885863e-05,
+      "loss": 0.1157,
+      "step": 10150
+    },
+    {
+      "epoch": 1.1524500907441015,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.3207100758129142e-05,
+      "loss": 0.1138,
+      "step": 10160
+    },
+    {
+      "epoch": 1.153584392014519,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.3190522328993368e-05,
+      "loss": 0.1219,
+      "step": 10170
+    },
+    {
+      "epoch": 1.1547186932849365,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.3173929632357384e-05,
+      "loss": 0.1202,
+      "step": 10180
+    },
+    {
+      "epoch": 1.155852994555354,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.315732269712487e-05,
+      "loss": 0.1164,
+      "step": 10190
+    },
+    {
+      "epoch": 1.1569872958257714,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.3140701552224322e-05,
+      "loss": 0.1069,
+      "step": 10200
+    },
+    {
+      "epoch": 1.1581215970961887,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.3124066226608982e-05,
+      "loss": 0.1171,
+      "step": 10210
+    },
+    {
+      "epoch": 1.1592558983666061,
+      "grad_norm": 2.875,
+      "learning_rate": 2.31074167492568e-05,
+      "loss": 0.1112,
+      "step": 10220
+    },
+    {
+      "epoch": 1.1603901996370236,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.309075314917036e-05,
+      "loss": 0.1074,
+      "step": 10230
+    },
+    {
+      "epoch": 1.161524500907441,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.307407545537687e-05,
+      "loss": 0.1097,
+      "step": 10240
+    },
+    {
+      "epoch": 1.1626588021778583,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.305738369692808e-05,
+      "loss": 0.1123,
+      "step": 10250
+    },
+    {
+      "epoch": 1.1637931034482758,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.304067790290025e-05,
+      "loss": 0.1149,
+      "step": 10260
+    },
+    {
+      "epoch": 1.1649274047186933,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.3023958102394056e-05,
+      "loss": 0.1195,
+      "step": 10270
+    },
+    {
+      "epoch": 1.1660617059891107,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.3007224324534607e-05,
+      "loss": 0.1117,
+      "step": 10280
+    },
+    {
+      "epoch": 1.1671960072595282,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.2990476598471343e-05,
+      "loss": 0.1161,
+      "step": 10290
+    },
+    {
+      "epoch": 1.1683303085299455,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.297371495337801e-05,
+      "loss": 0.1047,
+      "step": 10300
+    },
+    {
+      "epoch": 1.169464609800363,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.29569394184526e-05,
+      "loss": 0.1094,
+      "step": 10310
+    },
+    {
+      "epoch": 1.1705989110707804,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.2940150022917287e-05,
+      "loss": 0.1077,
+      "step": 10320
+    },
+    {
+      "epoch": 1.1717332123411979,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.292334679601841e-05,
+      "loss": 0.1146,
+      "step": 10330
+    },
+    {
+      "epoch": 1.1728675136116153,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.2906529767026383e-05,
+      "loss": 0.1164,
+      "step": 10340
+    },
+    {
+      "epoch": 1.1740018148820326,
+      "grad_norm": 3.125,
+      "learning_rate": 2.2889698965235667e-05,
+      "loss": 0.1192,
+      "step": 10350
+    },
+    {
+      "epoch": 1.17513611615245,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.287285441996473e-05,
+      "loss": 0.112,
+      "step": 10360
+    },
+    {
+      "epoch": 1.1762704174228675,
+      "grad_norm": 3.875,
+      "learning_rate": 2.2855996160555958e-05,
+      "loss": 0.106,
+      "step": 10370
+    },
+    {
+      "epoch": 1.177404718693285,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.2839124216375642e-05,
+      "loss": 0.1115,
+      "step": 10380
+    },
+    {
+      "epoch": 1.1785390199637025,
+      "grad_norm": 3.625,
+      "learning_rate": 2.282223861681391e-05,
+      "loss": 0.1174,
+      "step": 10390
+    },
+    {
+      "epoch": 1.1796733212341197,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.2805339391284664e-05,
+      "loss": 0.111,
+      "step": 10400
+    },
+    {
+      "epoch": 1.1808076225045372,
+      "grad_norm": 2.75,
+      "learning_rate": 2.2788426569225553e-05,
+      "loss": 0.111,
+      "step": 10410
+    },
+    {
+      "epoch": 1.1819419237749547,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.277150018009791e-05,
+      "loss": 0.1182,
+      "step": 10420
+    },
+    {
+      "epoch": 1.1830762250453721,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.2754560253386702e-05,
+      "loss": 0.1087,
+      "step": 10430
+    },
+    {
+      "epoch": 1.1842105263157894,
+      "grad_norm": 2.875,
+      "learning_rate": 2.2737606818600477e-05,
+      "loss": 0.1108,
+      "step": 10440
+    },
+    {
+      "epoch": 1.1853448275862069,
+      "grad_norm": 2.875,
+      "learning_rate": 2.27206399052713e-05,
+      "loss": 0.1139,
+      "step": 10450
+    },
+    {
+      "epoch": 1.1864791288566243,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.2703659542954735e-05,
+      "loss": 0.1268,
+      "step": 10460
+    },
+    {
+      "epoch": 1.1876134301270418,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.2686665761229772e-05,
+      "loss": 0.1029,
+      "step": 10470
+    },
+    {
+      "epoch": 1.1887477313974593,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.2669658589698753e-05,
+      "loss": 0.1098,
+      "step": 10480
+    },
+    {
+      "epoch": 1.1898820326678765,
+      "grad_norm": 2.625,
+      "learning_rate": 2.2652638057987382e-05,
+      "loss": 0.1091,
+      "step": 10490
+    },
+    {
+      "epoch": 1.191016333938294,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.2635604195744603e-05,
+      "loss": 0.109,
+      "step": 10500
+    },
+    {
+      "epoch": 1.1921506352087115,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.2618557032642593e-05,
+      "loss": 0.1198,
+      "step": 10510
+    },
+    {
+      "epoch": 1.193284936479129,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.2601496598376704e-05,
+      "loss": 0.1134,
+      "step": 10520
+    },
+    {
+      "epoch": 1.1944192377495462,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.2584422922665397e-05,
+      "loss": 0.1126,
+      "step": 10530
+    },
+    {
+      "epoch": 1.1955535390199636,
+      "grad_norm": 3.125,
+      "learning_rate": 2.2567336035250206e-05,
+      "loss": 0.1045,
+      "step": 10540
+    },
+    {
+      "epoch": 1.196687840290381,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.255023596589567e-05,
+      "loss": 0.1034,
+      "step": 10550
+    },
+    {
+      "epoch": 1.1978221415607986,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.2533122744389305e-05,
+      "loss": 0.1211,
+      "step": 10560
+    },
+    {
+      "epoch": 1.198956442831216,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.2515996400541527e-05,
+      "loss": 0.1096,
+      "step": 10570
+    },
+    {
+      "epoch": 1.2000907441016333,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.2498856964185605e-05,
+      "loss": 0.1083,
+      "step": 10580
+    },
+    {
+      "epoch": 1.2012250453720508,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.248170446517763e-05,
+      "loss": 0.1159,
+      "step": 10590
+    },
+    {
+      "epoch": 1.2023593466424682,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.2464538933396432e-05,
+      "loss": 0.1178,
+      "step": 10600
+    },
+    {
+      "epoch": 1.2034936479128857,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.244736039874355e-05,
+      "loss": 0.1129,
+      "step": 10610
+    },
+    {
+      "epoch": 1.2046279491833032,
+      "grad_norm": 3.5,
+      "learning_rate": 2.2430168891143184e-05,
+      "loss": 0.1203,
+      "step": 10620
+    },
+    {
+      "epoch": 1.2057622504537204,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.241296444054212e-05,
+      "loss": 0.1019,
+      "step": 10630
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.2395747076909682e-05,
+      "loss": 0.1109,
+      "step": 10640
+    },
+    {
+      "epoch": 1.2080308529945554,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.237851683023771e-05,
+      "loss": 0.1167,
+      "step": 10650
+    },
+    {
+      "epoch": 1.2091651542649728,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.2361273730540467e-05,
+      "loss": 0.1187,
+      "step": 10660
+    },
+    {
+      "epoch": 1.2102994555353903,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.2344017807854612e-05,
+      "loss": 0.1109,
+      "step": 10670
+    },
+    {
+      "epoch": 1.2114337568058076,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.2326749092239145e-05,
+      "loss": 0.1108,
+      "step": 10680
+    },
+    {
+      "epoch": 1.212568058076225,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.2309467613775347e-05,
+      "loss": 0.113,
+      "step": 10690
+    },
+    {
+      "epoch": 1.2137023593466425,
+      "grad_norm": 5.5,
+      "learning_rate": 2.2292173402566732e-05,
+      "loss": 0.1099,
+      "step": 10700
+    },
+    {
+      "epoch": 1.21483666061706,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.227486648873899e-05,
+      "loss": 0.1134,
+      "step": 10710
+    },
+    {
+      "epoch": 1.2159709618874772,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.225754690243994e-05,
+      "loss": 0.1123,
+      "step": 10720
+    },
+    {
+      "epoch": 1.2171052631578947,
+      "grad_norm": 3.75,
+      "learning_rate": 2.224021467383948e-05,
+      "loss": 0.1191,
+      "step": 10730
+    },
+    {
+      "epoch": 1.2182395644283122,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.222286983312953e-05,
+      "loss": 0.1143,
+      "step": 10740
+    },
+    {
+      "epoch": 1.2193738656987296,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.2205512410523973e-05,
+      "loss": 0.1087,
+      "step": 10750
+    },
+    {
+      "epoch": 1.220508166969147,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.2188142436258617e-05,
+      "loss": 0.114,
+      "step": 10760
+    },
+    {
+      "epoch": 1.2216424682395643,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.217075994059113e-05,
+      "loss": 0.109,
+      "step": 10770
+    },
+    {
+      "epoch": 1.2227767695099818,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.2153364953800985e-05,
+      "loss": 0.1086,
+      "step": 10780
+    },
+    {
+      "epoch": 1.2239110707803993,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.2135957506189432e-05,
+      "loss": 0.1137,
+      "step": 10790
+    },
+    {
+      "epoch": 1.2250453720508168,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.2118537628079412e-05,
+      "loss": 0.1193,
+      "step": 10800
+    },
+    {
+      "epoch": 1.226179673321234,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.210110534981552e-05,
+      "loss": 0.115,
+      "step": 10810
+    },
+    {
+      "epoch": 1.2273139745916515,
+      "grad_norm": 3.0,
+      "learning_rate": 2.2083660701763968e-05,
+      "loss": 0.1109,
+      "step": 10820
+    },
+    {
+      "epoch": 1.228448275862069,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.206620371431249e-05,
+      "loss": 0.1129,
+      "step": 10830
+    },
+    {
+      "epoch": 1.2295825771324864,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.2048734417870332e-05,
+      "loss": 0.1142,
+      "step": 10840
+    },
+    {
+      "epoch": 1.230716878402904,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.203125284286818e-05,
+      "loss": 0.1115,
+      "step": 10850
+    },
+    {
+      "epoch": 1.2318511796733211,
+      "grad_norm": 2.75,
+      "learning_rate": 2.20137590197581e-05,
+      "loss": 0.1111,
+      "step": 10860
+    },
+    {
+      "epoch": 1.2329854809437386,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.1996252979013505e-05,
+      "loss": 0.1162,
+      "step": 10870
+    },
+    {
+      "epoch": 1.234119782214156,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.197873475112908e-05,
+      "loss": 0.1047,
+      "step": 10880
+    },
+    {
+      "epoch": 1.2352540834845736,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.1961204366620745e-05,
+      "loss": 0.1046,
+      "step": 10890
+    },
+    {
+      "epoch": 1.236388384754991,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.1943661856025606e-05,
+      "loss": 0.1185,
+      "step": 10900
+    },
+    {
+      "epoch": 1.2375226860254083,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.192610724990187e-05,
+      "loss": 0.116,
+      "step": 10910
+    },
+    {
+      "epoch": 1.2386569872958257,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.190854057882883e-05,
+      "loss": 0.1153,
+      "step": 10920
+    },
+    {
+      "epoch": 1.2397912885662432,
+      "grad_norm": 3.25,
+      "learning_rate": 2.189096187340679e-05,
+      "loss": 0.109,
+      "step": 10930
+    },
+    {
+      "epoch": 1.2409255898366607,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.187337116425702e-05,
+      "loss": 0.1162,
+      "step": 10940
+    },
+    {
+      "epoch": 1.2420598911070782,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.18557684820217e-05,
+      "loss": 0.1093,
+      "step": 10950
+    },
+    {
+      "epoch": 1.2431941923774954,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.1838153857363858e-05,
+      "loss": 0.1195,
+      "step": 10960
+    },
+    {
+      "epoch": 1.2443284936479129,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.182052732096734e-05,
+      "loss": 0.111,
+      "step": 10970
+    },
+    {
+      "epoch": 1.2454627949183303,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.1802888903536728e-05,
+      "loss": 0.1067,
+      "step": 10980
+    },
+    {
+      "epoch": 1.2465970961887478,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.178523863579731e-05,
+      "loss": 0.117,
+      "step": 10990
+    },
+    {
+      "epoch": 1.247731397459165,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.1767576548495015e-05,
+      "loss": 0.1082,
+      "step": 11000
+    },
+    {
+      "epoch": 1.2488656987295825,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.174990267239635e-05,
+      "loss": 0.1089,
+      "step": 11010
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.1732217038288374e-05,
+      "loss": 0.1119,
+      "step": 11020
+    },
+    {
+      "epoch": 1.2511343012704175,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.1714519676978617e-05,
+      "loss": 0.1182,
+      "step": 11030
+    },
+    {
+      "epoch": 1.252268602540835,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.169681061929504e-05,
+      "loss": 0.1141,
+      "step": 11040
+    },
+    {
+      "epoch": 1.2534029038112522,
+      "grad_norm": 3.125,
+      "learning_rate": 2.167908989608598e-05,
+      "loss": 0.1219,
+      "step": 11050
+    },
+    {
+      "epoch": 1.2545372050816697,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.1661357538220098e-05,
+      "loss": 0.1128,
+      "step": 11060
+    },
+    {
+      "epoch": 1.2556715063520871,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.1643613576586308e-05,
+      "loss": 0.1197,
+      "step": 11070
+    },
+    {
+      "epoch": 1.2568058076225046,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.1625858042093754e-05,
+      "loss": 0.12,
+      "step": 11080
+    },
+    {
+      "epoch": 1.2579401088929218,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.160809096567173e-05,
+      "loss": 0.1188,
+      "step": 11090
+    },
+    {
+      "epoch": 1.2590744101633393,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.159031237826964e-05,
+      "loss": 0.1078,
+      "step": 11100
+    },
+    {
+      "epoch": 1.2602087114337568,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.1572522310856932e-05,
+      "loss": 0.1139,
+      "step": 11110
+    },
+    {
+      "epoch": 1.2613430127041743,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.1554720794423064e-05,
+      "loss": 0.1154,
+      "step": 11120
+    },
+    {
+      "epoch": 1.2624773139745917,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.153690785997743e-05,
+      "loss": 0.1126,
+      "step": 11130
+    },
+    {
+      "epoch": 1.2636116152450092,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1519083538549307e-05,
+      "loss": 0.1219,
+      "step": 11140
+    },
+    {
+      "epoch": 1.2647459165154264,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.1501247861187822e-05,
+      "loss": 0.1128,
+      "step": 11150
+    },
+    {
+      "epoch": 1.265880217785844,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.1483400858961877e-05,
+      "loss": 0.1108,
+      "step": 11160
+    },
+    {
+      "epoch": 1.2670145190562614,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.1465542562960087e-05,
+      "loss": 0.1058,
+      "step": 11170
+    },
+    {
+      "epoch": 1.2681488203266786,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.1447673004290775e-05,
+      "loss": 0.111,
+      "step": 11180
+    },
+    {
+      "epoch": 1.269283121597096,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.142979221408185e-05,
+      "loss": 0.1167,
+      "step": 11190
+    },
+    {
+      "epoch": 1.2704174228675136,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.14119002234808e-05,
+      "loss": 0.1087,
+      "step": 11200
+    },
+    {
+      "epoch": 1.271551724137931,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.139399706365462e-05,
+      "loss": 0.1118,
+      "step": 11210
+    },
+    {
+      "epoch": 1.2726860254083485,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.1376082765789766e-05,
+      "loss": 0.1213,
+      "step": 11220
+    },
+    {
+      "epoch": 1.273820326678766,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.135815736109209e-05,
+      "loss": 0.1106,
+      "step": 11230
+    },
+    {
+      "epoch": 1.2749546279491832,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.134022088078679e-05,
+      "loss": 0.1113,
+      "step": 11240
+    },
+    {
+      "epoch": 1.2760889292196007,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.132227335611837e-05,
+      "loss": 0.1143,
+      "step": 11250
+    },
+    {
+      "epoch": 1.2772232304900182,
+      "grad_norm": 4.90625,
+      "learning_rate": 2.1304314818350562e-05,
+      "loss": 0.111,
+      "step": 11260
+    },
+    {
+      "epoch": 1.2783575317604357,
+      "grad_norm": 3.625,
+      "learning_rate": 2.1286345298766277e-05,
+      "loss": 0.1042,
+      "step": 11270
+    },
+    {
+      "epoch": 1.279491833030853,
+      "grad_norm": 3.5,
+      "learning_rate": 2.126836482866757e-05,
+      "loss": 0.114,
+      "step": 11280
+    },
+    {
+      "epoch": 1.2806261343012704,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.1250373439375565e-05,
+      "loss": 0.1128,
+      "step": 11290
+    },
+    {
+      "epoch": 1.2817604355716878,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.12323711622304e-05,
+      "loss": 0.1148,
+      "step": 11300
+    },
+    {
+      "epoch": 1.2828947368421053,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.1214358028591194e-05,
+      "loss": 0.1113,
+      "step": 11310
+    },
+    {
+      "epoch": 1.2840290381125228,
+      "grad_norm": 3.25,
+      "learning_rate": 2.1196334069835958e-05,
+      "loss": 0.1047,
+      "step": 11320
+    },
+    {
+      "epoch": 1.28516333938294,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.1178299317361586e-05,
+      "loss": 0.1013,
+      "step": 11330
+    },
+    {
+      "epoch": 1.2862976406533575,
+      "grad_norm": 3.0,
+      "learning_rate": 2.1160253802583744e-05,
+      "loss": 0.1156,
+      "step": 11340
+    },
+    {
+      "epoch": 1.287431941923775,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.1142197556936862e-05,
+      "loss": 0.1106,
+      "step": 11350
+    },
+    {
+      "epoch": 1.2885662431941924,
+      "grad_norm": 4.3125,
+      "learning_rate": 2.112413061187407e-05,
+      "loss": 0.1153,
+      "step": 11360
+    },
+    {
+      "epoch": 1.2897005444646097,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.110605299886712e-05,
+      "loss": 0.1187,
+      "step": 11370
+    },
+    {
+      "epoch": 1.2908348457350272,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.1087964749406356e-05,
+      "loss": 0.1138,
+      "step": 11380
+    },
+    {
+      "epoch": 1.2919691470054446,
+      "grad_norm": 3.375,
+      "learning_rate": 2.1069865895000647e-05,
+      "loss": 0.1094,
+      "step": 11390
+    },
+    {
+      "epoch": 1.293103448275862,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.1051756467177333e-05,
+      "loss": 0.1122,
+      "step": 11400
+    },
+    {
+      "epoch": 1.2942377495462796,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.1033636497482182e-05,
+      "loss": 0.1146,
+      "step": 11410
+    },
+    {
+      "epoch": 1.295372050816697,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.1015506017479313e-05,
+      "loss": 0.1129,
+      "step": 11420
+    },
+    {
+      "epoch": 1.2965063520871143,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.0997365058751165e-05,
+      "loss": 0.106,
+      "step": 11430
+    },
+    {
+      "epoch": 1.2976406533575318,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.097921365289842e-05,
+      "loss": 0.1043,
+      "step": 11440
+    },
+    {
+      "epoch": 1.2987749546279492,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.096105183153997e-05,
+      "loss": 0.1038,
+      "step": 11450
+    },
+    {
+      "epoch": 1.2999092558983665,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.094287962631284e-05,
+      "loss": 0.1122,
+      "step": 11460
+    },
+    {
+      "epoch": 1.301043557168784,
+      "grad_norm": 3.0,
+      "learning_rate": 2.0924697068872145e-05,
+      "loss": 0.1196,
+      "step": 11470
+    },
+    {
+      "epoch": 1.3021778584392014,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.0906504190891033e-05,
+      "loss": 0.1119,
+      "step": 11480
+    },
+    {
+      "epoch": 1.303312159709619,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.088830102406064e-05,
+      "loss": 0.1202,
+      "step": 11490
+    },
+    {
+      "epoch": 1.3044464609800364,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.0870087600090012e-05,
+      "loss": 0.1222,
+      "step": 11500
+    },
+    {
+      "epoch": 1.3055807622504538,
+      "grad_norm": 2.5,
+      "learning_rate": 2.0851863950706068e-05,
+      "loss": 0.1144,
+      "step": 11510
+    },
+    {
+      "epoch": 1.306715063520871,
+      "grad_norm": 3.25,
+      "learning_rate": 2.0833630107653536e-05,
+      "loss": 0.115,
+      "step": 11520
+    },
+    {
+      "epoch": 1.3078493647912885,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.0815386102694904e-05,
+      "loss": 0.1162,
+      "step": 11530
+    },
+    {
+      "epoch": 1.308983666061706,
+      "grad_norm": 3.375,
+      "learning_rate": 2.079713196761036e-05,
+      "loss": 0.1115,
+      "step": 11540
+    },
+    {
+      "epoch": 1.3101179673321235,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.077886773419774e-05,
+      "loss": 0.1182,
+      "step": 11550
+    },
+    {
+      "epoch": 1.3112522686025407,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.076059343427247e-05,
+      "loss": 0.108,
+      "step": 11560
+    },
+    {
+      "epoch": 1.3123865698729582,
+      "grad_norm": 3.25,
+      "learning_rate": 2.074230909966751e-05,
+      "loss": 0.1109,
+      "step": 11570
+    },
+    {
+      "epoch": 1.3135208711433757,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.0724014762233306e-05,
+      "loss": 0.1094,
+      "step": 11580
+    },
+    {
+      "epoch": 1.3146551724137931,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.0705710453837717e-05,
+      "loss": 0.1003,
+      "step": 11590
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.0687396206365974e-05,
+      "loss": 0.1131,
+      "step": 11600
+    },
+    {
+      "epoch": 1.3169237749546279,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.0669072051720637e-05,
+      "loss": 0.1219,
+      "step": 11610
+    },
+    {
+      "epoch": 1.3180580762250453,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.06507380218215e-05,
+      "loss": 0.1085,
+      "step": 11620
+    },
+    {
+      "epoch": 1.3191923774954628,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.0632394148605577e-05,
+      "loss": 0.1036,
+      "step": 11630
+    },
+    {
+      "epoch": 1.3203266787658803,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.0614040464027023e-05,
+      "loss": 0.1012,
+      "step": 11640
+    },
+    {
+      "epoch": 1.3214609800362975,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.0595677000057077e-05,
+      "loss": 0.1046,
+      "step": 11650
+    },
+    {
+      "epoch": 1.322595281306715,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.0577303788684027e-05,
+      "loss": 0.1106,
+      "step": 11660
+    },
+    {
+      "epoch": 1.3237295825771325,
+      "grad_norm": 2.875,
+      "learning_rate": 2.0558920861913126e-05,
+      "loss": 0.111,
+      "step": 11670
+    },
+    {
+      "epoch": 1.32486388384755,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.054052825176656e-05,
+      "loss": 0.115,
+      "step": 11680
+    },
+    {
+      "epoch": 1.3259981851179674,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.0522125990283393e-05,
+      "loss": 0.1128,
+      "step": 11690
+    },
+    {
+      "epoch": 1.3271324863883849,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.0503714109519473e-05,
+      "loss": 0.1117,
+      "step": 11700
+    },
+    {
+      "epoch": 1.3282667876588021,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.048529264154742e-05,
+      "loss": 0.1119,
+      "step": 11710
+    },
+    {
+      "epoch": 1.3294010889292196,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.046686161845657e-05,
+      "loss": 0.1013,
+      "step": 11720
+    },
+    {
+      "epoch": 1.330535390199637,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.0448421072352873e-05,
+      "loss": 0.1034,
+      "step": 11730
+    },
+    {
+      "epoch": 1.3316696914700543,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.0429971035358885e-05,
+      "loss": 0.1193,
+      "step": 11740
+    },
+    {
+      "epoch": 1.3328039927404718,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.0411511539613702e-05,
+      "loss": 0.1195,
+      "step": 11750
+    },
+    {
+      "epoch": 1.3339382940108893,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.039304261727288e-05,
+      "loss": 0.1136,
+      "step": 11760
+    },
+    {
+      "epoch": 1.3350725952813067,
+      "grad_norm": 9.625,
+      "learning_rate": 2.037456430050841e-05,
+      "loss": 0.1127,
+      "step": 11770
+    },
+    {
+      "epoch": 1.3362068965517242,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.0356076621508627e-05,
+      "loss": 0.1055,
+      "step": 11780
+    },
+    {
+      "epoch": 1.3373411978221417,
+      "grad_norm": 3.25,
+      "learning_rate": 2.0337579612478204e-05,
+      "loss": 0.1042,
+      "step": 11790
+    },
+    {
+      "epoch": 1.338475499092559,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.0319073305638035e-05,
+      "loss": 0.1068,
+      "step": 11800
+    },
+    {
+      "epoch": 1.3396098003629764,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.0300557733225237e-05,
+      "loss": 0.1086,
+      "step": 11810
+    },
+    {
+      "epoch": 1.3407441016333939,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.028203292749306e-05,
+      "loss": 0.116,
+      "step": 11820
+    },
+    {
+      "epoch": 1.3418784029038113,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.026349892071082e-05,
+      "loss": 0.1021,
+      "step": 11830
+    },
+    {
+      "epoch": 1.3430127041742286,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.024495574516389e-05,
+      "loss": 0.1089,
+      "step": 11840
+    },
+    {
+      "epoch": 1.344147005444646,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.022640343315359e-05,
+      "loss": 0.1168,
+      "step": 11850
+    },
+    {
+      "epoch": 1.3452813067150635,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.020784201699716e-05,
+      "loss": 0.1244,
+      "step": 11860
+    },
+    {
+      "epoch": 1.346415607985481,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.0189271529027716e-05,
+      "loss": 0.1009,
+      "step": 11870
+    },
+    {
+      "epoch": 1.3475499092558985,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.017069200159416e-05,
+      "loss": 0.1107,
+      "step": 11880
+    },
+    {
+      "epoch": 1.3486842105263157,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.0152103467061148e-05,
+      "loss": 0.1188,
+      "step": 11890
+    },
+    {
+      "epoch": 1.3498185117967332,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.0133505957809014e-05,
+      "loss": 0.1125,
+      "step": 11900
+    },
+    {
+      "epoch": 1.3509528130671506,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.011489950623374e-05,
+      "loss": 0.1104,
+      "step": 11910
+    },
+    {
+      "epoch": 1.3520871143375681,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.009628414474688e-05,
+      "loss": 0.1113,
+      "step": 11920
+    },
+    {
+      "epoch": 1.3532214156079854,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.007765990577551e-05,
+      "loss": 0.1134,
+      "step": 11930
+    },
+    {
+      "epoch": 1.3543557168784028,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.005902682176216e-05,
+      "loss": 0.1032,
+      "step": 11940
+    },
+    {
+      "epoch": 1.3554900181488203,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.0040384925164782e-05,
+      "loss": 0.0965,
+      "step": 11950
+    },
+    {
+      "epoch": 1.3566243194192378,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.0021734248456676e-05,
+      "loss": 0.1084,
+      "step": 11960
+    },
+    {
+      "epoch": 1.3577586206896552,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.0003074824126426e-05,
+      "loss": 0.1147,
+      "step": 11970
+    },
+    {
+      "epoch": 1.3588929219600727,
+      "grad_norm": 3.375,
+      "learning_rate": 1.9984406684677864e-05,
+      "loss": 0.1209,
+      "step": 11980
+    },
+    {
+      "epoch": 1.36002722323049,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.9965729862630002e-05,
+      "loss": 0.1147,
+      "step": 11990
+    },
+    {
+      "epoch": 1.3611615245009074,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.994704439051698e-05,
+      "loss": 0.11,
+      "step": 12000
+    },
+    {
+      "epoch": 1.362295825771325,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.9928350300887987e-05,
+      "loss": 0.1083,
+      "step": 12010
+    },
+    {
+      "epoch": 1.3634301270417422,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.9909647626307256e-05,
+      "loss": 0.1135,
+      "step": 12020
+    },
+    {
+      "epoch": 1.3645644283121596,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.9890936399353942e-05,
+      "loss": 0.1055,
+      "step": 12030
+    },
+    {
+      "epoch": 1.365698729582577,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.987221665262212e-05,
+      "loss": 0.1147,
+      "step": 12040
+    },
+    {
+      "epoch": 1.3668330308529946,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.9853488418720694e-05,
+      "loss": 0.1184,
+      "step": 12050
+    },
+    {
+      "epoch": 1.367967332123412,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.9834751730273356e-05,
+      "loss": 0.1097,
+      "step": 12060
+    },
+    {
+      "epoch": 1.3691016333938295,
+      "grad_norm": 3.625,
+      "learning_rate": 1.9816006619918536e-05,
+      "loss": 0.1086,
+      "step": 12070
+    },
+    {
+      "epoch": 1.3702359346642468,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.979725312030931e-05,
+      "loss": 0.111,
+      "step": 12080
+    },
+    {
+      "epoch": 1.3713702359346642,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.977849126411339e-05,
+      "loss": 0.1112,
+      "step": 12090
+    },
+    {
+      "epoch": 1.3725045372050817,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.9759721084013035e-05,
+      "loss": 0.1131,
+      "step": 12100
+    },
+    {
+      "epoch": 1.3736388384754992,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.9740942612705003e-05,
+      "loss": 0.1119,
+      "step": 12110
+    },
+    {
+      "epoch": 1.3747731397459164,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.9722155882900506e-05,
+      "loss": 0.1182,
+      "step": 12120
+    },
+    {
+      "epoch": 1.3759074410163339,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.9703360927325124e-05,
+      "loss": 0.1113,
+      "step": 12130
+    },
+    {
+      "epoch": 1.3770417422867514,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.968455777871878e-05,
+      "loss": 0.1015,
+      "step": 12140
+    },
+    {
+      "epoch": 1.3781760435571688,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.9665746469835667e-05,
+      "loss": 0.1085,
+      "step": 12150
+    },
+    {
+      "epoch": 1.3793103448275863,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.9646927033444185e-05,
+      "loss": 0.1103,
+      "step": 12160
+    },
+    {
+      "epoch": 1.3804446460980035,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.9628099502326896e-05,
+      "loss": 0.1202,
+      "step": 12170
+    },
+    {
+      "epoch": 1.381578947368421,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.9609263909280468e-05,
+      "loss": 0.1222,
+      "step": 12180
+    },
+    {
+      "epoch": 1.3827132486388385,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.9590420287115603e-05,
+      "loss": 0.1132,
+      "step": 12190
+    },
+    {
+      "epoch": 1.383847549909256,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.9571568668657007e-05,
+      "loss": 0.1024,
+      "step": 12200
+    },
+    {
+      "epoch": 1.3849818511796732,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.9552709086743283e-05,
+      "loss": 0.1102,
+      "step": 12210
+    },
+    {
+      "epoch": 1.3861161524500907,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.953384157422694e-05,
+      "loss": 0.1162,
+      "step": 12220
+    },
+    {
+      "epoch": 1.3872504537205081,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.951496616397428e-05,
+      "loss": 0.1,
+      "step": 12230
+    },
+    {
+      "epoch": 1.3883847549909256,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.949608288886537e-05,
+      "loss": 0.1102,
+      "step": 12240
+    },
+    {
+      "epoch": 1.389519056261343,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.9477191781793984e-05,
+      "loss": 0.1115,
+      "step": 12250
+    },
+    {
+      "epoch": 1.3906533575317606,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.9458292875667524e-05,
+      "loss": 0.1105,
+      "step": 12260
+    },
+    {
+      "epoch": 1.3917876588021778,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.9439386203406987e-05,
+      "loss": 0.1124,
+      "step": 12270
+    },
+    {
+      "epoch": 1.3929219600725953,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.94204717979469e-05,
+      "loss": 0.1108,
+      "step": 12280
+    },
+    {
+      "epoch": 1.3940562613430127,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.940154969223525e-05,
+      "loss": 0.1052,
+      "step": 12290
+    },
+    {
+      "epoch": 1.39519056261343,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.9382619919233455e-05,
+      "loss": 0.1062,
+      "step": 12300
+    },
+    {
+      "epoch": 1.3963248638838475,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.936368251191627e-05,
+      "loss": 0.1024,
+      "step": 12310
+    },
+    {
+      "epoch": 1.397459165154265,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.934473750327176e-05,
+      "loss": 0.1051,
+      "step": 12320
+    },
+    {
+      "epoch": 1.3985934664246824,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.9325784926301235e-05,
+      "loss": 0.113,
+      "step": 12330
+    },
+    {
+      "epoch": 1.3997277676950999,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.9306824814019176e-05,
+      "loss": 0.1083,
+      "step": 12340
+    },
+    {
+      "epoch": 1.4008620689655173,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.9287857199453197e-05,
+      "loss": 0.1097,
+      "step": 12350
+    },
+    {
+      "epoch": 1.4019963702359346,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.9268882115643977e-05,
+      "loss": 0.1192,
+      "step": 12360
+    },
+    {
+      "epoch": 1.403130671506352,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.9249899595645214e-05,
+      "loss": 0.1065,
+      "step": 12370
+    },
+    {
+      "epoch": 1.4042649727767695,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.9230909672523557e-05,
+      "loss": 0.0986,
+      "step": 12380
+    },
+    {
+      "epoch": 1.405399274047187,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.9211912379358546e-05,
+      "loss": 0.1181,
+      "step": 12390
+    },
+    {
+      "epoch": 1.4065335753176043,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.9192907749242563e-05,
+      "loss": 0.1,
+      "step": 12400
+    },
+    {
+      "epoch": 1.4076678765880217,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.917389581528076e-05,
+      "loss": 0.1127,
+      "step": 12410
+    },
+    {
+      "epoch": 1.4088021778584392,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.915487661059103e-05,
+      "loss": 0.1057,
+      "step": 12420
+    },
+    {
+      "epoch": 1.4099364791288567,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.913585016830393e-05,
+      "loss": 0.1043,
+      "step": 12430
+    },
+    {
+      "epoch": 1.4110707803992741,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.9116816521562604e-05,
+      "loss": 0.1005,
+      "step": 12440
+    },
+    {
+      "epoch": 1.4122050816696916,
+      "grad_norm": 3.5,
+      "learning_rate": 1.9097775703522768e-05,
+      "loss": 0.1086,
+      "step": 12450
+    },
+    {
+      "epoch": 1.4133393829401089,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.9078727747352614e-05,
+      "loss": 0.1188,
+      "step": 12460
+    },
+    {
+      "epoch": 1.4144736842105263,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.9059672686232775e-05,
+      "loss": 0.11,
+      "step": 12470
+    },
+    {
+      "epoch": 1.4156079854809438,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.9040610553356273e-05,
+      "loss": 0.1142,
+      "step": 12480
+    },
+    {
+      "epoch": 1.416742286751361,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.902154138192842e-05,
+      "loss": 0.1234,
+      "step": 12490
+    },
+    {
+      "epoch": 1.4178765880217785,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.9002465205166817e-05,
+      "loss": 0.1041,
+      "step": 12500
+    },
+    {
+      "epoch": 1.419010889292196,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.8983382056301253e-05,
+      "loss": 0.1161,
+      "step": 12510
+    },
+    {
+      "epoch": 1.4201451905626135,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.8964291968573664e-05,
+      "loss": 0.1273,
+      "step": 12520
+    },
+    {
+      "epoch": 1.421279491833031,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.8945194975238082e-05,
+      "loss": 0.1168,
+      "step": 12530
+    },
+    {
+      "epoch": 1.4224137931034484,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.8926091109560548e-05,
+      "loss": 0.1124,
+      "step": 12540
+    },
+    {
+      "epoch": 1.4235480943738656,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.890698040481909e-05,
+      "loss": 0.1116,
+      "step": 12550
+    },
+    {
+      "epoch": 1.4246823956442831,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.8887862894303656e-05,
+      "loss": 0.1092,
+      "step": 12560
+    },
+    {
+      "epoch": 1.4258166969147006,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.8868738611316033e-05,
+      "loss": 0.1138,
+      "step": 12570
+    },
+    {
+      "epoch": 1.4269509981851178,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.8849607589169812e-05,
+      "loss": 0.1146,
+      "step": 12580
+    },
+    {
+      "epoch": 1.4280852994555353,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.8830469861190328e-05,
+      "loss": 0.1121,
+      "step": 12590
+    },
+    {
+      "epoch": 1.4292196007259528,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.8811325460714585e-05,
+      "loss": 0.1049,
+      "step": 12600
+    },
+    {
+      "epoch": 1.4303539019963702,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.879217442109122e-05,
+      "loss": 0.1076,
+      "step": 12610
+    },
+    {
+      "epoch": 1.4314882032667877,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.8773016775680435e-05,
+      "loss": 0.1217,
+      "step": 12620
+    },
+    {
+      "epoch": 1.4326225045372052,
+      "grad_norm": 3.625,
+      "learning_rate": 1.8753852557853943e-05,
+      "loss": 0.1099,
+      "step": 12630
+    },
+    {
+      "epoch": 1.4337568058076224,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.873468180099489e-05,
+      "loss": 0.1102,
+      "step": 12640
+    },
+    {
+      "epoch": 1.43489110707804,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.8715504538497823e-05,
+      "loss": 0.1166,
+      "step": 12650
+    },
+    {
+      "epoch": 1.4360254083484574,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.8696320803768626e-05,
+      "loss": 0.114,
+      "step": 12660
+    },
+    {
+      "epoch": 1.4371597096188748,
+      "grad_norm": 2.875,
+      "learning_rate": 1.8677130630224446e-05,
+      "loss": 0.1123,
+      "step": 12670
+    },
+    {
+      "epoch": 1.438294010889292,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.865793405129366e-05,
+      "loss": 0.1092,
+      "step": 12680
+    },
+    {
+      "epoch": 1.4394283121597096,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.8638731100415786e-05,
+      "loss": 0.1028,
+      "step": 12690
+    },
+    {
+      "epoch": 1.440562613430127,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.8619521811041458e-05,
+      "loss": 0.1122,
+      "step": 12700
+    },
+    {
+      "epoch": 1.4416969147005445,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.8600306216632347e-05,
+      "loss": 0.1125,
+      "step": 12710
+    },
+    {
+      "epoch": 1.442831215970962,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.8581084350661097e-05,
+      "loss": 0.1038,
+      "step": 12720
+    },
+    {
+      "epoch": 1.4439655172413794,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.8561856246611283e-05,
+      "loss": 0.1114,
+      "step": 12730
+    },
+    {
+      "epoch": 1.4450998185117967,
+      "grad_norm": 4.0,
+      "learning_rate": 1.854262193797735e-05,
+      "loss": 0.1102,
+      "step": 12740
+    },
+    {
+      "epoch": 1.4462341197822142,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.8523381458264548e-05,
+      "loss": 0.1135,
+      "step": 12750
+    },
+    {
+      "epoch": 1.4473684210526316,
+      "grad_norm": 2.875,
+      "learning_rate": 1.8504134840988887e-05,
+      "loss": 0.1095,
+      "step": 12760
+    },
+    {
+      "epoch": 1.4485027223230489,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.8484882119677042e-05,
+      "loss": 0.1062,
+      "step": 12770
+    },
+    {
+      "epoch": 1.4496370235934664,
+      "grad_norm": 3.0,
+      "learning_rate": 1.8465623327866346e-05,
+      "loss": 0.1137,
+      "step": 12780
+    },
+    {
+      "epoch": 1.4507713248638838,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.8446358499104703e-05,
+      "loss": 0.1077,
+      "step": 12790
+    },
+    {
+      "epoch": 1.4519056261343013,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8427087666950517e-05,
+      "loss": 0.106,
+      "step": 12800
+    },
+    {
+      "epoch": 1.4530399274047188,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.840781086497267e-05,
+      "loss": 0.1176,
+      "step": 12810
+    },
+    {
+      "epoch": 1.4541742286751362,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.838852812675043e-05,
+      "loss": 0.1109,
+      "step": 12820
+    },
+    {
+      "epoch": 1.4553085299455535,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.8369239485873406e-05,
+      "loss": 0.1048,
+      "step": 12830
+    },
+    {
+      "epoch": 1.456442831215971,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.83499449759415e-05,
+      "loss": 0.1205,
+      "step": 12840
+    },
+    {
+      "epoch": 1.4575771324863884,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.833064463056483e-05,
+      "loss": 0.1088,
+      "step": 12850
+    },
+    {
+      "epoch": 1.4587114337568057,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.8311338483363674e-05,
+      "loss": 0.117,
+      "step": 12860
+    },
+    {
+      "epoch": 1.4598457350272231,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.829202656796842e-05,
+      "loss": 0.1109,
+      "step": 12870
+    },
+    {
+      "epoch": 1.4609800362976406,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.8272708918019513e-05,
+      "loss": 0.109,
+      "step": 12880
+    },
+    {
+      "epoch": 1.462114337568058,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.8253385567167382e-05,
+      "loss": 0.1068,
+      "step": 12890
+    },
+    {
+      "epoch": 1.4632486388384756,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.8234056549072376e-05,
+      "loss": 0.1185,
+      "step": 12900
+    },
+    {
+      "epoch": 1.464382940108893,
+      "grad_norm": 4.9375,
+      "learning_rate": 1.821472189740473e-05,
+      "loss": 0.1217,
+      "step": 12910
+    },
+    {
+      "epoch": 1.4655172413793103,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.8195381645844486e-05,
+      "loss": 0.1164,
+      "step": 12920
+    },
+    {
+      "epoch": 1.4666515426497277,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.8176035828081435e-05,
+      "loss": 0.099,
+      "step": 12930
+    },
+    {
+      "epoch": 1.4677858439201452,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.815668447781509e-05,
+      "loss": 0.1227,
+      "step": 12940
+    },
+    {
+      "epoch": 1.4689201451905627,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.8137327628754563e-05,
+      "loss": 0.1114,
+      "step": 12950
+    },
+    {
+      "epoch": 1.47005444646098,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.811796531461858e-05,
+      "loss": 0.109,
+      "step": 12960
+    },
+    {
+      "epoch": 1.4711887477313974,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.8098597569135357e-05,
+      "loss": 0.1063,
+      "step": 12970
+    },
+    {
+      "epoch": 1.4723230490018149,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.807922442604259e-05,
+      "loss": 0.1092,
+      "step": 12980
+    },
+    {
+      "epoch": 1.4734573502722323,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.8059845919087373e-05,
+      "loss": 0.1126,
+      "step": 12990
+    },
+    {
+      "epoch": 1.4745916515426498,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.804046208202615e-05,
+      "loss": 0.1073,
+      "step": 13000
+    },
+    {
+      "epoch": 1.4757259528130673,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.8021072948624635e-05,
+      "loss": 0.1122,
+      "step": 13010
+    },
+    {
+      "epoch": 1.4768602540834845,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.8001678552657777e-05,
+      "loss": 0.1064,
+      "step": 13020
+    },
+    {
+      "epoch": 1.477994555353902,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.798227892790969e-05,
+      "loss": 0.111,
+      "step": 13030
+    },
+    {
+      "epoch": 1.4791288566243195,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.7962874108173605e-05,
+      "loss": 0.1181,
+      "step": 13040
+    },
+    {
+      "epoch": 1.4802631578947367,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.794346412725178e-05,
+      "loss": 0.1129,
+      "step": 13050
+    },
+    {
+      "epoch": 1.4813974591651542,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.7924049018955493e-05,
+      "loss": 0.1141,
+      "step": 13060
+    },
+    {
+      "epoch": 1.4825317604355717,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.790462881710493e-05,
+      "loss": 0.1165,
+      "step": 13070
+    },
+    {
+      "epoch": 1.4836660617059891,
+      "grad_norm": 4.125,
+      "learning_rate": 1.7885203555529165e-05,
+      "loss": 0.1122,
+      "step": 13080
+    },
+    {
+      "epoch": 1.4848003629764066,
+      "grad_norm": 3.125,
+      "learning_rate": 1.7865773268066077e-05,
+      "loss": 0.1033,
+      "step": 13090
+    },
+    {
+      "epoch": 1.485934664246824,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.7846337988562297e-05,
+      "loss": 0.111,
+      "step": 13100
+    },
+    {
+      "epoch": 1.4870689655172413,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.7826897750873165e-05,
+      "loss": 0.1159,
+      "step": 13110
+    },
+    {
+      "epoch": 1.4882032667876588,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.7807452588862645e-05,
+      "loss": 0.1098,
+      "step": 13120
+    },
+    {
+      "epoch": 1.4893375680580763,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.7788002536403287e-05,
+      "loss": 0.1136,
+      "step": 13130
+    },
+    {
+      "epoch": 1.4904718693284935,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.776854762737616e-05,
+      "loss": 0.1061,
+      "step": 13140
+    },
+    {
+      "epoch": 1.491606170598911,
+      "grad_norm": 3.0,
+      "learning_rate": 1.774908789567078e-05,
+      "loss": 0.1106,
+      "step": 13150
+    },
+    {
+      "epoch": 1.4927404718693285,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.7729623375185082e-05,
+      "loss": 0.1119,
+      "step": 13160
+    },
+    {
+      "epoch": 1.493874773139746,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.7710154099825336e-05,
+      "loss": 0.1161,
+      "step": 13170
+    },
+    {
+      "epoch": 1.4950090744101634,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.769068010350608e-05,
+      "loss": 0.1094,
+      "step": 13180
+    },
+    {
+      "epoch": 1.4961433756805809,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.7671201420150108e-05,
+      "loss": 0.11,
+      "step": 13190
+    },
+    {
+      "epoch": 1.4972776769509981,
+      "grad_norm": 2.375,
+      "learning_rate": 1.7651718083688348e-05,
+      "loss": 0.096,
+      "step": 13200
+    },
+    {
+      "epoch": 1.4984119782214156,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.7632230128059844e-05,
+      "loss": 0.1015,
+      "step": 13210
+    },
+    {
+      "epoch": 1.499546279491833,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.7612737587211696e-05,
+      "loss": 0.1099,
+      "step": 13220
+    },
+    {
+      "epoch": 1.5006805807622503,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.7593240495098976e-05,
+      "loss": 0.107,
+      "step": 13230
+    },
+    {
+      "epoch": 1.5018148820326678,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.7573738885684692e-05,
+      "loss": 0.0911,
+      "step": 13240
+    },
+    {
+      "epoch": 1.5029491833030852,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.755423279293972e-05,
+      "loss": 0.1108,
+      "step": 13250
+    },
+    {
+      "epoch": 1.5040834845735027,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.7534722250842743e-05,
+      "loss": 0.1073,
+      "step": 13260
+    },
+    {
+      "epoch": 1.5052177858439202,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.751520729338021e-05,
+      "loss": 0.1067,
+      "step": 13270
+    },
+    {
+      "epoch": 1.5063520871143377,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.7495687954546233e-05,
+      "loss": 0.1121,
+      "step": 13280
+    },
+    {
+      "epoch": 1.5074863883847551,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.7476164268342574e-05,
+      "loss": 0.1198,
+      "step": 13290
+    },
+    {
+      "epoch": 1.5086206896551724,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.745663626877857e-05,
+      "loss": 0.1084,
+      "step": 13300
+    },
+    {
+      "epoch": 1.5097549909255898,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.7437103989871065e-05,
+      "loss": 0.1016,
+      "step": 13310
+    },
+    {
+      "epoch": 1.510889292196007,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.741756746564437e-05,
+      "loss": 0.1114,
+      "step": 13320
+    },
+    {
+      "epoch": 1.5120235934664246,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.739802673013016e-05,
+      "loss": 0.1088,
+      "step": 13330
+    },
+    {
+      "epoch": 1.513157894736842,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.737848181736748e-05,
+      "loss": 0.12,
+      "step": 13340
+    },
+    {
+      "epoch": 1.5142921960072595,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.7358932761402638e-05,
+      "loss": 0.1211,
+      "step": 13350
+    },
+    {
+      "epoch": 1.515426497277677,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.733937959628915e-05,
+      "loss": 0.1125,
+      "step": 13360
+    },
+    {
+      "epoch": 1.5165607985480944,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.7319822356087718e-05,
+      "loss": 0.1125,
+      "step": 13370
+    },
+    {
+      "epoch": 1.517695099818512,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.7300261074866106e-05,
+      "loss": 0.1065,
+      "step": 13380
+    },
+    {
+      "epoch": 1.5188294010889292,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.7280695786699138e-05,
+      "loss": 0.1037,
+      "step": 13390
+    },
+    {
+      "epoch": 1.5199637023593466,
+      "grad_norm": 3.375,
+      "learning_rate": 1.7261126525668628e-05,
+      "loss": 0.1117,
+      "step": 13400
+    },
+    {
+      "epoch": 1.521098003629764,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.724155332586328e-05,
+      "loss": 0.114,
+      "step": 13410
+    },
+    {
+      "epoch": 1.5222323049001814,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.7221976221378695e-05,
+      "loss": 0.1159,
+      "step": 13420
+    },
+    {
+      "epoch": 1.5233666061705988,
+      "grad_norm": 3.5,
+      "learning_rate": 1.7202395246317235e-05,
+      "loss": 0.1049,
+      "step": 13430
+    },
+    {
+      "epoch": 1.5245009074410163,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.7182810434788044e-05,
+      "loss": 0.1024,
+      "step": 13440
+    },
+    {
+      "epoch": 1.5256352087114338,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.7163221820906934e-05,
+      "loss": 0.1152,
+      "step": 13450
+    },
+    {
+      "epoch": 1.5267695099818512,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.714362943879633e-05,
+      "loss": 0.1048,
+      "step": 13460
+    },
+    {
+      "epoch": 1.5279038112522687,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.712403332258523e-05,
+      "loss": 0.1104,
+      "step": 13470
+    },
+    {
+      "epoch": 1.5290381125226862,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.710443350640914e-05,
+      "loss": 0.1104,
+      "step": 13480
+    },
+    {
+      "epoch": 1.5301724137931034,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.7084830024409994e-05,
+      "loss": 0.1162,
+      "step": 13490
+    },
+    {
+      "epoch": 1.531306715063521,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.7065222910736148e-05,
+      "loss": 0.1157,
+      "step": 13500
+    },
+    {
+      "epoch": 1.5324410163339381,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.7045612199542244e-05,
+      "loss": 0.1097,
+      "step": 13510
+    },
+    {
+      "epoch": 1.5335753176043556,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.7025997924989216e-05,
+      "loss": 0.1115,
+      "step": 13520
+    },
+    {
+      "epoch": 1.534709618874773,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.7006380121244193e-05,
+      "loss": 0.1038,
+      "step": 13530
+    },
+    {
+      "epoch": 1.5358439201451906,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.6986758822480457e-05,
+      "loss": 0.114,
+      "step": 13540
+    },
+    {
+      "epoch": 1.536978221415608,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.6967134062877372e-05,
+      "loss": 0.1042,
+      "step": 13550
+    },
+    {
+      "epoch": 1.5381125226860255,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.6947505876620344e-05,
+      "loss": 0.1091,
+      "step": 13560
+    },
+    {
+      "epoch": 1.539246823956443,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.692787429790074e-05,
+      "loss": 0.1028,
+      "step": 13570
+    },
+    {
+      "epoch": 1.5403811252268602,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.6908239360915825e-05,
+      "loss": 0.1066,
+      "step": 13580
+    },
+    {
+      "epoch": 1.5415154264972777,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.688860109986873e-05,
+      "loss": 0.117,
+      "step": 13590
+    },
+    {
+      "epoch": 1.542649727767695,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.6868959548968384e-05,
+      "loss": 0.108,
+      "step": 13600
+    },
+    {
+      "epoch": 1.5437840290381124,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.6849314742429416e-05,
+      "loss": 0.1147,
+      "step": 13610
+    },
+    {
+      "epoch": 1.5449183303085299,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.6829666714472155e-05,
+      "loss": 0.1136,
+      "step": 13620
+    },
+    {
+      "epoch": 1.5460526315789473,
+      "grad_norm": 3.375,
+      "learning_rate": 1.6810015499322525e-05,
+      "loss": 0.1154,
+      "step": 13630
+    },
+    {
+      "epoch": 1.5471869328493648,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.679036113121201e-05,
+      "loss": 0.1079,
+      "step": 13640
+    },
+    {
+      "epoch": 1.5483212341197823,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.6770703644377594e-05,
+      "loss": 0.1086,
+      "step": 13650
+    },
+    {
+      "epoch": 1.5494555353901998,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.6751043073061664e-05,
+      "loss": 0.1087,
+      "step": 13660
+    },
+    {
+      "epoch": 1.550589836660617,
+      "grad_norm": 3.75,
+      "learning_rate": 1.6731379451512013e-05,
+      "loss": 0.11,
+      "step": 13670
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.6711712813981733e-05,
+      "loss": 0.1181,
+      "step": 13680
+    },
+    {
+      "epoch": 1.552858439201452,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.669204319472917e-05,
+      "loss": 0.1071,
+      "step": 13690
+    },
+    {
+      "epoch": 1.5539927404718692,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.6672370628017865e-05,
+      "loss": 0.1183,
+      "step": 13700
+    },
+    {
+      "epoch": 1.5551270417422867,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.665269514811649e-05,
+      "loss": 0.1085,
+      "step": 13710
+    },
+    {
+      "epoch": 1.5562613430127041,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.66330167892988e-05,
+      "loss": 0.1081,
+      "step": 13720
+    },
+    {
+      "epoch": 1.5573956442831216,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.661333558584356e-05,
+      "loss": 0.1082,
+      "step": 13730
+    },
+    {
+      "epoch": 1.558529945553539,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.659365157203448e-05,
+      "loss": 0.1237,
+      "step": 13740
+    },
+    {
+      "epoch": 1.5596642468239565,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.6573964782160194e-05,
+      "loss": 0.1061,
+      "step": 13750
+    },
+    {
+      "epoch": 1.560798548094374,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.6554275250514134e-05,
+      "loss": 0.1081,
+      "step": 13760
+    },
+    {
+      "epoch": 1.5619328493647913,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.6534583011394537e-05,
+      "loss": 0.1143,
+      "step": 13770
+    },
+    {
+      "epoch": 1.5630671506352087,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.6514888099104346e-05,
+      "loss": 0.1073,
+      "step": 13780
+    },
+    {
+      "epoch": 1.564201451905626,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.649519054795116e-05,
+      "loss": 0.11,
+      "step": 13790
+    },
+    {
+      "epoch": 1.5653357531760435,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.647549039224718e-05,
+      "loss": 0.101,
+      "step": 13800
+    },
+    {
+      "epoch": 1.566470054446461,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.645578766630913e-05,
+      "loss": 0.118,
+      "step": 13810
+    },
+    {
+      "epoch": 1.5676043557168784,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.643608240445823e-05,
+      "loss": 0.1126,
+      "step": 13820
+    },
+    {
+      "epoch": 1.5687386569872959,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.6416374641020108e-05,
+      "loss": 0.1114,
+      "step": 13830
+    },
+    {
+      "epoch": 1.5698729582577133,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.639666441032475e-05,
+      "loss": 0.1112,
+      "step": 13840
+    },
+    {
+      "epoch": 1.5710072595281308,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.6376951746706433e-05,
+      "loss": 0.1085,
+      "step": 13850
+    },
+    {
+      "epoch": 1.572141560798548,
+      "grad_norm": 2.875,
+      "learning_rate": 1.635723668450369e-05,
+      "loss": 0.1164,
+      "step": 13860
+    },
+    {
+      "epoch": 1.5732758620689655,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.6337519258059212e-05,
+      "loss": 0.1173,
+      "step": 13870
+    },
+    {
+      "epoch": 1.574410163339383,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.6317799501719826e-05,
+      "loss": 0.1058,
+      "step": 13880
+    },
+    {
+      "epoch": 1.5755444646098002,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.6298077449836406e-05,
+      "loss": 0.1036,
+      "step": 13890
+    },
+    {
+      "epoch": 1.5766787658802177,
+      "grad_norm": 4.5,
+      "learning_rate": 1.627835313676382e-05,
+      "loss": 0.1165,
+      "step": 13900
+    },
+    {
+      "epoch": 1.5778130671506352,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.62586265968609e-05,
+      "loss": 0.0998,
+      "step": 13910
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.6238897864490323e-05,
+      "loss": 0.11,
+      "step": 13920
+    },
+    {
+      "epoch": 1.5800816696914701,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.621916697401861e-05,
+      "loss": 0.1113,
+      "step": 13930
+    },
+    {
+      "epoch": 1.5812159709618876,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.6199433959816038e-05,
+      "loss": 0.1094,
+      "step": 13940
+    },
+    {
+      "epoch": 1.582350272232305,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.617969885625657e-05,
+      "loss": 0.1118,
+      "step": 13950
+    },
+    {
+      "epoch": 1.5834845735027223,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.615996169771782e-05,
+      "loss": 0.1112,
+      "step": 13960
+    },
+    {
+      "epoch": 1.5846188747731398,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.6140222518580982e-05,
+      "loss": 0.1116,
+      "step": 13970
+    },
+    {
+      "epoch": 1.585753176043557,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.6120481353230768e-05,
+      "loss": 0.1011,
+      "step": 13980
+    },
+    {
+      "epoch": 1.5868874773139745,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.610073823605534e-05,
+      "loss": 0.1046,
+      "step": 13990
+    },
+    {
+      "epoch": 1.588021778584392,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.6080993201446273e-05,
+      "loss": 0.1075,
+      "step": 14000
+    },
+    {
+      "epoch": 1.5891560798548094,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.606124628379849e-05,
+      "loss": 0.1068,
+      "step": 14010
+    },
+    {
+      "epoch": 1.590290381125227,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.6041497517510165e-05,
+      "loss": 0.1089,
+      "step": 14020
+    },
+    {
+      "epoch": 1.5914246823956444,
+      "grad_norm": 3.25,
+      "learning_rate": 1.602174693698272e-05,
+      "loss": 0.1018,
+      "step": 14030
+    },
+    {
+      "epoch": 1.5925589836660619,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.6001994576620722e-05,
+      "loss": 0.1138,
+      "step": 14040
+    },
+    {
+      "epoch": 1.593693284936479,
+      "grad_norm": 3.5,
+      "learning_rate": 1.5982240470831836e-05,
+      "loss": 0.1128,
+      "step": 14050
+    },
+    {
+      "epoch": 1.5948275862068966,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.596248465402679e-05,
+      "loss": 0.1235,
+      "step": 14060
+    },
+    {
+      "epoch": 1.5959618874773138,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.594272716061927e-05,
+      "loss": 0.1167,
+      "step": 14070
+    },
+    {
+      "epoch": 1.5970961887477313,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5922968025025886e-05,
+      "loss": 0.1112,
+      "step": 14080
+    },
+    {
+      "epoch": 1.5982304900181488,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.5903207281666112e-05,
+      "loss": 0.1025,
+      "step": 14090
+    },
+    {
+      "epoch": 1.5993647912885662,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.588344496496223e-05,
+      "loss": 0.1116,
+      "step": 14100
+    },
+    {
+      "epoch": 1.6004990925589837,
+      "grad_norm": 2.875,
+      "learning_rate": 1.5863681109339248e-05,
+      "loss": 0.1101,
+      "step": 14110
+    },
+    {
+      "epoch": 1.6016333938294012,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.5843915749224864e-05,
+      "loss": 0.1069,
+      "step": 14120
+    },
+    {
+      "epoch": 1.6027676950998186,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.5824148919049397e-05,
+      "loss": 0.1151,
+      "step": 14130
+    },
+    {
+      "epoch": 1.603901996370236,
+      "grad_norm": 3.0,
+      "learning_rate": 1.5804380653245727e-05,
+      "loss": 0.105,
+      "step": 14140
+    },
+    {
+      "epoch": 1.6050362976406534,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.5784610986249224e-05,
+      "loss": 0.1088,
+      "step": 14150
+    },
+    {
+      "epoch": 1.6061705989110708,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.576483995249772e-05,
+      "loss": 0.1202,
+      "step": 14160
+    },
+    {
+      "epoch": 1.607304900181488,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.5745067586431398e-05,
+      "loss": 0.1208,
+      "step": 14170
+    },
+    {
+      "epoch": 1.6084392014519056,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.5725293922492796e-05,
+      "loss": 0.1102,
+      "step": 14180
+    },
+    {
+      "epoch": 1.609573502722323,
+      "grad_norm": 3.0,
+      "learning_rate": 1.5705518995126683e-05,
+      "loss": 0.1135,
+      "step": 14190
+    },
+    {
+      "epoch": 1.6107078039927405,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.5685742838780054e-05,
+      "loss": 0.0999,
+      "step": 14200
+    },
+    {
+      "epoch": 1.611842105263158,
+      "grad_norm": 3.375,
+      "learning_rate": 1.566596548790202e-05,
+      "loss": 0.1103,
+      "step": 14210
+    },
+    {
+      "epoch": 1.6129764065335754,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.5646186976943794e-05,
+      "loss": 0.1084,
+      "step": 14220
+    },
+    {
+      "epoch": 1.614110707803993,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.5626407340358596e-05,
+      "loss": 0.1099,
+      "step": 14230
+    },
+    {
+      "epoch": 1.6152450090744102,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.5606626612601612e-05,
+      "loss": 0.1176,
+      "step": 14240
+    },
+    {
+      "epoch": 1.6163793103448276,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.558684482812993e-05,
+      "loss": 0.1063,
+      "step": 14250
+    },
+    {
+      "epoch": 1.6175136116152449,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5567062021402484e-05,
+      "loss": 0.1056,
+      "step": 14260
+    },
+    {
+      "epoch": 1.6186479128856623,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.5547278226879976e-05,
+      "loss": 0.1092,
+      "step": 14270
+    },
+    {
+      "epoch": 1.6197822141560798,
+      "grad_norm": 3.75,
+      "learning_rate": 1.552749347902483e-05,
+      "loss": 0.1146,
+      "step": 14280
+    },
+    {
+      "epoch": 1.6209165154264973,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.5507707812301144e-05,
+      "loss": 0.1051,
+      "step": 14290
+    },
+    {
+      "epoch": 1.6220508166969148,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.54879212611746e-05,
+      "loss": 0.1114,
+      "step": 14300
+    },
+    {
+      "epoch": 1.6231851179673322,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.5468133860112436e-05,
+      "loss": 0.101,
+      "step": 14310
+    },
+    {
+      "epoch": 1.6243194192377497,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.544834564358336e-05,
+      "loss": 0.1029,
+      "step": 14320
+    },
+    {
+      "epoch": 1.625453720508167,
+      "grad_norm": 3.875,
+      "learning_rate": 1.54285566460575e-05,
+      "loss": 0.0977,
+      "step": 14330
+    },
+    {
+      "epoch": 1.6265880217785844,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.540876690200636e-05,
+      "loss": 0.1231,
+      "step": 14340
+    },
+    {
+      "epoch": 1.6277223230490017,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.538897644590272e-05,
+      "loss": 0.1082,
+      "step": 14350
+    },
+    {
+      "epoch": 1.6288566243194191,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.5369185312220614e-05,
+      "loss": 0.1009,
+      "step": 14360
+    },
+    {
+      "epoch": 1.6299909255898366,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.5349393535435262e-05,
+      "loss": 0.1146,
+      "step": 14370
+    },
+    {
+      "epoch": 1.631125226860254,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.5329601150022995e-05,
+      "loss": 0.1181,
+      "step": 14380
+    },
+    {
+      "epoch": 1.6322595281306715,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.530980819046121e-05,
+      "loss": 0.1013,
+      "step": 14390
+    },
+    {
+      "epoch": 1.633393829401089,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.529001469122829e-05,
+      "loss": 0.1069,
+      "step": 14400
+    },
+    {
+      "epoch": 1.6345281306715065,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.527022068680358e-05,
+      "loss": 0.101,
+      "step": 14410
+    },
+    {
+      "epoch": 1.6356624319419237,
+      "grad_norm": 3.75,
+      "learning_rate": 1.525042621166729e-05,
+      "loss": 0.117,
+      "step": 14420
+    },
+    {
+      "epoch": 1.6367967332123412,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.5230631300300447e-05,
+      "loss": 0.1137,
+      "step": 14430
+    },
+    {
+      "epoch": 1.6379310344827587,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5210835987184861e-05,
+      "loss": 0.1092,
+      "step": 14440
+    },
+    {
+      "epoch": 1.639065335753176,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.5191040306803011e-05,
+      "loss": 0.1073,
+      "step": 14450
+    },
+    {
+      "epoch": 1.6401996370235934,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.5171244293638037e-05,
+      "loss": 0.1184,
+      "step": 14460
+    },
+    {
+      "epoch": 1.6413339382940109,
+      "grad_norm": 3.25,
+      "learning_rate": 1.5151447982173655e-05,
+      "loss": 0.1048,
+      "step": 14470
+    },
+    {
+      "epoch": 1.6424682395644283,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.5131651406894088e-05,
+      "loss": 0.1046,
+      "step": 14480
+    },
+    {
+      "epoch": 1.6436025408348458,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.5111854602284037e-05,
+      "loss": 0.1045,
+      "step": 14490
+    },
+    {
+      "epoch": 1.6447368421052633,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.5092057602828588e-05,
+      "loss": 0.1134,
+      "step": 14500
+    },
+    {
+      "epoch": 1.6458711433756807,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.5072260443013175e-05,
+      "loss": 0.1127,
+      "step": 14510
+    },
+    {
+      "epoch": 1.647005444646098,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.5052463157323511e-05,
+      "loss": 0.1067,
+      "step": 14520
+    },
+    {
+      "epoch": 1.6481397459165155,
+      "grad_norm": 3.75,
+      "learning_rate": 1.5032665780245521e-05,
+      "loss": 0.1042,
+      "step": 14530
+    },
+    {
+      "epoch": 1.6492740471869327,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.5012868346265298e-05,
+      "loss": 0.1183,
+      "step": 14540
+    },
+    {
+      "epoch": 1.6504083484573502,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.4993070889869025e-05,
+      "loss": 0.1162,
+      "step": 14550
+    },
+    {
+      "epoch": 1.6515426497277677,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.4973273445542936e-05,
+      "loss": 0.1083,
+      "step": 14560
+    },
+    {
+      "epoch": 1.6526769509981851,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.4953476047773229e-05,
+      "loss": 0.1077,
+      "step": 14570
+    },
+    {
+      "epoch": 1.6538112522686026,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4933678731046038e-05,
+      "loss": 0.1106,
+      "step": 14580
+    },
+    {
+      "epoch": 1.65494555353902,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.4913881529847336e-05,
+      "loss": 0.1175,
+      "step": 14590
+    },
+    {
+      "epoch": 1.6560798548094375,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.4894084478662913e-05,
+      "loss": 0.1115,
+      "step": 14600
+    },
+    {
+      "epoch": 1.6572141560798548,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.4874287611978293e-05,
+      "loss": 0.1091,
+      "step": 14610
+    },
+    {
+      "epoch": 1.6583484573502723,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.485449096427867e-05,
+      "loss": 0.1159,
+      "step": 14620
+    },
+    {
+      "epoch": 1.6594827586206895,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.4834694570048866e-05,
+      "loss": 0.108,
+      "step": 14630
+    },
+    {
+      "epoch": 1.660617059891107,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.4814898463773253e-05,
+      "loss": 0.1024,
+      "step": 14640
+    },
+    {
+      "epoch": 1.6617513611615244,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.4795102679935707e-05,
+      "loss": 0.1091,
+      "step": 14650
+    },
+    {
+      "epoch": 1.662885662431942,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.4775307253019551e-05,
+      "loss": 0.1104,
+      "step": 14660
+    },
+    {
+      "epoch": 1.6640199637023594,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.4755512217507465e-05,
+      "loss": 0.1112,
+      "step": 14670
+    },
+    {
+      "epoch": 1.6651542649727769,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.4735717607881468e-05,
+      "loss": 0.1019,
+      "step": 14680
+    },
+    {
+      "epoch": 1.6662885662431943,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.4715923458622822e-05,
+      "loss": 0.1138,
+      "step": 14690
+    },
+    {
+      "epoch": 1.6674228675136116,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.4696129804211999e-05,
+      "loss": 0.1143,
+      "step": 14700
+    },
+    {
+      "epoch": 1.668557168784029,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.4676336679128592e-05,
+      "loss": 0.1158,
+      "step": 14710
+    },
+    {
+      "epoch": 1.6696914700544465,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.4656544117851294e-05,
+      "loss": 0.1084,
+      "step": 14720
+    },
+    {
+      "epoch": 1.6708257713248638,
+      "grad_norm": 3.0,
+      "learning_rate": 1.4636752154857806e-05,
+      "loss": 0.1137,
+      "step": 14730
+    },
+    {
+      "epoch": 1.6719600725952812,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.4616960824624778e-05,
+      "loss": 0.1084,
+      "step": 14740
+    },
+    {
+      "epoch": 1.6730943738656987,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.4597170161627775e-05,
+      "loss": 0.1169,
+      "step": 14750
+    },
+    {
+      "epoch": 1.6742286751361162,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.4577380200341175e-05,
+      "loss": 0.1118,
+      "step": 14760
+    },
+    {
+      "epoch": 1.6753629764065336,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.4557590975238165e-05,
+      "loss": 0.1068,
+      "step": 14770
+    },
+    {
+      "epoch": 1.6764972776769511,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.4537802520790629e-05,
+      "loss": 0.1133,
+      "step": 14780
+    },
+    {
+      "epoch": 1.6776315789473686,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.451801487146911e-05,
+      "loss": 0.1063,
+      "step": 14790
+    },
+    {
+      "epoch": 1.6787658802177858,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.4498228061742757e-05,
+      "loss": 0.1129,
+      "step": 14800
+    },
+    {
+      "epoch": 1.6799001814882033,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.4478442126079238e-05,
+      "loss": 0.1181,
+      "step": 14810
+    },
+    {
+      "epoch": 1.6810344827586206,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.4458657098944729e-05,
+      "loss": 0.109,
+      "step": 14820
+    },
+    {
+      "epoch": 1.682168784029038,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.4438873014803796e-05,
+      "loss": 0.1163,
+      "step": 14830
+    },
+    {
+      "epoch": 1.6833030852994555,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.4419089908119373e-05,
+      "loss": 0.1085,
+      "step": 14840
+    },
+    {
+      "epoch": 1.684437386569873,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.4399307813352697e-05,
+      "loss": 0.113,
+      "step": 14850
+    },
+    {
+      "epoch": 1.6855716878402904,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.4379526764963223e-05,
+      "loss": 0.0985,
+      "step": 14860
+    },
+    {
+      "epoch": 1.686705989110708,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.435974679740861e-05,
+      "loss": 0.106,
+      "step": 14870
+    },
+    {
+      "epoch": 1.6878402903811254,
+      "grad_norm": 3.125,
+      "learning_rate": 1.4339967945144612e-05,
+      "loss": 0.1155,
+      "step": 14880
+    },
+    {
+      "epoch": 1.6889745916515426,
+      "grad_norm": 4.5,
+      "learning_rate": 1.4320190242625056e-05,
+      "loss": 0.1091,
+      "step": 14890
+    },
+    {
+      "epoch": 1.69010889292196,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.430041372430175e-05,
+      "loss": 0.1046,
+      "step": 14900
+    },
+    {
+      "epoch": 1.6912431941923773,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.4280638424624446e-05,
+      "loss": 0.1128,
+      "step": 14910
+    },
+    {
+      "epoch": 1.6923774954627948,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.4260864378040788e-05,
+      "loss": 0.1127,
+      "step": 14920
+    },
+    {
+      "epoch": 1.6935117967332123,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.4241091618996215e-05,
+      "loss": 0.1048,
+      "step": 14930
+    },
+    {
+      "epoch": 1.6946460980036298,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.422132018193394e-05,
+      "loss": 0.1041,
+      "step": 14940
+    },
+    {
+      "epoch": 1.6957803992740472,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4201550101294854e-05,
+      "loss": 0.1092,
+      "step": 14950
+    },
+    {
+      "epoch": 1.6969147005444647,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.4181781411517503e-05,
+      "loss": 0.0999,
+      "step": 14960
+    },
+    {
+      "epoch": 1.6980490018148822,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.416201414703801e-05,
+      "loss": 0.1142,
+      "step": 14970
+    },
+    {
+      "epoch": 1.6991833030852994,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.4142248342290001e-05,
+      "loss": 0.1048,
+      "step": 14980
+    },
+    {
+      "epoch": 1.7003176043557169,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.4122484031704572e-05,
+      "loss": 0.1244,
+      "step": 14990
+    },
+    {
+      "epoch": 1.7014519056261344,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.410272124971021e-05,
+      "loss": 0.1077,
+      "step": 15000
+    },
+    {
+      "epoch": 1.7025862068965516,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.4082960030732742e-05,
+      "loss": 0.1038,
+      "step": 15010
+    },
+    {
+      "epoch": 1.703720508166969,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.4063200409195264e-05,
+      "loss": 0.1034,
+      "step": 15020
+    },
+    {
+      "epoch": 1.7048548094373865,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4043442419518107e-05,
+      "loss": 0.103,
+      "step": 15030
+    },
+    {
+      "epoch": 1.705989110707804,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.4023686096118747e-05,
+      "loss": 0.114,
+      "step": 15040
+    },
+    {
+      "epoch": 1.7071234119782215,
+      "grad_norm": 2.875,
+      "learning_rate": 1.4003931473411754e-05,
+      "loss": 0.1178,
+      "step": 15050
+    },
+    {
+      "epoch": 1.708257713248639,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.3984178585808747e-05,
+      "loss": 0.1091,
+      "step": 15060
+    },
+    {
+      "epoch": 1.7093920145190564,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.3964427467718307e-05,
+      "loss": 0.1071,
+      "step": 15070
+    },
+    {
+      "epoch": 1.7105263157894737,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.3944678153545959e-05,
+      "loss": 0.1202,
+      "step": 15080
+    },
+    {
+      "epoch": 1.7116606170598911,
+      "grad_norm": 3.5,
+      "learning_rate": 1.3924930677694049e-05,
+      "loss": 0.1182,
+      "step": 15090
+    },
+    {
+      "epoch": 1.7127949183303084,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.3905185074561752e-05,
+      "loss": 0.1087,
+      "step": 15100
+    },
+    {
+      "epoch": 1.7139292196007259,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.3885441378544972e-05,
+      "loss": 0.1112,
+      "step": 15110
+    },
+    {
+      "epoch": 1.7150635208711433,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.3865699624036273e-05,
+      "loss": 0.1118,
+      "step": 15120
+    },
+    {
+      "epoch": 1.7161978221415608,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.3845959845424874e-05,
+      "loss": 0.1065,
+      "step": 15130
+    },
+    {
+      "epoch": 1.7173321234119783,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.3826222077096509e-05,
+      "loss": 0.0999,
+      "step": 15140
+    },
+    {
+      "epoch": 1.7184664246823957,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.3806486353433443e-05,
+      "loss": 0.1054,
+      "step": 15150
+    },
+    {
+      "epoch": 1.7196007259528132,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.3786752708814369e-05,
+      "loss": 0.1114,
+      "step": 15160
+    },
+    {
+      "epoch": 1.7207350272232305,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.3767021177614345e-05,
+      "loss": 0.1098,
+      "step": 15170
+    },
+    {
+      "epoch": 1.721869328493648,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.374729179420478e-05,
+      "loss": 0.1096,
+      "step": 15180
+    },
+    {
+      "epoch": 1.7230036297640652,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.3727564592953299e-05,
+      "loss": 0.1045,
+      "step": 15190
+    },
+    {
+      "epoch": 1.7241379310344827,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.3707839608223764e-05,
+      "loss": 0.1121,
+      "step": 15200
+    },
+    {
+      "epoch": 1.7252722323049001,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.368811687437615e-05,
+      "loss": 0.1176,
+      "step": 15210
+    },
+    {
+      "epoch": 1.7264065335753176,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.3668396425766522e-05,
+      "loss": 0.1027,
+      "step": 15220
+    },
+    {
+      "epoch": 1.727540834845735,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.3648678296746978e-05,
+      "loss": 0.116,
+      "step": 15230
+    },
+    {
+      "epoch": 1.7286751361161525,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.3628962521665538e-05,
+      "loss": 0.1153,
+      "step": 15240
+    },
+    {
+      "epoch": 1.72980943738657,
+      "grad_norm": 3.625,
+      "learning_rate": 1.360924913486616e-05,
+      "loss": 0.1105,
+      "step": 15250
+    },
+    {
+      "epoch": 1.7309437386569873,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.358953817068862e-05,
+      "loss": 0.1058,
+      "step": 15260
+    },
+    {
+      "epoch": 1.7320780399274047,
+      "grad_norm": 3.0,
+      "learning_rate": 1.3569829663468483e-05,
+      "loss": 0.1057,
+      "step": 15270
+    },
+    {
+      "epoch": 1.7332123411978222,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.3550123647537026e-05,
+      "loss": 0.1126,
+      "step": 15280
+    },
+    {
+      "epoch": 1.7343466424682394,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.353042015722119e-05,
+      "loss": 0.1049,
+      "step": 15290
+    },
+    {
+      "epoch": 1.735480943738657,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.3510719226843529e-05,
+      "loss": 0.1077,
+      "step": 15300
+    },
+    {
+      "epoch": 1.7366152450090744,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.3491020890722115e-05,
+      "loss": 0.1081,
+      "step": 15310
+    },
+    {
+      "epoch": 1.7377495462794919,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.347132518317052e-05,
+      "loss": 0.1027,
+      "step": 15320
+    },
+    {
+      "epoch": 1.7388838475499093,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.3451632138497721e-05,
+      "loss": 0.1179,
+      "step": 15330
+    },
+    {
+      "epoch": 1.7400181488203268,
+      "grad_norm": 3.25,
+      "learning_rate": 1.3431941791008069e-05,
+      "loss": 0.1131,
+      "step": 15340
+    },
+    {
+      "epoch": 1.7411524500907443,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.3412254175001218e-05,
+      "loss": 0.1187,
+      "step": 15350
+    },
+    {
+      "epoch": 1.7422867513611615,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.339256932477205e-05,
+      "loss": 0.1082,
+      "step": 15360
+    },
+    {
+      "epoch": 1.743421052631579,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.3372887274610646e-05,
+      "loss": 0.1069,
+      "step": 15370
+    },
+    {
+      "epoch": 1.7445553539019962,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.3353208058802192e-05,
+      "loss": 0.1052,
+      "step": 15380
+    },
+    {
+      "epoch": 1.7456896551724137,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.3333531711626953e-05,
+      "loss": 0.1064,
+      "step": 15390
+    },
+    {
+      "epoch": 1.7468239564428312,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.331385826736018e-05,
+      "loss": 0.1036,
+      "step": 15400
+    },
+    {
+      "epoch": 1.7479582577132486,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.3294187760272088e-05,
+      "loss": 0.1144,
+      "step": 15410
+    },
+    {
+      "epoch": 1.7490925589836661,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.327452022462776e-05,
+      "loss": 0.1114,
+      "step": 15420
+    },
+    {
+      "epoch": 1.7502268602540836,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.3254855694687105e-05,
+      "loss": 0.1041,
+      "step": 15430
+    },
+    {
+      "epoch": 1.751361161524501,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.3235194204704806e-05,
+      "loss": 0.112,
+      "step": 15440
+    },
+    {
+      "epoch": 1.7524954627949183,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.3215535788930229e-05,
+      "loss": 0.1017,
+      "step": 15450
+    },
+    {
+      "epoch": 1.7536297640653358,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.3195880481607413e-05,
+      "loss": 0.1084,
+      "step": 15460
+    },
+    {
+      "epoch": 1.754764065335753,
+      "grad_norm": 3.375,
+      "learning_rate": 1.3176228316974965e-05,
+      "loss": 0.1053,
+      "step": 15470
+    },
+    {
+      "epoch": 1.7558983666061705,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.3156579329266014e-05,
+      "loss": 0.1143,
+      "step": 15480
+    },
+    {
+      "epoch": 1.757032667876588,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.3136933552708173e-05,
+      "loss": 0.1181,
+      "step": 15490
+    },
+    {
+      "epoch": 1.7581669691470054,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.3117291021523435e-05,
+      "loss": 0.1136,
+      "step": 15500
+    },
+    {
+      "epoch": 1.759301270417423,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3097651769928174e-05,
+      "loss": 0.1031,
+      "step": 15510
+    },
+    {
+      "epoch": 1.7604355716878404,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.3078015832133011e-05,
+      "loss": 0.1063,
+      "step": 15520
+    },
+    {
+      "epoch": 1.7615698729582578,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.305838324234283e-05,
+      "loss": 0.1054,
+      "step": 15530
+    },
+    {
+      "epoch": 1.762704174228675,
+      "grad_norm": 3.5,
+      "learning_rate": 1.3038754034756668e-05,
+      "loss": 0.1132,
+      "step": 15540
+    },
+    {
+      "epoch": 1.7638384754990926,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.3019128243567658e-05,
+      "loss": 0.1131,
+      "step": 15550
+    },
+    {
+      "epoch": 1.76497277676951,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.2999505902963014e-05,
+      "loss": 0.1173,
+      "step": 15560
+    },
+    {
+      "epoch": 1.7661070780399273,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.29798870471239e-05,
+      "loss": 0.1099,
+      "step": 15570
+    },
+    {
+      "epoch": 1.7672413793103448,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.2960271710225445e-05,
+      "loss": 0.1137,
+      "step": 15580
+    },
+    {
+      "epoch": 1.7683756805807622,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.2940659926436622e-05,
+      "loss": 0.1086,
+      "step": 15590
+    },
+    {
+      "epoch": 1.7695099818511797,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.2921051729920227e-05,
+      "loss": 0.1078,
+      "step": 15600
+    },
+    {
+      "epoch": 1.7706442831215972,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.2901447154832818e-05,
+      "loss": 0.1069,
+      "step": 15610
+    },
+    {
+      "epoch": 1.7717785843920146,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.2881846235324611e-05,
+      "loss": 0.1117,
+      "step": 15620
+    },
+    {
+      "epoch": 1.772912885662432,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.2862249005539493e-05,
+      "loss": 0.1169,
+      "step": 15630
+    },
+    {
+      "epoch": 1.7740471869328494,
+      "grad_norm": 2.75,
+      "learning_rate": 1.2842655499614896e-05,
+      "loss": 0.1054,
+      "step": 15640
+    },
+    {
+      "epoch": 1.7751814882032668,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.2823065751681776e-05,
+      "loss": 0.1127,
+      "step": 15650
+    },
+    {
+      "epoch": 1.776315789473684,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.2803479795864552e-05,
+      "loss": 0.1136,
+      "step": 15660
+    },
+    {
+      "epoch": 1.7774500907441015,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.278389766628101e-05,
+      "loss": 0.1087,
+      "step": 15670
+    },
+    {
+      "epoch": 1.778584392014519,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.2764319397042304e-05,
+      "loss": 0.1088,
+      "step": 15680
+    },
+    {
+      "epoch": 1.7797186932849365,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.274474502225284e-05,
+      "loss": 0.1205,
+      "step": 15690
+    },
+    {
+      "epoch": 1.780852994555354,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.2725174576010248e-05,
+      "loss": 0.1089,
+      "step": 15700
+    },
+    {
+      "epoch": 1.7819872958257714,
+      "grad_norm": 3.875,
+      "learning_rate": 1.2705608092405314e-05,
+      "loss": 0.1022,
+      "step": 15710
+    },
+    {
+      "epoch": 1.783121597096189,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.2686045605521917e-05,
+      "loss": 0.1032,
+      "step": 15720
+    },
+    {
+      "epoch": 1.7842558983666061,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.266648714943699e-05,
+      "loss": 0.1145,
+      "step": 15730
+    },
+    {
+      "epoch": 1.7853901996370236,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.2646932758220425e-05,
+      "loss": 0.1087,
+      "step": 15740
+    },
+    {
+      "epoch": 1.7865245009074409,
+      "grad_norm": 3.25,
+      "learning_rate": 1.2627382465935046e-05,
+      "loss": 0.1089,
+      "step": 15750
+    },
+    {
+      "epoch": 1.7876588021778583,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.2607836306636525e-05,
+      "loss": 0.1063,
+      "step": 15760
+    },
+    {
+      "epoch": 1.7887931034482758,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.2588294314373349e-05,
+      "loss": 0.1141,
+      "step": 15770
+    },
+    {
+      "epoch": 1.7899274047186933,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.2568756523186727e-05,
+      "loss": 0.1054,
+      "step": 15780
+    },
+    {
+      "epoch": 1.7910617059891107,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.2549222967110576e-05,
+      "loss": 0.1071,
+      "step": 15790
+    },
+    {
+      "epoch": 1.7921960072595282,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.2529693680171415e-05,
+      "loss": 0.1066,
+      "step": 15800
+    },
+    {
+      "epoch": 1.7933303085299457,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.251016869638833e-05,
+      "loss": 0.1095,
+      "step": 15810
+    },
+    {
+      "epoch": 1.794464609800363,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.2490648049772923e-05,
+      "loss": 0.1035,
+      "step": 15820
+    },
+    {
+      "epoch": 1.7955989110707804,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.2471131774329217e-05,
+      "loss": 0.1105,
+      "step": 15830
+    },
+    {
+      "epoch": 1.7967332123411979,
+      "grad_norm": 3.75,
+      "learning_rate": 1.2451619904053647e-05,
+      "loss": 0.1179,
+      "step": 15840
+    },
+    {
+      "epoch": 1.7978675136116151,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2432112472934963e-05,
+      "loss": 0.1158,
+      "step": 15850
+    },
+    {
+      "epoch": 1.7990018148820326,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.2412609514954172e-05,
+      "loss": 0.1125,
+      "step": 15860
+    },
+    {
+      "epoch": 1.80013611615245,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.239311106408451e-05,
+      "loss": 0.1119,
+      "step": 15870
+    },
+    {
+      "epoch": 1.8012704174228675,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.2373617154291342e-05,
+      "loss": 0.1048,
+      "step": 15880
+    },
+    {
+      "epoch": 1.802404718693285,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.2354127819532137e-05,
+      "loss": 0.1102,
+      "step": 15890
+    },
+    {
+      "epoch": 1.8035390199637025,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.2334643093756388e-05,
+      "loss": 0.1077,
+      "step": 15900
+    },
+    {
+      "epoch": 1.80467332123412,
+      "grad_norm": 3.625,
+      "learning_rate": 1.2315163010905557e-05,
+      "loss": 0.1129,
+      "step": 15910
+    },
+    {
+      "epoch": 1.8058076225045372,
+      "grad_norm": 4.0,
+      "learning_rate": 1.2295687604913027e-05,
+      "loss": 0.1129,
+      "step": 15920
+    },
+    {
+      "epoch": 1.8069419237749547,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.227621690970402e-05,
+      "loss": 0.1168,
+      "step": 15930
+    },
+    {
+      "epoch": 1.808076225045372,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.2256750959195574e-05,
+      "loss": 0.105,
+      "step": 15940
+    },
+    {
+      "epoch": 1.8092105263157894,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.2237289787296435e-05,
+      "loss": 0.1086,
+      "step": 15950
+    },
+    {
+      "epoch": 1.8103448275862069,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.2217833427907046e-05,
+      "loss": 0.1004,
+      "step": 15960
+    },
+    {
+      "epoch": 1.8114791288566243,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.2198381914919461e-05,
+      "loss": 0.1123,
+      "step": 15970
+    },
+    {
+      "epoch": 1.8126134301270418,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.2178935282217282e-05,
+      "loss": 0.1095,
+      "step": 15980
+    },
+    {
+      "epoch": 1.8137477313974593,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.2159493563675627e-05,
+      "loss": 0.1043,
+      "step": 15990
+    },
+    {
+      "epoch": 1.8148820326678767,
+      "grad_norm": 2.875,
+      "learning_rate": 1.2140056793161037e-05,
+      "loss": 0.1148,
+      "step": 16000
+    },
+    {
+      "epoch": 1.816016333938294,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.212062500453145e-05,
+      "loss": 0.1128,
+      "step": 16010
+    },
+    {
+      "epoch": 1.8171506352087115,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.210119823163611e-05,
+      "loss": 0.1069,
+      "step": 16020
+    },
+    {
+      "epoch": 1.8182849364791287,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2081776508315528e-05,
+      "loss": 0.1032,
+      "step": 16030
+    },
+    {
+      "epoch": 1.8194192377495462,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.2062359868401436e-05,
+      "loss": 0.1087,
+      "step": 16040
+    },
+    {
+      "epoch": 1.8205535390199636,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2042948345716685e-05,
+      "loss": 0.1136,
+      "step": 16050
+    },
+    {
+      "epoch": 1.821687840290381,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.202354197407523e-05,
+      "loss": 0.1101,
+      "step": 16060
+    },
+    {
+      "epoch": 1.8228221415607986,
+      "grad_norm": 3.0,
+      "learning_rate": 1.2004140787282039e-05,
+      "loss": 0.1018,
+      "step": 16070
+    },
+    {
+      "epoch": 1.823956442831216,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.1984744819133069e-05,
+      "loss": 0.1137,
+      "step": 16080
+    },
+    {
+      "epoch": 1.8250907441016335,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.1965354103415159e-05,
+      "loss": 0.104,
+      "step": 16090
+    },
+    {
+      "epoch": 1.8262250453720508,
+      "grad_norm": 3.25,
+      "learning_rate": 1.1945968673906021e-05,
+      "loss": 0.1095,
+      "step": 16100
+    },
+    {
+      "epoch": 1.8273593466424682,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.1926588564374159e-05,
+      "loss": 0.1069,
+      "step": 16110
+    },
+    {
+      "epoch": 1.8284936479128857,
+      "grad_norm": 3.875,
+      "learning_rate": 1.1907213808578785e-05,
+      "loss": 0.1154,
+      "step": 16120
+    },
+    {
+      "epoch": 1.829627949183303,
+      "grad_norm": 3.125,
+      "learning_rate": 1.1887844440269818e-05,
+      "loss": 0.1016,
+      "step": 16130
+    },
+    {
+      "epoch": 1.8307622504537204,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.186848049318776e-05,
+      "loss": 0.12,
+      "step": 16140
+    },
+    {
+      "epoch": 1.831896551724138,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.1849122001063697e-05,
+      "loss": 0.1114,
+      "step": 16150
+    },
+    {
+      "epoch": 1.8330308529945554,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.1829768997619202e-05,
+      "loss": 0.113,
+      "step": 16160
+    },
+    {
+      "epoch": 1.8341651542649728,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.1810421516566279e-05,
+      "loss": 0.1018,
+      "step": 16170
+    },
+    {
+      "epoch": 1.8352994555353903,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.1791079591607325e-05,
+      "loss": 0.1084,
+      "step": 16180
+    },
+    {
+      "epoch": 1.8364337568058078,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.1771743256435044e-05,
+      "loss": 0.1107,
+      "step": 16190
+    },
+    {
+      "epoch": 1.837568058076225,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.175241254473243e-05,
+      "loss": 0.1104,
+      "step": 16200
+    },
+    {
+      "epoch": 1.8387023593466425,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.1733087490172643e-05,
+      "loss": 0.1077,
+      "step": 16210
+    },
+    {
+      "epoch": 1.8398366606170597,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.171376812641902e-05,
+      "loss": 0.113,
+      "step": 16220
+    },
+    {
+      "epoch": 1.8409709618874772,
+      "grad_norm": 3.875,
+      "learning_rate": 1.1694454487124972e-05,
+      "loss": 0.1119,
+      "step": 16230
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.1675146605933935e-05,
+      "loss": 0.0998,
+      "step": 16240
+    },
+    {
+      "epoch": 1.8432395644283122,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.1655844516479332e-05,
+      "loss": 0.1026,
+      "step": 16250
+    },
+    {
+      "epoch": 1.8443738656987296,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.1636548252384465e-05,
+      "loss": 0.1174,
+      "step": 16260
+    },
+    {
+      "epoch": 1.845508166969147,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.1617257847262526e-05,
+      "loss": 0.1126,
+      "step": 16270
+    },
+    {
+      "epoch": 1.8466424682395646,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.1597973334716473e-05,
+      "loss": 0.1218,
+      "step": 16280
+    },
+    {
+      "epoch": 1.8477767695099818,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.1578694748339008e-05,
+      "loss": 0.1109,
+      "step": 16290
+    },
+    {
+      "epoch": 1.8489110707803993,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1559422121712524e-05,
+      "loss": 0.1029,
+      "step": 16300
+    },
+    {
+      "epoch": 1.8500453720508165,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.1540155488409001e-05,
+      "loss": 0.1077,
+      "step": 16310
+    },
+    {
+      "epoch": 1.851179673321234,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.1520894881990016e-05,
+      "loss": 0.1092,
+      "step": 16320
+    },
+    {
+      "epoch": 1.8523139745916515,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.1501640336006616e-05,
+      "loss": 0.1026,
+      "step": 16330
+    },
+    {
+      "epoch": 1.853448275862069,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.1482391883999306e-05,
+      "loss": 0.1043,
+      "step": 16340
+    },
+    {
+      "epoch": 1.8545825771324864,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.1463149559497984e-05,
+      "loss": 0.1143,
+      "step": 16350
+    },
+    {
+      "epoch": 1.855716878402904,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.1443913396021846e-05,
+      "loss": 0.116,
+      "step": 16360
+    },
+    {
+      "epoch": 1.8568511796733214,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.1424683427079392e-05,
+      "loss": 0.1077,
+      "step": 16370
+    },
+    {
+      "epoch": 1.8579854809437386,
+      "grad_norm": 3.125,
+      "learning_rate": 1.14054596861683e-05,
+      "loss": 0.1002,
+      "step": 16380
+    },
+    {
+      "epoch": 1.859119782214156,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.138624220677542e-05,
+      "loss": 0.1082,
+      "step": 16390
+    },
+    {
+      "epoch": 1.8602540834845736,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1367031022376679e-05,
+      "loss": 0.1026,
+      "step": 16400
+    },
+    {
+      "epoch": 1.8613883847549908,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.1347826166437049e-05,
+      "loss": 0.0937,
+      "step": 16410
+    },
+    {
+      "epoch": 1.8625226860254083,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.1328627672410482e-05,
+      "loss": 0.1066,
+      "step": 16420
+    },
+    {
+      "epoch": 1.8636569872958257,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.1309435573739831e-05,
+      "loss": 0.1068,
+      "step": 16430
+    },
+    {
+      "epoch": 1.8647912885662432,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.129024990385683e-05,
+      "loss": 0.1064,
+      "step": 16440
+    },
+    {
+      "epoch": 1.8659255898366607,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.1271070696181991e-05,
+      "loss": 0.1123,
+      "step": 16450
+    },
+    {
+      "epoch": 1.8670598911070782,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.1251897984124593e-05,
+      "loss": 0.1203,
+      "step": 16460
+    },
+    {
+      "epoch": 1.8681941923774956,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.1232731801082574e-05,
+      "loss": 0.1027,
+      "step": 16470
+    },
+    {
+      "epoch": 1.8693284936479129,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.1213572180442526e-05,
+      "loss": 0.1159,
+      "step": 16480
+    },
+    {
+      "epoch": 1.8704627949183303,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.1194419155579594e-05,
+      "loss": 0.1003,
+      "step": 16490
+    },
+    {
+      "epoch": 1.8715970961887476,
+      "grad_norm": 3.25,
+      "learning_rate": 1.1175272759857432e-05,
+      "loss": 0.1145,
+      "step": 16500
+    },
+    {
+      "epoch": 1.872731397459165,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.1156133026628153e-05,
+      "loss": 0.1132,
+      "step": 16510
+    },
+    {
+      "epoch": 1.8738656987295825,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.1136999989232257e-05,
+      "loss": 0.118,
+      "step": 16520
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.1117873680998593e-05,
+      "loss": 0.1,
+      "step": 16530
+    },
+    {
+      "epoch": 1.8761343012704175,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.1098754135244276e-05,
+      "loss": 0.1104,
+      "step": 16540
+    },
+    {
+      "epoch": 1.877268602540835,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.1079641385274644e-05,
+      "loss": 0.1003,
+      "step": 16550
+    },
+    {
+      "epoch": 1.8784029038112524,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.1060535464383202e-05,
+      "loss": 0.1116,
+      "step": 16560
+    },
+    {
+      "epoch": 1.8795372050816697,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.1041436405851546e-05,
+      "loss": 0.1036,
+      "step": 16570
+    },
+    {
+      "epoch": 1.8806715063520871,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1022344242949342e-05,
+      "loss": 0.1171,
+      "step": 16580
+    },
+    {
+      "epoch": 1.8818058076225044,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.1003259008934212e-05,
+      "loss": 0.111,
+      "step": 16590
+    },
+    {
+      "epoch": 1.8829401088929218,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0984180737051737e-05,
+      "loss": 0.1059,
+      "step": 16600
+    },
+    {
+      "epoch": 1.8840744101633393,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.0965109460535362e-05,
+      "loss": 0.1118,
+      "step": 16610
+    },
+    {
+      "epoch": 1.8852087114337568,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.094604521260633e-05,
+      "loss": 0.1199,
+      "step": 16620
+    },
+    {
+      "epoch": 1.8863430127041743,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.092698802647367e-05,
+      "loss": 0.1119,
+      "step": 16630
+    },
+    {
+      "epoch": 1.8874773139745917,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.0907937935334078e-05,
+      "loss": 0.114,
+      "step": 16640
+    },
+    {
+      "epoch": 1.8886116152450092,
+      "grad_norm": 3.875,
+      "learning_rate": 1.0888894972371916e-05,
+      "loss": 0.1143,
+      "step": 16650
+    },
+    {
+      "epoch": 1.8897459165154264,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.086985917075912e-05,
+      "loss": 0.0993,
+      "step": 16660
+    },
+    {
+      "epoch": 1.890880217785844,
+      "grad_norm": 3.0,
+      "learning_rate": 1.0850830563655139e-05,
+      "loss": 0.1201,
+      "step": 16670
+    },
+    {
+      "epoch": 1.8920145190562614,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.0831809184206923e-05,
+      "loss": 0.1116,
+      "step": 16680
+    },
+    {
+      "epoch": 1.8931488203266786,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.0812795065548789e-05,
+      "loss": 0.1027,
+      "step": 16690
+    },
+    {
+      "epoch": 1.894283121597096,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.0793788240802438e-05,
+      "loss": 0.1088,
+      "step": 16700
+    },
+    {
+      "epoch": 1.8954174228675136,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.077478874307685e-05,
+      "loss": 0.1068,
+      "step": 16710
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.0755796605468245e-05,
+      "loss": 0.1001,
+      "step": 16720
+    },
+    {
+      "epoch": 1.8976860254083485,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.0736811861060035e-05,
+      "loss": 0.0977,
+      "step": 16730
+    },
+    {
+      "epoch": 1.898820326678766,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.0717834542922722e-05,
+      "loss": 0.1007,
+      "step": 16740
+    },
+    {
+      "epoch": 1.8999546279491835,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.0698864684113909e-05,
+      "loss": 0.1147,
+      "step": 16750
+    },
+    {
+      "epoch": 1.9010889292196007,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.0679902317678177e-05,
+      "loss": 0.1131,
+      "step": 16760
+    },
+    {
+      "epoch": 1.9022232304900182,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.0660947476647071e-05,
+      "loss": 0.1144,
+      "step": 16770
+    },
+    {
+      "epoch": 1.9033575317604354,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.064200019403902e-05,
+      "loss": 0.1097,
+      "step": 16780
+    },
+    {
+      "epoch": 1.904491833030853,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.0623060502859285e-05,
+      "loss": 0.1126,
+      "step": 16790
+    },
+    {
+      "epoch": 1.9056261343012704,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.0604128436099916e-05,
+      "loss": 0.1127,
+      "step": 16800
+    },
+    {
+      "epoch": 1.9067604355716878,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0585204026739669e-05,
+      "loss": 0.1117,
+      "step": 16810
+    },
+    {
+      "epoch": 1.9078947368421053,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0566287307743963e-05,
+      "loss": 0.113,
+      "step": 16820
+    },
+    {
+      "epoch": 1.9090290381125228,
+      "grad_norm": 3.125,
+      "learning_rate": 1.0547378312064822e-05,
+      "loss": 0.1067,
+      "step": 16830
+    },
+    {
+      "epoch": 1.9101633393829403,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.0528477072640813e-05,
+      "loss": 0.1159,
+      "step": 16840
+    },
+    {
+      "epoch": 1.9112976406533575,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.0509583622397008e-05,
+      "loss": 0.0994,
+      "step": 16850
+    },
+    {
+      "epoch": 1.912431941923775,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.049069799424489e-05,
+      "loss": 0.1097,
+      "step": 16860
+    },
+    {
+      "epoch": 1.9135662431941922,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.0471820221082326e-05,
+      "loss": 0.1087,
+      "step": 16870
+    },
+    {
+      "epoch": 1.9147005444646097,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.0452950335793498e-05,
+      "loss": 0.1083,
+      "step": 16880
+    },
+    {
+      "epoch": 1.9158348457350272,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.043408837124885e-05,
+      "loss": 0.1113,
+      "step": 16890
+    },
+    {
+      "epoch": 1.9169691470054446,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.0415234360305022e-05,
+      "loss": 0.1169,
+      "step": 16900
+    },
+    {
+      "epoch": 1.918103448275862,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.0396388335804809e-05,
+      "loss": 0.1094,
+      "step": 16910
+    },
+    {
+      "epoch": 1.9192377495462796,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0377550330577093e-05,
+      "loss": 0.1027,
+      "step": 16920
+    },
+    {
+      "epoch": 1.920372050816697,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.0358720377436771e-05,
+      "loss": 0.112,
+      "step": 16930
+    },
+    {
+      "epoch": 1.9215063520871143,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.0339898509184739e-05,
+      "loss": 0.1091,
+      "step": 16940
+    },
+    {
+      "epoch": 1.9226406533575318,
+      "grad_norm": 3.25,
+      "learning_rate": 1.0321084758607781e-05,
+      "loss": 0.1095,
+      "step": 16950
+    },
+    {
+      "epoch": 1.9237749546279492,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.030227915847857e-05,
+      "loss": 0.1136,
+      "step": 16960
+    },
+    {
+      "epoch": 1.9249092558983665,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.028348174155556e-05,
+      "loss": 0.1035,
+      "step": 16970
+    },
+    {
+      "epoch": 1.926043557168784,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.0264692540582956e-05,
+      "loss": 0.1061,
+      "step": 16980
+    },
+    {
+      "epoch": 1.9271778584392014,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.0245911588290655e-05,
+      "loss": 0.1067,
+      "step": 16990
+    },
+    {
+      "epoch": 1.928312159709619,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.0227138917394176e-05,
+      "loss": 0.1088,
+      "step": 17000
+    },
+    {
+      "epoch": 1.9294464609800364,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.0208374560594629e-05,
+      "loss": 0.1175,
+      "step": 17010
+    },
+    {
+      "epoch": 1.9305807622504538,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.0189618550578623e-05,
+      "loss": 0.1093,
+      "step": 17020
+    },
+    {
+      "epoch": 1.9317150635208713,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.0170870920018236e-05,
+      "loss": 0.1151,
+      "step": 17030
+    },
+    {
+      "epoch": 1.9328493647912885,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.0152131701570953e-05,
+      "loss": 0.108,
+      "step": 17040
+    },
+    {
+      "epoch": 1.933983666061706,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0133400927879593e-05,
+      "loss": 0.1119,
+      "step": 17050
+    },
+    {
+      "epoch": 1.9351179673321233,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.0114678631572283e-05,
+      "loss": 0.1133,
+      "step": 17060
+    },
+    {
+      "epoch": 1.9362522686025407,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.0095964845262363e-05,
+      "loss": 0.1116,
+      "step": 17070
+    },
+    {
+      "epoch": 1.9373865698729582,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0077259601548366e-05,
+      "loss": 0.1205,
+      "step": 17080
+    },
+    {
+      "epoch": 1.9385208711433757,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.0058562933013928e-05,
+      "loss": 0.1147,
+      "step": 17090
+    },
+    {
+      "epoch": 1.9396551724137931,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.003987487222776e-05,
+      "loss": 0.1134,
+      "step": 17100
+    },
+    {
+      "epoch": 1.9407894736842106,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.002119545174358e-05,
+      "loss": 0.1027,
+      "step": 17110
+    },
+    {
+      "epoch": 1.941923774954628,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.0002524704100044e-05,
+      "loss": 0.099,
+      "step": 17120
+    },
+    {
+      "epoch": 1.9430580762250453,
+      "grad_norm": 3.375,
+      "learning_rate": 9.983862661820708e-06,
+      "loss": 0.1079,
+      "step": 17130
+    },
+    {
+      "epoch": 1.9441923774954628,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.96520935741396e-06,
+      "loss": 0.1122,
+      "step": 17140
+    },
+    {
+      "epoch": 1.94532667876588,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.946564823372966e-06,
+      "loss": 0.1139,
+      "step": 17150
+    },
+    {
+      "epoch": 1.9464609800362975,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.92792909217563e-06,
+      "loss": 0.1209,
+      "step": 17160
+    },
+    {
+      "epoch": 1.947595281306715,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.909302196284494e-06,
+      "loss": 0.1049,
+      "step": 17170
+    },
+    {
+      "epoch": 1.9487295825771325,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.890684168146736e-06,
+      "loss": 0.1214,
+      "step": 17180
+    },
+    {
+      "epoch": 1.94986388384755,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.872075040194066e-06,
+      "loss": 0.1065,
+      "step": 17190
+    },
+    {
+      "epoch": 1.9509981851179674,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.853474844842707e-06,
+      "loss": 0.1122,
+      "step": 17200
+    },
+    {
+      "epoch": 1.9521324863883849,
+      "grad_norm": 4.0,
+      "learning_rate": 9.834883614493304e-06,
+      "loss": 0.1036,
+      "step": 17210
+    },
+    {
+      "epoch": 1.9532667876588021,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.816301381530906e-06,
+      "loss": 0.11,
+      "step": 17220
+    },
+    {
+      "epoch": 1.9544010889292196,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.797728178324878e-06,
+      "loss": 0.1206,
+      "step": 17230
+    },
+    {
+      "epoch": 1.955535390199637,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.77916403722885e-06,
+      "loss": 0.1071,
+      "step": 17240
+    },
+    {
+      "epoch": 1.9566696914700543,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.760608990580681e-06,
+      "loss": 0.114,
+      "step": 17250
+    },
+    {
+      "epoch": 1.9578039927404718,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.742063070702369e-06,
+      "loss": 0.1157,
+      "step": 17260
+    },
+    {
+      "epoch": 1.9589382940108893,
+      "grad_norm": 3.625,
+      "learning_rate": 9.723526309900037e-06,
+      "loss": 0.1197,
+      "step": 17270
+    },
+    {
+      "epoch": 1.9600725952813067,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.704998740463825e-06,
+      "loss": 0.1131,
+      "step": 17280
+    },
+    {
+      "epoch": 1.9612068965517242,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.68648039466789e-06,
+      "loss": 0.1115,
+      "step": 17290
+    },
+    {
+      "epoch": 1.9623411978221417,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.667971304770305e-06,
+      "loss": 0.114,
+      "step": 17300
+    },
+    {
+      "epoch": 1.9634754990925591,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.64947150301302e-06,
+      "loss": 0.1026,
+      "step": 17310
+    },
+    {
+      "epoch": 1.9646098003629764,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.630981021621823e-06,
+      "loss": 0.1002,
+      "step": 17320
+    },
+    {
+      "epoch": 1.9657441016333939,
+      "grad_norm": 3.625,
+      "learning_rate": 9.612499892806234e-06,
+      "loss": 0.0983,
+      "step": 17330
+    },
+    {
+      "epoch": 1.966878402903811,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.59402814875951e-06,
+      "loss": 0.1035,
+      "step": 17340
+    },
+    {
+      "epoch": 1.9680127041742286,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.575565821658555e-06,
+      "loss": 0.1155,
+      "step": 17350
+    },
+    {
+      "epoch": 1.969147005444646,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.557112943663853e-06,
+      "loss": 0.1143,
+      "step": 17360
+    },
+    {
+      "epoch": 1.9702813067150635,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.538669546919454e-06,
+      "loss": 0.1064,
+      "step": 17370
+    },
+    {
+      "epoch": 1.971415607985481,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.52023566355286e-06,
+      "loss": 0.1037,
+      "step": 17380
+    },
+    {
+      "epoch": 1.9725499092558985,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.501811325675037e-06,
+      "loss": 0.1101,
+      "step": 17390
+    },
+    {
+      "epoch": 1.973684210526316,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.48339656538029e-06,
+      "loss": 0.1139,
+      "step": 17400
+    },
+    {
+      "epoch": 1.9748185117967332,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.46499141474626e-06,
+      "loss": 0.12,
+      "step": 17410
+    },
+    {
+      "epoch": 1.9759528130671506,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.446595905833852e-06,
+      "loss": 0.1056,
+      "step": 17420
+    },
+    {
+      "epoch": 1.977087114337568,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.428210070687154e-06,
+      "loss": 0.1123,
+      "step": 17430
+    },
+    {
+      "epoch": 1.9782214156079854,
+      "grad_norm": 3.125,
+      "learning_rate": 9.409833941333426e-06,
+      "loss": 0.1076,
+      "step": 17440
+    },
+    {
+      "epoch": 1.9793557168784028,
+      "grad_norm": 4.0,
+      "learning_rate": 9.391467549783007e-06,
+      "loss": 0.1176,
+      "step": 17450
+    },
+    {
+      "epoch": 1.9804900181488203,
+      "grad_norm": 3.0,
+      "learning_rate": 9.37311092802928e-06,
+      "loss": 0.1157,
+      "step": 17460
+    },
+    {
+      "epoch": 1.9816243194192378,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.3547641080486e-06,
+      "loss": 0.1053,
+      "step": 17470
+    },
+    {
+      "epoch": 1.9827586206896552,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.336427121800258e-06,
+      "loss": 0.1063,
+      "step": 17480
+    },
+    {
+      "epoch": 1.9838929219600727,
+      "grad_norm": 4.375,
+      "learning_rate": 9.318100001226419e-06,
+      "loss": 0.1156,
+      "step": 17490
+    },
+    {
+      "epoch": 1.98502722323049,
+      "grad_norm": 3.75,
+      "learning_rate": 9.299782778252045e-06,
+      "loss": 0.1096,
+      "step": 17500
+    },
+    {
+      "epoch": 1.9861615245009074,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.281475484784873e-06,
+      "loss": 0.1078,
+      "step": 17510
+    },
+    {
+      "epoch": 1.987295825771325,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.263178152715332e-06,
+      "loss": 0.1096,
+      "step": 17520
+    },
+    {
+      "epoch": 1.9884301270417422,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.244890813916506e-06,
+      "loss": 0.1133,
+      "step": 17530
+    },
+    {
+      "epoch": 1.9895644283121596,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.226613500244075e-06,
+      "loss": 0.11,
+      "step": 17540
+    },
+    {
+      "epoch": 1.990698729582577,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.20834624353624e-06,
+      "loss": 0.1155,
+      "step": 17550
+    },
+    {
+      "epoch": 1.9918330308529946,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.190089075613705e-06,
+      "loss": 0.1065,
+      "step": 17560
+    },
+    {
+      "epoch": 1.992967332123412,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.171842028279575e-06,
+      "loss": 0.1103,
+      "step": 17570
+    },
+    {
+      "epoch": 1.9941016333938295,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.153605133319347e-06,
+      "loss": 0.0966,
+      "step": 17580
+    },
+    {
+      "epoch": 1.995235934664247,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.135378422500816e-06,
+      "loss": 0.1032,
+      "step": 17590
+    },
+    {
+      "epoch": 1.9963702359346642,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.117161927574054e-06,
+      "loss": 0.106,
+      "step": 17600
+    },
+    {
+      "epoch": 1.9975045372050817,
+      "grad_norm": 3.625,
+      "learning_rate": 9.098955680271327e-06,
+      "loss": 0.1231,
+      "step": 17610
+    },
+    {
+      "epoch": 1.998638838475499,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.080759712307047e-06,
+      "loss": 0.1087,
+      "step": 17620
+    },
+    {
+      "epoch": 1.9997731397459164,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.06257405537773e-06,
+      "loss": 0.0979,
+      "step": 17630
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9406647807637907,
+      "eval_f1": 0.939783772842014,
+      "eval_loss": 0.14500758051872253,
+      "eval_precision": 0.9391720119486796,
+      "eval_recall": 0.9406647807637907,
+      "eval_runtime": 122.8041,
+      "eval_samples_per_second": 483.599,
+      "eval_steps_per_second": 3.778,
+      "step": 17632
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 26448,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.925619303619494e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}