diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7156 @@
+{
+  "best_metric": 0.8900473933649289,
+  "best_model_checkpoint": "videomae-large-finetuned-kinetics-finetuned-videomae-large-kitchen/checkpoint-10878",
+  "epoch": 49.02,
+  "eval_steps": 500,
+  "global_step": 11100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.504504504504505e-07,
+      "loss": 3.7012,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.00900900900901e-07,
+      "loss": 3.6854,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 3.7679,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.801801801801802e-06,
+      "loss": 3.7882,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2522522522522524e-06,
+      "loss": 3.7113,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 3.7206,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.153153153153153e-06,
+      "loss": 3.7404,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.603603603603604e-06,
+      "loss": 3.7362,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 3.6743,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.504504504504505e-06,
+      "loss": 3.6672,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954954954954955e-06,
+      "loss": 3.6228,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 3.7014,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.855855855855856e-06,
+      "loss": 3.6683,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.306306306306306e-06,
+      "loss": 3.7223,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 3.6466,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.207207207207208e-06,
+      "loss": 3.6262,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.657657657657658e-06,
+      "loss": 3.6034,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 3.6218,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.558558558558558e-06,
+      "loss": 3.6263,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.00900900900901e-06,
+      "loss": 3.5178,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 3.4525,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.90990990990991e-06,
+      "loss": 3.5158,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "eval_accuracy": 0.058767772511848344,
+      "eval_loss": 3.6067442893981934,
+      "eval_runtime": 248.9798,
+      "eval_samples_per_second": 4.237,
+      "eval_steps_per_second": 2.121,
+      "step": 222
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0360360360360361e-05,
+      "loss": 3.5524,
+      "step": 230
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 3.3292,
+      "step": 240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1261261261261261e-05,
+      "loss": 3.3694,
+      "step": 250
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1711711711711713e-05,
+      "loss": 3.4374,
+      "step": 260
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2162162162162164e-05,
+      "loss": 3.4399,
+      "step": 270
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2612612612612611e-05,
+      "loss": 3.3363,
+      "step": 280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3063063063063064e-05,
+      "loss": 3.2906,
+      "step": 290
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 3.3264,
+      "step": 300
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3963963963963963e-05,
+      "loss": 3.4182,
+      "step": 310
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4414414414414416e-05,
+      "loss": 3.2261,
+      "step": 320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4864864864864867e-05,
+      "loss": 3.052,
+      "step": 330
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5315315315315316e-05,
+      "loss": 3.0721,
+      "step": 340
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5765765765765765e-05,
+      "loss": 2.8128,
+      "step": 350
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 2.9125,
+      "step": 360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 3.2123,
+      "step": 370
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7117117117117117e-05,
+      "loss": 3.0112,
+      "step": 380
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.756756756756757e-05,
+      "loss": 2.7736,
+      "step": 390
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.801801801801802e-05,
+      "loss": 2.8531,
+      "step": 400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.846846846846847e-05,
+      "loss": 3.1244,
+      "step": 410
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 2.7662,
+      "step": 420
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.936936936936937e-05,
+      "loss": 3.0974,
+      "step": 430
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.981981981981982e-05,
+      "loss": 2.8571,
+      "step": 440
+    },
+    {
+      "epoch": 1.02,
+      "eval_accuracy": 0.3014218009478673,
+      "eval_loss": 3.1445391178131104,
+      "eval_runtime": 245.3111,
+      "eval_samples_per_second": 4.301,
+      "eval_steps_per_second": 2.152,
+      "step": 444
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.0270270270270273e-05,
+      "loss": 2.5959,
+      "step": 450
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.0720720720720722e-05,
+      "loss": 2.4969,
+      "step": 460
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.117117117117117e-05,
+      "loss": 2.4735,
+      "step": 470
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1621621621621624e-05,
+      "loss": 2.288,
+      "step": 480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2072072072072073e-05,
+      "loss": 2.35,
+      "step": 490
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2522522522522523e-05,
+      "loss": 2.3779,
+      "step": 500
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2972972972972976e-05,
+      "loss": 2.635,
+      "step": 510
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3423423423423425e-05,
+      "loss": 2.1435,
+      "step": 520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3873873873873874e-05,
+      "loss": 2.5303,
+      "step": 530
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4324324324324327e-05,
+      "loss": 2.2849,
+      "step": 540
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4774774774774777e-05,
+      "loss": 2.387,
+      "step": 550
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5225225225225222e-05,
+      "loss": 2.0807,
+      "step": 560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5675675675675675e-05,
+      "loss": 1.8191,
+      "step": 570
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.6126126126126128e-05,
+      "loss": 2.1925,
+      "step": 580
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.6576576576576577e-05,
+      "loss": 1.9418,
+      "step": 590
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 2.3101,
+      "step": 600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.7477477477477483e-05,
+      "loss": 2.4531,
+      "step": 610
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7927927927927926e-05,
+      "loss": 2.0981,
+      "step": 620
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.8378378378378378e-05,
+      "loss": 2.2399,
+      "step": 630
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.882882882882883e-05,
+      "loss": 1.6239,
+      "step": 640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.927927927927928e-05,
+      "loss": 2.0464,
+      "step": 650
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9729729729729733e-05,
+      "loss": 1.8854,
+      "step": 660
+    },
+    {
+      "epoch": 2.02,
+      "eval_accuracy": 0.4606635071090047,
+      "eval_loss": 2.364382743835449,
+      "eval_runtime": 250.5205,
+      "eval_samples_per_second": 4.211,
+      "eval_steps_per_second": 2.108,
+      "step": 666
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.0180180180180183e-05,
+      "loss": 1.4952,
+      "step": 670
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.063063063063063e-05,
+      "loss": 1.6293,
+      "step": 680
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.108108108108108e-05,
+      "loss": 1.732,
+      "step": 690
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.153153153153153e-05,
+      "loss": 1.3844,
+      "step": 700
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.198198198198199e-05,
+      "loss": 1.7163,
+      "step": 710
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.2432432432432436e-05,
+      "loss": 1.638,
+      "step": 720
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.2882882882882886e-05,
+      "loss": 2.005,
+      "step": 730
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.3738,
+      "step": 740
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.3783783783783784e-05,
+      "loss": 1.5599,
+      "step": 750
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.4234234234234234e-05,
+      "loss": 1.565,
+      "step": 760
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.468468468468469e-05,
+      "loss": 2.0372,
+      "step": 770
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.513513513513514e-05,
+      "loss": 1.8249,
+      "step": 780
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.558558558558558e-05,
+      "loss": 1.3251,
+      "step": 790
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.603603603603604e-05,
+      "loss": 0.9725,
+      "step": 800
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.648648648648649e-05,
+      "loss": 1.4739,
+      "step": 810
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.693693693693694e-05,
+      "loss": 1.2461,
+      "step": 820
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.738738738738739e-05,
+      "loss": 1.4122,
+      "step": 830
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.783783783783784e-05,
+      "loss": 1.5089,
+      "step": 840
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.8288288288288285e-05,
+      "loss": 1.5946,
+      "step": 850
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.873873873873874e-05,
+      "loss": 1.3672,
+      "step": 860
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.918918918918919e-05,
+      "loss": 1.4603,
+      "step": 870
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.963963963963964e-05,
+      "loss": 1.5533,
+      "step": 880
+    },
+    {
+      "epoch": 3.02,
+      "eval_accuracy": 0.5620853080568721,
+      "eval_loss": 1.796697974205017,
+      "eval_runtime": 247.4604,
+      "eval_samples_per_second": 4.263,
+      "eval_steps_per_second": 2.134,
+      "step": 888
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.0090090090090096e-05,
+      "loss": 1.4654,
+      "step": 890
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.0540540540540545e-05,
+      "loss": 1.2046,
+      "step": 900
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.099099099099099e-05,
+      "loss": 1.319,
+      "step": 910
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.1441441441441444e-05,
+      "loss": 1.1716,
+      "step": 920
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.189189189189189e-05,
+      "loss": 0.8349,
+      "step": 930
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.234234234234234e-05,
+      "loss": 1.06,
+      "step": 940
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.27927927927928e-05,
+      "loss": 0.7343,
+      "step": 950
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.324324324324325e-05,
+      "loss": 1.4815,
+      "step": 960
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.369369369369369e-05,
+      "loss": 1.369,
+      "step": 970
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.414414414414415e-05,
+      "loss": 1.3616,
+      "step": 980
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.4594594594594596e-05,
+      "loss": 1.49,
+      "step": 990
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.5045045045045046e-05,
+      "loss": 1.0277,
+      "step": 1000
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.54954954954955e-05,
+      "loss": 0.9602,
+      "step": 1010
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.594594594594595e-05,
+      "loss": 0.8809,
+      "step": 1020
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.6396396396396394e-05,
+      "loss": 1.607,
+      "step": 1030
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.684684684684685e-05,
+      "loss": 1.272,
+      "step": 1040
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 4.72972972972973e-05,
+      "loss": 0.9436,
+      "step": 1050
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 4.774774774774775e-05,
+      "loss": 1.3687,
+      "step": 1060
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 4.8198198198198205e-05,
+      "loss": 0.9016,
+      "step": 1070
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 4.8648648648648654e-05,
+      "loss": 1.0162,
+      "step": 1080
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 4.90990990990991e-05,
+      "loss": 1.2452,
+      "step": 1090
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 4.954954954954955e-05,
+      "loss": 0.9199,
+      "step": 1100
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5e-05,
+      "loss": 1.3935,
+      "step": 1110
+    },
+    {
+      "epoch": 4.02,
+      "eval_accuracy": 0.6502369668246446,
+      "eval_loss": 1.375472068786621,
+      "eval_runtime": 246.8667,
+      "eval_samples_per_second": 4.274,
+      "eval_steps_per_second": 2.139,
+      "step": 1110
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.994994994994995e-05,
+      "loss": 0.6373,
+      "step": 1120
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.98998998998999e-05,
+      "loss": 0.871,
+      "step": 1130
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.984984984984985e-05,
+      "loss": 0.9121,
+      "step": 1140
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.97997997997998e-05,
+      "loss": 0.6519,
+      "step": 1150
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.974974974974975e-05,
+      "loss": 1.1523,
+      "step": 1160
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.96996996996997e-05,
+      "loss": 1.2733,
+      "step": 1170
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.964964964964965e-05,
+      "loss": 1.4579,
+      "step": 1180
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.95995995995996e-05,
+      "loss": 0.8392,
+      "step": 1190
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.954954954954955e-05,
+      "loss": 1.2018,
+      "step": 1200
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.94994994994995e-05,
+      "loss": 1.1601,
+      "step": 1210
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.944944944944945e-05,
+      "loss": 1.0143,
+      "step": 1220
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.93993993993994e-05,
+      "loss": 0.8343,
+      "step": 1230
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.9349349349349347e-05,
+      "loss": 0.8297,
+      "step": 1240
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.92992992992993e-05,
+      "loss": 0.6976,
+      "step": 1250
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.9249249249249253e-05,
+      "loss": 0.7075,
+      "step": 1260
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.9199199199199204e-05,
+      "loss": 0.9578,
+      "step": 1270
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 4.9149149149149154e-05,
+      "loss": 0.8347,
+      "step": 1280
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 4.90990990990991e-05,
+      "loss": 0.8478,
+      "step": 1290
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 4.9049049049049054e-05,
+      "loss": 1.1399,
+      "step": 1300
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 4.8998998998999004e-05,
+      "loss": 0.782,
+      "step": 1310
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 4.8948948948948954e-05,
+      "loss": 0.5463,
+      "step": 1320
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 4.8898898898898904e-05,
+      "loss": 1.1722,
+      "step": 1330
+    },
+    {
+      "epoch": 5.02,
+      "eval_accuracy": 0.7109004739336493,
+      "eval_loss": 1.2232290506362915,
+      "eval_runtime": 250.1719,
+      "eval_samples_per_second": 4.217,
+      "eval_steps_per_second": 2.111,
+      "step": 1332
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.884884884884885e-05,
+      "loss": 1.1194,
+      "step": 1340
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.87987987987988e-05,
+      "loss": 0.6365,
+      "step": 1350
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.8748748748748754e-05,
+      "loss": 1.0437,
+      "step": 1360
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.8698698698698704e-05,
+      "loss": 0.8387,
+      "step": 1370
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.8648648648648654e-05,
+      "loss": 0.7191,
+      "step": 1380
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.85985985985986e-05,
+      "loss": 0.5812,
+      "step": 1390
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.854854854854855e-05,
+      "loss": 1.2726,
+      "step": 1400
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.8498498498498504e-05,
+      "loss": 0.3544,
+      "step": 1410
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.8448448448448455e-05,
+      "loss": 0.6736,
+      "step": 1420
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.83983983983984e-05,
+      "loss": 1.1335,
+      "step": 1430
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.834834834834835e-05,
+      "loss": 0.9236,
+      "step": 1440
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.82982982982983e-05,
+      "loss": 0.3673,
+      "step": 1450
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.824824824824825e-05,
+      "loss": 0.6396,
+      "step": 1460
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.8198198198198205e-05,
+      "loss": 1.0521,
+      "step": 1470
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.7888,
+      "step": 1480
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 4.80980980980981e-05,
+      "loss": 0.8377,
+      "step": 1490
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 4.804804804804805e-05,
+      "loss": 0.6352,
+      "step": 1500
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 4.7997997997998e-05,
+      "loss": 1.33,
+      "step": 1510
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 4.7947947947947955e-05,
+      "loss": 1.0159,
+      "step": 1520
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 4.78978978978979e-05,
+      "loss": 0.5105,
+      "step": 1530
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 4.784784784784785e-05,
+      "loss": 0.7281,
+      "step": 1540
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 4.77977977977978e-05,
+      "loss": 0.2896,
+      "step": 1550
+    },
+    {
+      "epoch": 6.02,
+      "eval_accuracy": 0.6255924170616114,
+      "eval_loss": 1.2858810424804688,
+      "eval_runtime": 244.5305,
+      "eval_samples_per_second": 4.314,
+      "eval_steps_per_second": 2.159,
+      "step": 1554
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.774774774774775e-05,
+      "loss": 0.8307,
+      "step": 1560
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.76976976976977e-05,
+      "loss": 0.8164,
+      "step": 1570
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.764764764764765e-05,
+      "loss": 0.2429,
+      "step": 1580
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.75975975975976e-05,
+      "loss": 0.8349,
+      "step": 1590
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.754754754754755e-05,
+      "loss": 0.7847,
+      "step": 1600
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.74974974974975e-05,
+      "loss": 0.5592,
+      "step": 1610
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.744744744744745e-05,
+      "loss": 1.1295,
+      "step": 1620
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.73973973973974e-05,
+      "loss": 0.4537,
+      "step": 1630
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.734734734734735e-05,
+      "loss": 0.7464,
+      "step": 1640
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.72972972972973e-05,
+      "loss": 0.7417,
+      "step": 1650
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.724724724724725e-05,
+      "loss": 1.1476,
+      "step": 1660
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.71971971971972e-05,
+      "loss": 0.6166,
+      "step": 1670
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.714714714714715e-05,
+      "loss": 1.0584,
+      "step": 1680
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.70970970970971e-05,
+      "loss": 0.8838,
+      "step": 1690
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.704704704704705e-05,
+      "loss": 0.6595,
+      "step": 1700
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.6996996996997e-05,
+      "loss": 1.0565,
+      "step": 1710
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.694694694694695e-05,
+      "loss": 0.9766,
+      "step": 1720
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 4.68968968968969e-05,
+      "loss": 0.5502,
+      "step": 1730
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 4.684684684684685e-05,
+      "loss": 0.868,
+      "step": 1740
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 4.67967967967968e-05,
+      "loss": 0.7801,
+      "step": 1750
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 4.674674674674675e-05,
+      "loss": 0.3967,
+      "step": 1760
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 4.66966966966967e-05,
+      "loss": 0.3166,
+      "step": 1770
+    },
+    {
+      "epoch": 7.02,
+      "eval_accuracy": 0.6720379146919432,
+      "eval_loss": 1.291029453277588,
+      "eval_runtime": 251.3717,
+      "eval_samples_per_second": 4.197,
+      "eval_steps_per_second": 2.1,
+      "step": 1776
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.6646646646646644e-05,
+      "loss": 0.5002,
+      "step": 1780
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.65965965965966e-05,
+      "loss": 0.5258,
+      "step": 1790
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.654654654654655e-05,
+      "loss": 0.1499,
+      "step": 1800
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.64964964964965e-05,
+      "loss": 0.3391,
+      "step": 1810
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.644644644644645e-05,
+      "loss": 0.2682,
+      "step": 1820
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.6396396396396394e-05,
+      "loss": 0.7928,
+      "step": 1830
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.634634634634635e-05,
+      "loss": 0.3695,
+      "step": 1840
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.8298,
+      "step": 1850
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.624624624624625e-05,
+      "loss": 0.5369,
+      "step": 1860
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.61961961961962e-05,
+      "loss": 0.1909,
+      "step": 1870
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.6146146146146144e-05,
+      "loss": 0.6153,
+      "step": 1880
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.6096096096096094e-05,
+      "loss": 0.5458,
+      "step": 1890
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.604604604604605e-05,
+      "loss": 0.3381,
+      "step": 1900
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.5995995995996e-05,
+      "loss": 0.8188,
+      "step": 1910
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.594594594594595e-05,
+      "loss": 0.9806,
+      "step": 1920
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.5895895895895895e-05,
+      "loss": 0.176,
+      "step": 1930
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.5845845845845845e-05,
+      "loss": 0.6741,
+      "step": 1940
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 4.57957957957958e-05,
+      "loss": 0.5312,
+      "step": 1950
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 4.574574574574575e-05,
+      "loss": 0.9062,
+      "step": 1960
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 4.56956956956957e-05,
+      "loss": 0.3351,
+      "step": 1970
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 4.5645645645645645e-05,
+      "loss": 0.2313,
+      "step": 1980
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 4.5595595595595595e-05,
+      "loss": 0.6902,
+      "step": 1990
+    },
+    {
+      "epoch": 8.02,
+      "eval_accuracy": 0.6995260663507109,
+      "eval_loss": 1.270196557044983,
+      "eval_runtime": 250.5107,
+      "eval_samples_per_second": 4.211,
+      "eval_steps_per_second": 2.108,
+      "step": 1998
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.5545545545545545e-05,
+      "loss": 0.5693,
+      "step": 2000
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.54954954954955e-05,
+      "loss": 0.3676,
+      "step": 2010
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.544544544544545e-05,
+      "loss": 0.6409,
+      "step": 2020
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.5395395395395395e-05,
+      "loss": 0.9375,
+      "step": 2030
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.5345345345345345e-05,
+      "loss": 0.2074,
+      "step": 2040
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.5295295295295295e-05,
+      "loss": 0.8579,
+      "step": 2050
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.524524524524525e-05,
+      "loss": 0.8591,
+      "step": 2060
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.5195195195195196e-05,
+      "loss": 0.9443,
+      "step": 2070
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.5145145145145146e-05,
+      "loss": 0.7148,
+      "step": 2080
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.5095095095095096e-05,
+      "loss": 0.3355,
+      "step": 2090
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.5045045045045046e-05,
+      "loss": 0.5742,
+      "step": 2100
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.4994994994994996e-05,
+      "loss": 0.7839,
+      "step": 2110
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.4944944944944946e-05,
+      "loss": 0.643,
+      "step": 2120
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.4894894894894896e-05,
+      "loss": 0.5693,
+      "step": 2130
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.4844844844844846e-05,
+      "loss": 0.4774,
+      "step": 2140
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.4794794794794796e-05,
+      "loss": 0.1836,
+      "step": 2150
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.4744744744744746e-05,
+      "loss": 0.6807,
+      "step": 2160
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 4.4694694694694696e-05,
+      "loss": 0.1184,
+      "step": 2170
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 4.4644644644644646e-05,
+      "loss": 0.645,
+      "step": 2180
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 4.4594594594594596e-05,
+      "loss": 0.721,
+      "step": 2190
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 4.4544544544544546e-05,
+      "loss": 0.4942,
+      "step": 2200
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 4.4494494494494497e-05,
+      "loss": 0.1889,
+      "step": 2210
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.4193,
+      "step": 2220
+    },
+    {
+      "epoch": 9.02,
+      "eval_accuracy": 0.7137440758293839,
+      "eval_loss": 1.208659291267395,
+      "eval_runtime": 251.2654,
+      "eval_samples_per_second": 4.199,
+      "eval_steps_per_second": 2.101,
+      "step": 2220
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4.43943943943944e-05,
+      "loss": 0.4347,
+      "step": 2230
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4.434434434434435e-05,
+      "loss": 0.4797,
+      "step": 2240
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4.42942942942943e-05,
+      "loss": 0.225,
+      "step": 2250
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4.424424424424425e-05,
+      "loss": 0.3253,
+      "step": 2260
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4.41941941941942e-05,
+      "loss": 0.7713,
+      "step": 2270
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.414414414414415e-05,
+      "loss": 0.9326,
+      "step": 2280
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.40940940940941e-05,
+      "loss": 0.5887,
+      "step": 2290
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.404404404404405e-05,
+      "loss": 1.0222,
+      "step": 2300
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.3993993993994e-05,
+      "loss": 0.5595,
+      "step": 2310
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.394394394394394e-05,
+      "loss": 0.2286,
+      "step": 2320
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.38938938938939e-05,
+      "loss": 0.6498,
+      "step": 2330
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.384384384384385e-05,
+      "loss": 0.7378,
+      "step": 2340
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.37937937937938e-05,
+      "loss": 0.3545,
+      "step": 2350
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.374374374374375e-05,
+      "loss": 0.875,
+      "step": 2360
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.369369369369369e-05,
+      "loss": 0.5076,
+      "step": 2370
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 4.364364364364365e-05,
+      "loss": 1.0156,
+      "step": 2380
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 4.35935935935936e-05,
+      "loss": 1.0267,
+      "step": 2390
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 4.354354354354355e-05,
+      "loss": 0.313,
+      "step": 2400
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 4.34934934934935e-05,
+      "loss": 0.5326,
+      "step": 2410
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 4.344344344344344e-05,
+      "loss": 0.5772,
+      "step": 2420
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 4.339339339339339e-05,
+      "loss": 0.1869,
+      "step": 2430
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 4.334334334334335e-05,
+      "loss": 0.1889,
+      "step": 2440
+    },
+    {
+      "epoch": 10.02,
+      "eval_accuracy": 0.7611374407582938,
+      "eval_loss": 1.0499961376190186,
+      "eval_runtime": 250.2797,
+      "eval_samples_per_second": 4.215,
+      "eval_steps_per_second": 2.11,
+      "step": 2442
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 4.32932932932933e-05,
+      "loss": 0.3387,
+      "step": 2450
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 4.324324324324325e-05,
+      "loss": 0.5305,
+      "step": 2460
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 4.319319319319319e-05,
+      "loss": 0.2868,
+      "step": 2470
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 4.314314314314314e-05,
+      "loss": 0.8214,
+      "step": 2480
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 4.30930930930931e-05,
+      "loss": 0.8556,
+      "step": 2490
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.304304304304305e-05,
+      "loss": 0.7684,
+      "step": 2500
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.2992992992993e-05,
+      "loss": 0.6128,
+      "step": 2510
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.294294294294294e-05,
+      "loss": 0.7031,
+      "step": 2520
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.289289289289289e-05,
+      "loss": 1.0707,
+      "step": 2530
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.284284284284284e-05,
+      "loss": 1.0715,
+      "step": 2540
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.27927927927928e-05,
+      "loss": 0.1988,
+      "step": 2550
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.274274274274275e-05,
+      "loss": 0.3882,
+      "step": 2560
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.269269269269269e-05,
+      "loss": 0.2426,
+      "step": 2570
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.264264264264264e-05,
+      "loss": 0.4747,
+      "step": 2580
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.4644,
+      "step": 2590
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.254254254254255e-05,
+      "loss": 0.2122,
+      "step": 2600
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 4.24924924924925e-05,
+      "loss": 0.0621,
+      "step": 2610
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 4.244244244244244e-05,
+      "loss": 0.3805,
+      "step": 2620
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 4.239239239239239e-05,
+      "loss": 1.1612,
+      "step": 2630
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 4.234234234234234e-05,
+      "loss": 0.3846,
+      "step": 2640
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 4.229229229229229e-05,
+      "loss": 0.153,
+      "step": 2650
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 4.224224224224225e-05,
+      "loss": 0.4502,
+      "step": 2660
+    },
+    {
+      "epoch": 11.02,
+      "eval_accuracy": 0.7118483412322275,
+      "eval_loss": 1.1647076606750488,
+      "eval_runtime": 251.766,
+      "eval_samples_per_second": 4.19,
+      "eval_steps_per_second": 2.097,
+      "step": 2664
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.219219219219219e-05,
+      "loss": 0.0562,
+      "step": 2670
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.214214214214214e-05,
+      "loss": 0.1186,
+      "step": 2680
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.209209209209209e-05,
+      "loss": 0.8681,
+      "step": 2690
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.204204204204204e-05,
+      "loss": 0.2302,
+      "step": 2700
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.199199199199199e-05,
+      "loss": 0.8436,
+      "step": 2710
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.194194194194194e-05,
+      "loss": 0.6694,
+      "step": 2720
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.189189189189189e-05,
+      "loss": 0.6777,
+      "step": 2730
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1841841841841843e-05,
+      "loss": 0.6254,
+      "step": 2740
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1791791791791793e-05,
+      "loss": 0.4268,
+      "step": 2750
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1741741741741744e-05,
+      "loss": 0.6122,
+      "step": 2760
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1691691691691694e-05,
+      "loss": 0.1103,
+      "step": 2770
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1641641641641644e-05,
+      "loss": 0.0529,
+      "step": 2780
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1591591591591594e-05,
+      "loss": 0.4995,
+      "step": 2790
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1541541541541544e-05,
+      "loss": 0.2944,
+      "step": 2800
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1491491491491494e-05,
+      "loss": 0.6641,
+      "step": 2810
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1441441441441444e-05,
+      "loss": 0.0549,
+      "step": 2820
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.1391391391391394e-05,
+      "loss": 0.5331,
+      "step": 2830
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 4.1341341341341344e-05,
+      "loss": 0.7761,
+      "step": 2840
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 4.1291291291291294e-05,
+      "loss": 0.0295,
+      "step": 2850
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 4.124124124124124e-05,
+      "loss": 0.833,
+      "step": 2860
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 4.1191191191191194e-05,
+      "loss": 0.7441,
+      "step": 2870
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 4.1141141141141144e-05,
+      "loss": 0.7703,
+      "step": 2880
+    },
+    {
+      "epoch": 12.02,
+      "eval_accuracy": 0.7241706161137441,
+      "eval_loss": 1.1037365198135376,
+      "eval_runtime": 251.3045,
+      "eval_samples_per_second": 4.198,
+      "eval_steps_per_second": 2.101,
+      "step": 2886
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 4.1091091091091094e-05,
+      "loss": 0.5453,
+      "step": 2890
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 4.1041041041041045e-05,
+      "loss": 0.6633,
+      "step": 2900
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 4.099099099099099e-05,
+      "loss": 0.8383,
+      "step": 2910
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 4.0940940940940945e-05,
+      "loss": 1.0613,
+      "step": 2920
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 4.0890890890890895e-05,
+      "loss": 0.4538,
+      "step": 2930
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 4.0840840840840845e-05,
+      "loss": 0.6571,
+      "step": 2940
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.0790790790790795e-05,
+      "loss": 0.1228,
+      "step": 2950
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.0441,
+      "step": 2960
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.069069069069069e-05,
+      "loss": 1.2068,
+      "step": 2970
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.0640640640640645e-05,
+      "loss": 0.256,
+      "step": 2980
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.0590590590590595e-05,
+      "loss": 0.2736,
+      "step": 2990
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.0540540540540545e-05,
+      "loss": 0.3077,
+      "step": 3000
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.049049049049049e-05,
+      "loss": 0.4051,
+      "step": 3010
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.044044044044044e-05,
+      "loss": 0.1085,
+      "step": 3020
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.0390390390390395e-05,
+      "loss": 0.1995,
+      "step": 3030
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.0340340340340346e-05,
+      "loss": 0.0099,
+      "step": 3040
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 4.0290290290290296e-05,
+      "loss": 0.6994,
+      "step": 3050
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 4.024024024024024e-05,
+      "loss": 0.1602,
+      "step": 3060
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 4.019019019019019e-05,
+      "loss": 0.0428,
+      "step": 3070
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 4.014014014014014e-05,
+      "loss": 0.4445,
+      "step": 3080
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 4.0090090090090096e-05,
+      "loss": 0.7188,
+      "step": 3090
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 4.0040040040040046e-05,
+      "loss": 0.0957,
+      "step": 3100
+    },
+    {
+      "epoch": 13.02,
+      "eval_accuracy": 0.7706161137440758,
+      "eval_loss": 1.096709966659546,
+      "eval_runtime": 252.2941,
+      "eval_samples_per_second": 4.182,
+      "eval_steps_per_second": 2.093,
+      "step": 3108
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.998998998998999e-05,
+      "loss": 0.3589,
+      "step": 3110
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.993993993993994e-05,
+      "loss": 0.4206,
+      "step": 3120
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.988988988988989e-05,
+      "loss": 0.4723,
+      "step": 3130
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.9839839839839846e-05,
+      "loss": 0.8298,
+      "step": 3140
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.9789789789789796e-05,
+      "loss": 0.1881,
+      "step": 3150
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.973973973973974e-05,
+      "loss": 0.1694,
+      "step": 3160
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.968968968968969e-05,
+      "loss": 0.8879,
+      "step": 3170
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.963963963963964e-05,
+      "loss": 0.0839,
+      "step": 3180
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.958958958958959e-05,
+      "loss": 0.4495,
+      "step": 3190
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.953953953953955e-05,
+      "loss": 0.4337,
+      "step": 3200
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.948948948948949e-05,
+      "loss": 0.2326,
+      "step": 3210
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.943943943943944e-05,
+      "loss": 0.2812,
+      "step": 3220
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.938938938938939e-05,
+      "loss": 0.8772,
+      "step": 3230
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.933933933933934e-05,
+      "loss": 0.2491,
+      "step": 3240
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.92892892892893e-05,
+      "loss": 0.0997,
+      "step": 3250
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.923923923923924e-05,
+      "loss": 0.1355,
+      "step": 3260
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.918918918918919e-05,
+      "loss": 0.0281,
+      "step": 3270
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 3.913913913913914e-05,
+      "loss": 0.3556,
+      "step": 3280
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 3.908908908908909e-05,
+      "loss": 0.7046,
+      "step": 3290
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 3.903903903903904e-05,
+      "loss": 0.6311,
+      "step": 3300
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 3.898898898898899e-05,
+      "loss": 1.3494,
+      "step": 3310
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 3.893893893893894e-05,
+      "loss": 0.3301,
+      "step": 3320
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.3202,
+      "step": 3330
+    },
+    {
+      "epoch": 14.02,
+      "eval_accuracy": 0.7545023696682465,
+      "eval_loss": 1.0478529930114746,
+      "eval_runtime": 245.4987,
+      "eval_samples_per_second": 4.297,
+      "eval_steps_per_second": 2.151,
+      "step": 3330
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.883883883883884e-05,
+      "loss": 0.2185,
+      "step": 3340
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.878878878878879e-05,
+      "loss": 0.969,
+      "step": 3350
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.873873873873874e-05,
+      "loss": 0.0583,
+      "step": 3360
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.868868868868869e-05,
+      "loss": 0.6673,
+      "step": 3370
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.863863863863864e-05,
+      "loss": 0.5023,
+      "step": 3380
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.858858858858859e-05,
+      "loss": 0.2498,
+      "step": 3390
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.8538538538538534e-05,
+      "loss": 0.2372,
+      "step": 3400
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.848848848848849e-05,
+      "loss": 1.0618,
+      "step": 3410
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.843843843843844e-05,
+      "loss": 0.4096,
+      "step": 3420
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.838838838838839e-05,
+      "loss": 0.358,
+      "step": 3430
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.833833833833834e-05,
+      "loss": 0.1354,
+      "step": 3440
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.8288288288288285e-05,
+      "loss": 0.2454,
+      "step": 3450
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.823823823823824e-05,
+      "loss": 0.3936,
+      "step": 3460
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.818818818818819e-05,
+      "loss": 0.5306,
+      "step": 3470
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.813813813813814e-05,
+      "loss": 0.058,
+      "step": 3480
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 3.808808808808809e-05,
+      "loss": 0.1881,
+      "step": 3490
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 3.8038038038038035e-05,
+      "loss": 0.6403,
+      "step": 3500
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 3.7987987987987985e-05,
+      "loss": 0.2682,
+      "step": 3510
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 3.793793793793794e-05,
+      "loss": 0.1623,
+      "step": 3520
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 3.788788788788789e-05,
+      "loss": 0.2505,
+      "step": 3530
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 3.783783783783784e-05,
+      "loss": 0.1224,
+      "step": 3540
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 3.7787787787787786e-05,
+      "loss": 0.3634,
+      "step": 3550
+    },
+    {
+      "epoch": 15.02,
+      "eval_accuracy": 0.8056872037914692,
+      "eval_loss": 1.0713841915130615,
+      "eval_runtime": 256.4525,
+      "eval_samples_per_second": 4.114,
+      "eval_steps_per_second": 2.059,
+      "step": 3552
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 3.7737737737737736e-05,
+      "loss": 0.6577,
+      "step": 3560
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 3.768768768768769e-05,
+      "loss": 0.148,
+      "step": 3570
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 3.763763763763764e-05,
+      "loss": 0.3399,
+      "step": 3580
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 3.758758758758759e-05,
+      "loss": 0.421,
+      "step": 3590
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 3.7537537537537536e-05,
+      "loss": 0.5354,
+      "step": 3600
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.7487487487487486e-05,
+      "loss": 0.7543,
+      "step": 3610
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.7437437437437436e-05,
+      "loss": 0.1728,
+      "step": 3620
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.738738738738739e-05,
+      "loss": 0.3597,
+      "step": 3630
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.733733733733734e-05,
+      "loss": 0.5684,
+      "step": 3640
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.7287287287287286e-05,
+      "loss": 0.4652,
+      "step": 3650
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.7237237237237236e-05,
+      "loss": 0.1103,
+      "step": 3660
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.7187187187187186e-05,
+      "loss": 0.6614,
+      "step": 3670
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.713713713713714e-05,
+      "loss": 0.7315,
+      "step": 3680
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.708708708708709e-05,
+      "loss": 0.3703,
+      "step": 3690
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.465,
+      "step": 3700
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 3.698698698698699e-05,
+      "loss": 0.0523,
+      "step": 3710
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 3.693693693693694e-05,
+      "loss": 0.8318,
+      "step": 3720
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 3.688688688688689e-05,
+      "loss": 0.3518,
+      "step": 3730
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 3.6836836836836844e-05,
+      "loss": 0.0482,
+      "step": 3740
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 3.678678678678679e-05,
+      "loss": 0.2184,
+      "step": 3750
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 3.673673673673674e-05,
+      "loss": 0.4402,
+      "step": 3760
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 3.668668668668669e-05,
+      "loss": 0.3883,
+      "step": 3770
+    },
+    {
+      "epoch": 16.02,
+      "eval_accuracy": 0.7497630331753554,
+      "eval_loss": 1.2322851419448853,
+      "eval_runtime": 255.5358,
+      "eval_samples_per_second": 4.129,
+      "eval_steps_per_second": 2.066,
+      "step": 3774
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 3.663663663663664e-05,
+      "loss": 0.4998,
+      "step": 3780
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 3.6586586586586594e-05,
+      "loss": 0.1621,
+      "step": 3790
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 3.653653653653654e-05,
+      "loss": 0.1899,
+      "step": 3800
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 3.648648648648649e-05,
+      "loss": 0.3906,
+      "step": 3810
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 3.643643643643644e-05,
+      "loss": 0.0129,
+      "step": 3820
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.638638638638639e-05,
+      "loss": 0.4866,
+      "step": 3830
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.633633633633634e-05,
+      "loss": 0.4291,
+      "step": 3840
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.628628628628629e-05,
+      "loss": 0.3115,
+      "step": 3850
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.623623623623624e-05,
+      "loss": 0.4325,
+      "step": 3860
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.618618618618619e-05,
+      "loss": 0.949,
+      "step": 3870
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.613613613613614e-05,
+      "loss": 0.7044,
+      "step": 3880
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.608608608608609e-05,
+      "loss": 0.2485,
+      "step": 3890
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.603603603603604e-05,
+      "loss": 0.478,
+      "step": 3900
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.598598598598599e-05,
+      "loss": 0.2318,
+      "step": 3910
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.593593593593594e-05,
+      "loss": 0.309,
+      "step": 3920
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.588588588588589e-05,
+      "loss": 0.2111,
+      "step": 3930
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.583583583583583e-05,
+      "loss": 0.5333,
+      "step": 3940
+    },
+    {
+      "epoch": 17.02,
+      "learning_rate": 3.578578578578579e-05,
+      "loss": 0.0069,
+      "step": 3950
+    },
+    {
+      "epoch": 17.02,
+      "learning_rate": 3.573573573573574e-05,
+      "loss": 0.3088,
+      "step": 3960
+    },
+    {
+      "epoch": 17.02,
+      "learning_rate": 3.568568568568569e-05,
+      "loss": 0.0138,
+      "step": 3970
+    },
+    {
+      "epoch": 17.02,
+      "learning_rate": 3.563563563563564e-05,
+      "loss": 0.1217,
+      "step": 3980
+    },
+    {
+      "epoch": 17.02,
+      "learning_rate": 3.558558558558558e-05,
+      "loss": 0.0322,
+      "step": 3990
+    },
+    {
+      "epoch": 17.02,
+      "eval_accuracy": 0.7848341232227488,
+      "eval_loss": 1.0504473447799683,
+      "eval_runtime": 248.5352,
+      "eval_samples_per_second": 4.245,
+      "eval_steps_per_second": 2.124,
+      "step": 3996
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 3.553553553553554e-05,
+      "loss": 0.1216,
+      "step": 4000
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 3.548548548548549e-05,
+      "loss": 0.277,
+      "step": 4010
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 3.543543543543544e-05,
+      "loss": 0.1433,
+      "step": 4020
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 3.538538538538539e-05,
+      "loss": 0.0116,
+      "step": 4030
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 3.533533533533533e-05,
+      "loss": 0.2064,
+      "step": 4040
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 3.528528528528528e-05,
+      "loss": 0.0396,
+      "step": 4050
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.523523523523524e-05,
+      "loss": 0.0992,
+      "step": 4060
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.3158,
+      "step": 4070
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.513513513513514e-05,
+      "loss": 0.4886,
+      "step": 4080
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.508508508508508e-05,
+      "loss": 0.3274,
+      "step": 4090
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.503503503503503e-05,
+      "loss": 0.2986,
+      "step": 4100
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.498498498498499e-05,
+      "loss": 0.5,
+      "step": 4110
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.493493493493494e-05,
+      "loss": 0.5617,
+      "step": 4120
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.488488488488489e-05,
+      "loss": 0.155,
+      "step": 4130
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.483483483483483e-05,
+      "loss": 0.136,
+      "step": 4140
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.478478478478478e-05,
+      "loss": 0.328,
+      "step": 4150
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 3.473473473473473e-05,
+      "loss": 0.5026,
+      "step": 4160
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 3.468468468468469e-05,
+      "loss": 0.1072,
+      "step": 4170
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 3.463463463463464e-05,
+      "loss": 0.0615,
+      "step": 4180
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 3.458458458458458e-05,
+      "loss": 0.1075,
+      "step": 4190
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 3.453453453453453e-05,
+      "loss": 0.1155,
+      "step": 4200
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 3.448448448448448e-05,
+      "loss": 0.5108,
+      "step": 4210
+    },
+    {
+      "epoch": 18.02,
+      "eval_accuracy": 0.7914691943127962,
+      "eval_loss": 1.1356470584869385,
+      "eval_runtime": 247.1171,
+      "eval_samples_per_second": 4.269,
+      "eval_steps_per_second": 2.137,
+      "step": 4218
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 3.443443443443444e-05,
+      "loss": 0.2598,
+      "step": 4220
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 3.438438438438439e-05,
+      "loss": 0.231,
+      "step": 4230
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 3.4334334334334334e-05,
+      "loss": 0.0449,
+      "step": 4240
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 3.4284284284284284e-05,
+      "loss": 0.4002,
+      "step": 4250
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 3.4234234234234234e-05,
+      "loss": 0.1483,
+      "step": 4260
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 3.4184184184184184e-05,
+      "loss": 0.1712,
+      "step": 4270
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.413413413413414e-05,
+      "loss": 0.0387,
+      "step": 4280
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.4084084084084084e-05,
+      "loss": 0.4603,
+      "step": 4290
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.4034034034034034e-05,
+      "loss": 0.2635,
+      "step": 4300
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.3983983983983984e-05,
+      "loss": 0.2796,
+      "step": 4310
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.3933933933933934e-05,
+      "loss": 0.3961,
+      "step": 4320
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.388388388388389e-05,
+      "loss": 0.114,
+      "step": 4330
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.3833833833833834e-05,
+      "loss": 0.0239,
+      "step": 4340
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.3783783783783784e-05,
+      "loss": 0.0925,
+      "step": 4350
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.3733733733733734e-05,
+      "loss": 0.0064,
+      "step": 4360
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.3683683683683684e-05,
+      "loss": 0.0837,
+      "step": 4370
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 3.3633633633633635e-05,
+      "loss": 0.777,
+      "step": 4380
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 3.3583583583583585e-05,
+      "loss": 0.3358,
+      "step": 4390
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 3.3533533533533535e-05,
+      "loss": 0.1277,
+      "step": 4400
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 3.3483483483483485e-05,
+      "loss": 0.19,
+      "step": 4410
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 3.3433433433433435e-05,
+      "loss": 0.024,
+      "step": 4420
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 3.3383383383383385e-05,
+      "loss": 0.1083,
+      "step": 4430
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.309,
+      "step": 4440
+    },
+    {
+      "epoch": 19.02,
+      "eval_accuracy": 0.7592417061611374,
+      "eval_loss": 1.1408809423446655,
+      "eval_runtime": 253.4268,
+      "eval_samples_per_second": 4.163,
+      "eval_steps_per_second": 2.083,
+      "step": 4440
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 3.3283283283283285e-05,
+      "loss": 0.4995,
+      "step": 4450
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 3.3233233233233235e-05,
+      "loss": 0.0753,
+      "step": 4460
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 3.3183183183183185e-05,
+      "loss": 0.3524,
+      "step": 4470
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 3.3133133133133135e-05,
+      "loss": 0.4853,
+      "step": 4480
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 3.3083083083083085e-05,
+      "loss": 0.4007,
+      "step": 4490
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.3033033033033035e-05,
+      "loss": 0.0571,
+      "step": 4500
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.2982982982982985e-05,
+      "loss": 0.0153,
+      "step": 4510
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.2932932932932935e-05,
+      "loss": 0.1357,
+      "step": 4520
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.2882882882882886e-05,
+      "loss": 0.4965,
+      "step": 4530
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.2832832832832836e-05,
+      "loss": 0.0903,
+      "step": 4540
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.2782782782782786e-05,
+      "loss": 0.1452,
+      "step": 4550
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.2732732732732736e-05,
+      "loss": 0.1465,
+      "step": 4560
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.2682682682682686e-05,
+      "loss": 0.4399,
+      "step": 4570
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.263263263263263e-05,
+      "loss": 0.5114,
+      "step": 4580
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.258258258258258e-05,
+      "loss": 0.4951,
+      "step": 4590
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.2532532532532536e-05,
+      "loss": 0.4028,
+      "step": 4600
+    },
+    {
+      "epoch": 20.02,
+      "learning_rate": 3.2482482482482486e-05,
+      "loss": 0.5305,
+      "step": 4610
+    },
+    {
+      "epoch": 20.02,
+      "learning_rate": 3.2432432432432436e-05,
+      "loss": 0.4779,
+      "step": 4620
+    },
+    {
+      "epoch": 20.02,
+      "learning_rate": 3.238238238238238e-05,
+      "loss": 0.0929,
+      "step": 4630
+    },
+    {
+      "epoch": 20.02,
+      "learning_rate": 3.233233233233233e-05,
+      "loss": 0.3879,
+      "step": 4640
+    },
+    {
+      "epoch": 20.02,
+      "learning_rate": 3.2282282282282286e-05,
+      "loss": 0.6666,
+      "step": 4650
+    },
+    {
+      "epoch": 20.02,
+      "learning_rate": 3.2232232232232236e-05,
+      "loss": 0.56,
+      "step": 4660
+    },
+    {
+      "epoch": 20.02,
+      "eval_accuracy": 0.7914691943127962,
+      "eval_loss": 1.082774043083191,
+      "eval_runtime": 243.1084,
+      "eval_samples_per_second": 4.34,
+      "eval_steps_per_second": 2.172,
+      "step": 4662
+    },
+    {
+      "epoch": 21.0,
+      "learning_rate": 3.2182182182182187e-05,
+      "loss": 0.6778,
+      "step": 4670
+    },
+    {
+      "epoch": 21.0,
+      "learning_rate": 3.213213213213213e-05,
+      "loss": 0.3136,
+      "step": 4680
+    },
+    {
+      "epoch": 21.0,
+      "learning_rate": 3.208208208208208e-05,
+      "loss": 0.0267,
+      "step": 4690
+    },
+    {
+      "epoch": 21.0,
+      "learning_rate": 3.203203203203203e-05,
+      "loss": 0.929,
+      "step": 4700
+    },
+    {
+      "epoch": 21.0,
+      "learning_rate": 3.198198198198199e-05,
+      "loss": 0.3136,
+      "step": 4710
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.193193193193194e-05,
+      "loss": 0.3475,
+      "step": 4720
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.188188188188188e-05,
+      "loss": 0.3227,
+      "step": 4730
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.183183183183183e-05,
+      "loss": 0.3799,
+      "step": 4740
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.178178178178178e-05,
+      "loss": 0.0775,
+      "step": 4750
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.173173173173174e-05,
+      "loss": 0.5132,
+      "step": 4760
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.168168168168169e-05,
+      "loss": 0.4308,
+      "step": 4770
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.163163163163163e-05,
+      "loss": 0.9897,
+      "step": 4780
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.158158158158158e-05,
+      "loss": 0.1667,
+      "step": 4790
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.153153153153153e-05,
+      "loss": 0.0132,
+      "step": 4800
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.0172,
+      "step": 4810
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 3.143143143143144e-05,
+      "loss": 0.1472,
+      "step": 4820
+    },
+    {
+      "epoch": 21.02,
+      "learning_rate": 3.138138138138138e-05,
+      "loss": 0.1148,
+      "step": 4830
+    },
+    {
+      "epoch": 21.02,
+      "learning_rate": 3.133133133133133e-05,
+      "loss": 0.209,
+      "step": 4840
+    },
+    {
+      "epoch": 21.02,
+      "learning_rate": 3.128128128128128e-05,
+      "loss": 0.3609,
+      "step": 4850
+    },
+    {
+      "epoch": 21.02,
+      "learning_rate": 3.123123123123123e-05,
+      "loss": 0.4272,
+      "step": 4860
+    },
+    {
+      "epoch": 21.02,
+      "learning_rate": 3.118118118118119e-05,
+      "loss": 0.2125,
+      "step": 4870
+    },
+    {
+      "epoch": 21.02,
+      "learning_rate": 3.113113113113113e-05,
+      "loss": 0.3675,
+      "step": 4880
+    },
+    {
+      "epoch": 21.02,
+      "eval_accuracy": 0.8123222748815165,
+      "eval_loss": 0.9153758883476257,
+      "eval_runtime": 248.3053,
+      "eval_samples_per_second": 4.249,
+      "eval_steps_per_second": 2.126,
+      "step": 4884
+    },
+    {
+      "epoch": 22.0,
+      "learning_rate": 3.108108108108108e-05,
+      "loss": 0.268,
+      "step": 4890
+    },
+    {
+      "epoch": 22.0,
+      "learning_rate": 3.103103103103103e-05,
+      "loss": 0.0923,
+      "step": 4900
+    },
+    {
+      "epoch": 22.0,
+      "learning_rate": 3.098098098098098e-05,
+      "loss": 0.4398,
+      "step": 4910
+    },
+    {
+      "epoch": 22.0,
+      "learning_rate": 3.093093093093093e-05,
+      "loss": 0.0604,
+      "step": 4920
+    },
+    {
+      "epoch": 22.0,
+      "learning_rate": 3.088088088088088e-05,
+      "loss": 0.5577,
+      "step": 4930
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.083083083083083e-05,
+      "loss": 0.124,
+      "step": 4940
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.078078078078078e-05,
+      "loss": 0.213,
+      "step": 4950
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.073073073073073e-05,
+      "loss": 0.1574,
+      "step": 4960
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.068068068068068e-05,
+      "loss": 0.6631,
+      "step": 4970
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.063063063063063e-05,
+      "loss": 0.3951,
+      "step": 4980
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.058058058058058e-05,
+      "loss": 0.0888,
+      "step": 4990
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.053053053053053e-05,
+      "loss": 0.5769,
+      "step": 5000
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.0480480480480482e-05,
+      "loss": 0.5553,
+      "step": 5010
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.0430430430430436e-05,
+      "loss": 0.0928,
+      "step": 5020
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.038038038038038e-05,
+      "loss": 0.03,
+      "step": 5030
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.0330330330330332e-05,
+      "loss": 0.0306,
+      "step": 5040
+    },
+    {
+      "epoch": 22.01,
+      "learning_rate": 3.0280280280280282e-05,
+      "loss": 0.2021,
+      "step": 5050
+    },
+    {
+      "epoch": 22.02,
+      "learning_rate": 3.0230230230230232e-05,
+      "loss": 0.599,
+      "step": 5060
+    },
+    {
+      "epoch": 22.02,
+      "learning_rate": 3.0180180180180183e-05,
+      "loss": 0.7723,
+      "step": 5070
+    },
+    {
+      "epoch": 22.02,
+      "learning_rate": 3.013013013013013e-05,
+      "loss": 0.345,
+      "step": 5080
+    },
+    {
+      "epoch": 22.02,
+      "learning_rate": 3.008008008008008e-05,
+      "loss": 0.0989,
+      "step": 5090
+    },
+    {
+      "epoch": 22.02,
+      "learning_rate": 3.0030030030030033e-05,
+      "loss": 0.0076,
+      "step": 5100
+    },
+    {
+      "epoch": 22.02,
+      "eval_accuracy": 0.8132701421800947,
+      "eval_loss": 1.0973728895187378,
+      "eval_runtime": 245.203,
+      "eval_samples_per_second": 4.303,
+      "eval_steps_per_second": 2.153,
+      "step": 5106
+    },
+    {
+      "epoch": 23.0,
+      "learning_rate": 2.9979979979979983e-05,
+      "loss": 0.2067,
+      "step": 5110
+    },
+    {
+      "epoch": 23.0,
+      "learning_rate": 2.9929929929929933e-05,
+      "loss": 0.2051,
+      "step": 5120
+    },
+    {
+      "epoch": 23.0,
+      "learning_rate": 2.987987987987988e-05,
+      "loss": 0.1336,
+      "step": 5130
+    },
+    {
+      "epoch": 23.0,
+      "learning_rate": 2.982982982982983e-05,
+      "loss": 0.0022,
+      "step": 5140
+    },
+    {
+      "epoch": 23.0,
+      "learning_rate": 2.9779779779779783e-05,
+      "loss": 0.6114,
+      "step": 5150
+    },
+    {
+      "epoch": 23.0,
+      "learning_rate": 2.9729729729729733e-05,
+      "loss": 0.0087,
+      "step": 5160
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.967967967967968e-05,
+      "loss": 0.0708,
+      "step": 5170
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.0075,
+      "step": 5180
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.957957957957958e-05,
+      "loss": 0.5418,
+      "step": 5190
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.952952952952953e-05,
+      "loss": 0.594,
+      "step": 5200
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.9479479479479484e-05,
+      "loss": 0.4395,
+      "step": 5210
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.9429429429429427e-05,
+      "loss": 0.2405,
+      "step": 5220
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.937937937937938e-05,
+      "loss": 0.1534,
+      "step": 5230
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.932932932932933e-05,
+      "loss": 0.0035,
+      "step": 5240
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.927927927927928e-05,
+      "loss": 0.4587,
+      "step": 5250
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.9229229229229234e-05,
+      "loss": 0.1474,
+      "step": 5260
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 2.9179179179179177e-05,
+      "loss": 0.4108,
+      "step": 5270
+    },
+    {
+      "epoch": 23.02,
+      "learning_rate": 2.912912912912913e-05,
+      "loss": 0.0074,
+      "step": 5280
+    },
+    {
+      "epoch": 23.02,
+      "learning_rate": 2.907907907907908e-05,
+      "loss": 0.2801,
+      "step": 5290
+    },
+    {
+      "epoch": 23.02,
+      "learning_rate": 2.902902902902903e-05,
+      "loss": 0.0016,
+      "step": 5300
+    },
+    {
+      "epoch": 23.02,
+      "learning_rate": 2.897897897897898e-05,
+      "loss": 0.2913,
+      "step": 5310
+    },
+    {
+      "epoch": 23.02,
+      "learning_rate": 2.8928928928928928e-05,
+      "loss": 0.0451,
+      "step": 5320
+    },
+    {
+      "epoch": 23.02,
+      "eval_accuracy": 0.8151658767772512,
+      "eval_loss": 1.0360599756240845,
+      "eval_runtime": 247.1589,
+      "eval_samples_per_second": 4.269,
+      "eval_steps_per_second": 2.136,
+      "step": 5328
+    },
+    {
+      "epoch": 24.0,
+      "learning_rate": 2.8878878878878878e-05,
+      "loss": 0.0749,
+      "step": 5330
+    },
+    {
+      "epoch": 24.0,
+      "learning_rate": 2.882882882882883e-05,
+      "loss": 0.0118,
+      "step": 5340
+    },
+    {
+      "epoch": 24.0,
+      "learning_rate": 2.877877877877878e-05,
+      "loss": 0.0215,
+      "step": 5350
+    },
+    {
+      "epoch": 24.0,
+      "learning_rate": 2.872872872872873e-05,
+      "loss": 0.0017,
+      "step": 5360
+    },
+    {
+      "epoch": 24.0,
+      "learning_rate": 2.8678678678678678e-05,
+      "loss": 0.1047,
+      "step": 5370
+    },
+    {
+      "epoch": 24.0,
+      "learning_rate": 2.8628628628628628e-05,
+      "loss": 0.379,
+      "step": 5380
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.857857857857858e-05,
+      "loss": 0.5577,
+      "step": 5390
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.852852852852853e-05,
+      "loss": 0.4075,
+      "step": 5400
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.847847847847848e-05,
+      "loss": 0.0151,
+      "step": 5410
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.8428428428428428e-05,
+      "loss": 0.2632,
+      "step": 5420
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.8378378378378378e-05,
+      "loss": 0.0274,
+      "step": 5430
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.832832832832833e-05,
+      "loss": 0.4542,
+      "step": 5440
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.8278278278278282e-05,
+      "loss": 0.0043,
+      "step": 5450
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.8228228228228232e-05,
+      "loss": 0.3751,
+      "step": 5460
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.817817817817818e-05,
+      "loss": 0.0041,
+      "step": 5470
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.812812812812813e-05,
+      "loss": 0.1011,
+      "step": 5480
+    },
+    {
+      "epoch": 24.01,
+      "learning_rate": 2.807807807807808e-05,
+      "loss": 0.2128,
+      "step": 5490
+    },
+    {
+      "epoch": 24.02,
+      "learning_rate": 2.8028028028028032e-05,
+      "loss": 0.0048,
+      "step": 5500
+    },
+    {
+      "epoch": 24.02,
+      "learning_rate": 2.7977977977977982e-05,
+      "loss": 0.0063,
+      "step": 5510
+    },
+    {
+      "epoch": 24.02,
+      "learning_rate": 2.7927927927927926e-05,
+      "loss": 0.0029,
+      "step": 5520
+    },
+    {
+      "epoch": 24.02,
+      "learning_rate": 2.787787787787788e-05,
+      "loss": 0.6566,
+      "step": 5530
+    },
+    {
+      "epoch": 24.02,
+      "learning_rate": 2.782782782782783e-05,
+      "loss": 0.1477,
+      "step": 5540
+    },
+    {
+      "epoch": 24.02,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.2558,
+      "step": 5550
+    },
+    {
+      "epoch": 24.02,
+      "eval_accuracy": 0.8236966824644549,
+      "eval_loss": 0.7830020785331726,
+      "eval_runtime": 248.2906,
+      "eval_samples_per_second": 4.249,
+      "eval_steps_per_second": 2.127,
+      "step": 5550
+    },
+    {
+      "epoch": 25.0,
+      "learning_rate": 2.7727727727727733e-05,
+      "loss": 0.6345,
+      "step": 5560
+    },
+    {
+      "epoch": 25.0,
+      "learning_rate": 2.7677677677677676e-05,
+      "loss": 0.4474,
+      "step": 5570
+    },
+    {
+      "epoch": 25.0,
+      "learning_rate": 2.762762762762763e-05,
+      "loss": 0.01,
+      "step": 5580
+    },
+    {
+      "epoch": 25.0,
+      "learning_rate": 2.757757757757758e-05,
+      "loss": 0.1454,
+      "step": 5590
+    },
+    {
+      "epoch": 25.0,
+      "learning_rate": 2.752752752752753e-05,
+      "loss": 0.2356,
+      "step": 5600
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.7477477477477483e-05,
+      "loss": 0.2639,
+      "step": 5610
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.7427427427427426e-05,
+      "loss": 0.0431,
+      "step": 5620
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.7377377377377376e-05,
+      "loss": 0.1564,
+      "step": 5630
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.732732732732733e-05,
+      "loss": 0.1669,
+      "step": 5640
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.727727727727728e-05,
+      "loss": 0.2806,
+      "step": 5650
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.722722722722723e-05,
+      "loss": 0.3531,
+      "step": 5660
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.7177177177177177e-05,
+      "loss": 0.2176,
+      "step": 5670
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.7127127127127127e-05,
+      "loss": 0.155,
+      "step": 5680
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.707707707707708e-05,
+      "loss": 0.0943,
+      "step": 5690
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 0.0181,
+      "step": 5700
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 2.697697697697698e-05,
+      "loss": 0.5348,
+      "step": 5710
+    },
+    {
+      "epoch": 25.02,
+      "learning_rate": 2.6926926926926927e-05,
+      "loss": 1.0172,
+      "step": 5720
+    },
+    {
+      "epoch": 25.02,
+      "learning_rate": 2.6876876876876877e-05,
+      "loss": 0.0848,
+      "step": 5730
+    },
+    {
+      "epoch": 25.02,
+      "learning_rate": 2.6826826826826827e-05,
+      "loss": 0.5269,
+      "step": 5740
+    },
+    {
+      "epoch": 25.02,
+      "learning_rate": 2.677677677677678e-05,
+      "loss": 0.1968,
+      "step": 5750
+    },
+    {
+      "epoch": 25.02,
+      "learning_rate": 2.672672672672673e-05,
+      "loss": 0.0014,
+      "step": 5760
+    },
+    {
+      "epoch": 25.02,
+      "learning_rate": 2.6676676676676677e-05,
+      "loss": 0.0125,
+      "step": 5770
+    },
+    {
+      "epoch": 25.02,
+      "eval_accuracy": 0.8170616113744076,
+      "eval_loss": 0.8727856278419495,
+      "eval_runtime": 245.6983,
+      "eval_samples_per_second": 4.294,
+      "eval_steps_per_second": 2.149,
+      "step": 5772
+    },
+    {
+      "epoch": 26.0,
+      "learning_rate": 2.6626626626626627e-05,
+      "loss": 0.249,
+      "step": 5780
+    },
+    {
+      "epoch": 26.0,
+      "learning_rate": 2.6576576576576577e-05,
+      "loss": 0.0209,
+      "step": 5790
+    },
+    {
+      "epoch": 26.0,
+      "learning_rate": 2.652652652652653e-05,
+      "loss": 0.0565,
+      "step": 5800
+    },
+    {
+      "epoch": 26.0,
+      "learning_rate": 2.6476476476476474e-05,
+      "loss": 0.0247,
+      "step": 5810
+    },
+    {
+      "epoch": 26.0,
+      "learning_rate": 2.6426426426426428e-05,
+      "loss": 0.5166,
+      "step": 5820
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.6376376376376378e-05,
+      "loss": 0.7852,
+      "step": 5830
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.6326326326326328e-05,
+      "loss": 0.4739,
+      "step": 5840
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.6276276276276278e-05,
+      "loss": 0.3672,
+      "step": 5850
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.6226226226226224e-05,
+      "loss": 0.0842,
+      "step": 5860
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.6176176176176175e-05,
+      "loss": 0.0076,
+      "step": 5870
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.6126126126126128e-05,
+      "loss": 0.3222,
+      "step": 5880
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.6076076076076078e-05,
+      "loss": 0.6115,
+      "step": 5890
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.6026026026026028e-05,
+      "loss": 0.0009,
+      "step": 5900
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.5975975975975975e-05,
+      "loss": 0.0648,
+      "step": 5910
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.2629,
+      "step": 5920
+    },
+    {
+      "epoch": 26.01,
+      "learning_rate": 2.587587587587588e-05,
+      "loss": 0.0009,
+      "step": 5930
+    },
+    {
+      "epoch": 26.02,
+      "learning_rate": 2.582582582582583e-05,
+      "loss": 0.4119,
+      "step": 5940
+    },
+    {
+      "epoch": 26.02,
+      "learning_rate": 2.577577577577578e-05,
+      "loss": 0.0849,
+      "step": 5950
+    },
+    {
+      "epoch": 26.02,
+      "learning_rate": 2.5725725725725725e-05,
+      "loss": 0.8525,
+      "step": 5960
+    },
+    {
+      "epoch": 26.02,
+      "learning_rate": 2.5675675675675675e-05,
+      "loss": 0.8865,
+      "step": 5970
+    },
+    {
+      "epoch": 26.02,
+      "learning_rate": 2.5625625625625625e-05,
+      "loss": 0.005,
+      "step": 5980
+    },
+    {
+      "epoch": 26.02,
+      "learning_rate": 2.557557557557558e-05,
+      "loss": 0.4184,
+      "step": 5990
+    },
+    {
+      "epoch": 26.02,
+      "eval_accuracy": 0.8265402843601896,
+      "eval_loss": 0.8413429856300354,
+      "eval_runtime": 245.7335,
+      "eval_samples_per_second": 4.293,
+      "eval_steps_per_second": 2.149,
+      "step": 5994
+    },
+    {
+      "epoch": 27.0,
+      "learning_rate": 2.552552552552553e-05,
+      "loss": 0.0181,
+      "step": 6000
+    },
+    {
+      "epoch": 27.0,
+      "learning_rate": 2.5475475475475476e-05,
+      "loss": 0.0033,
+      "step": 6010
+    },
+    {
+      "epoch": 27.0,
+      "learning_rate": 2.5425425425425426e-05,
+      "loss": 0.4205,
+      "step": 6020
+    },
+    {
+      "epoch": 27.0,
+      "learning_rate": 2.5375375375375376e-05,
+      "loss": 0.1357,
+      "step": 6030
+    },
+    {
+      "epoch": 27.0,
+      "learning_rate": 2.532532532532533e-05,
+      "loss": 0.0352,
+      "step": 6040
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.527527527527528e-05,
+      "loss": 0.1965,
+      "step": 6050
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.5225225225225222e-05,
+      "loss": 0.2975,
+      "step": 6060
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.5175175175175176e-05,
+      "loss": 0.6702,
+      "step": 6070
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.5125125125125126e-05,
+      "loss": 0.6015,
+      "step": 6080
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.5075075075075076e-05,
+      "loss": 0.333,
+      "step": 6090
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.502502502502503e-05,
+      "loss": 0.2907,
+      "step": 6100
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.4974974974974976e-05,
+      "loss": 0.3042,
+      "step": 6110
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.4924924924924926e-05,
+      "loss": 0.5136,
+      "step": 6120
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.4874874874874876e-05,
+      "loss": 0.0567,
+      "step": 6130
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.4824824824824826e-05,
+      "loss": 0.2974,
+      "step": 6140
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.4774774774774777e-05,
+      "loss": 0.0103,
+      "step": 6150
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 2.4724724724724727e-05,
+      "loss": 0.1177,
+      "step": 6160
+    },
+    {
+      "epoch": 27.02,
+      "learning_rate": 2.4674674674674673e-05,
+      "loss": 0.0542,
+      "step": 6170
+    },
+    {
+      "epoch": 27.02,
+      "learning_rate": 2.4624624624624627e-05,
+      "loss": 0.2662,
+      "step": 6180
+    },
+    {
+      "epoch": 27.02,
+      "learning_rate": 2.4574574574574577e-05,
+      "loss": 0.0018,
+      "step": 6190
+    },
+    {
+      "epoch": 27.02,
+      "learning_rate": 2.4524524524524527e-05,
+      "loss": 0.0329,
+      "step": 6200
+    },
+    {
+      "epoch": 27.02,
+      "learning_rate": 2.4474474474474477e-05,
+      "loss": 0.2566,
+      "step": 6210
+    },
+    {
+      "epoch": 27.02,
+      "eval_accuracy": 0.8009478672985783,
+      "eval_loss": 1.0643999576568604,
+      "eval_runtime": 248.0961,
+      "eval_samples_per_second": 4.252,
+      "eval_steps_per_second": 2.128,
+      "step": 6216
+    },
+    {
+      "epoch": 28.0,
+      "learning_rate": 2.4424424424424424e-05,
+      "loss": 0.0261,
+      "step": 6220
+    },
+    {
+      "epoch": 28.0,
+      "learning_rate": 2.4374374374374377e-05,
+      "loss": 0.48,
+      "step": 6230
+    },
+    {
+      "epoch": 28.0,
+      "learning_rate": 2.4324324324324327e-05,
+      "loss": 0.4585,
+      "step": 6240
+    },
+    {
+      "epoch": 28.0,
+      "learning_rate": 2.4274274274274274e-05,
+      "loss": 0.2948,
+      "step": 6250
+    },
+    {
+      "epoch": 28.0,
+      "learning_rate": 2.4224224224224227e-05,
+      "loss": 0.7788,
+      "step": 6260
+    },
+    {
+      "epoch": 28.0,
+      "learning_rate": 2.4174174174174174e-05,
+      "loss": 0.0556,
+      "step": 6270
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.4124124124124124e-05,
+      "loss": 0.0544,
+      "step": 6280
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.0015,
+      "step": 6290
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.4024024024024024e-05,
+      "loss": 0.2303,
+      "step": 6300
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.3973973973973978e-05,
+      "loss": 0.0352,
+      "step": 6310
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.3923923923923924e-05,
+      "loss": 0.5362,
+      "step": 6320
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.3873873873873874e-05,
+      "loss": 0.0233,
+      "step": 6330
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.3823823823823824e-05,
+      "loss": 0.5377,
+      "step": 6340
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.3773773773773775e-05,
+      "loss": 0.1257,
+      "step": 6350
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.3723723723723725e-05,
+      "loss": 0.0522,
+      "step": 6360
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.3673673673673675e-05,
+      "loss": 0.4811,
+      "step": 6370
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 2.3623623623623625e-05,
+      "loss": 0.1918,
+      "step": 6380
+    },
+    {
+      "epoch": 28.02,
+      "learning_rate": 2.3573573573573575e-05,
+      "loss": 0.5281,
+      "step": 6390
+    },
+    {
+      "epoch": 28.02,
+      "learning_rate": 2.3523523523523525e-05,
+      "loss": 0.0989,
+      "step": 6400
+    },
+    {
+      "epoch": 28.02,
+      "learning_rate": 2.3473473473473475e-05,
+      "loss": 0.0942,
+      "step": 6410
+    },
+    {
+      "epoch": 28.02,
+      "learning_rate": 2.3423423423423425e-05,
+      "loss": 0.3984,
+      "step": 6420
+    },
+    {
+      "epoch": 28.02,
+      "learning_rate": 2.3373373373373375e-05,
+      "loss": 0.1257,
+      "step": 6430
+    },
+    {
+      "epoch": 28.02,
+      "eval_accuracy": 0.8265402843601896,
+      "eval_loss": 0.8641291856765747,
+      "eval_runtime": 247.586,
+      "eval_samples_per_second": 4.261,
+      "eval_steps_per_second": 2.133,
+      "step": 6438
+    },
+    {
+      "epoch": 29.0,
+      "learning_rate": 2.3323323323323322e-05,
+      "loss": 0.3926,
+      "step": 6440
+    },
+    {
+      "epoch": 29.0,
+      "learning_rate": 2.3273273273273275e-05,
+      "loss": 0.413,
+      "step": 6450
+    },
+    {
+      "epoch": 29.0,
+      "learning_rate": 2.3223223223223225e-05,
+      "loss": 0.0019,
+      "step": 6460
+    },
+    {
+      "epoch": 29.0,
+      "learning_rate": 2.3173173173173175e-05,
+      "loss": 0.5808,
+      "step": 6470
+    },
+    {
+      "epoch": 29.0,
+      "learning_rate": 2.3123123123123125e-05,
+      "loss": 0.6509,
+      "step": 6480
+    },
+    {
+      "epoch": 29.0,
+      "learning_rate": 2.3073073073073072e-05,
+      "loss": 0.0258,
+      "step": 6490
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.3023023023023026e-05,
+      "loss": 0.2004,
+      "step": 6500
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2972972972972976e-05,
+      "loss": 0.1049,
+      "step": 6510
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2922922922922922e-05,
+      "loss": 0.1371,
+      "step": 6520
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2872872872872876e-05,
+      "loss": 0.0012,
+      "step": 6530
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2822822822822822e-05,
+      "loss": 0.0027,
+      "step": 6540
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2772772772772773e-05,
+      "loss": 0.4307,
+      "step": 6550
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2722722722722726e-05,
+      "loss": 0.3665,
+      "step": 6560
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2672672672672673e-05,
+      "loss": 0.3267,
+      "step": 6570
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2622622622622626e-05,
+      "loss": 0.0028,
+      "step": 6580
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2572572572572573e-05,
+      "loss": 0.475,
+      "step": 6590
+    },
+    {
+      "epoch": 29.01,
+      "learning_rate": 2.2522522522522523e-05,
+      "loss": 0.3067,
+      "step": 6600
+    },
+    {
+      "epoch": 29.02,
+      "learning_rate": 2.2472472472472473e-05,
+      "loss": 0.1204,
+      "step": 6610
+    },
+    {
+      "epoch": 29.02,
+      "learning_rate": 2.2422422422422423e-05,
+      "loss": 0.0012,
+      "step": 6620
+    },
+    {
+      "epoch": 29.02,
+      "learning_rate": 2.2372372372372373e-05,
+      "loss": 0.1838,
+      "step": 6630
+    },
+    {
+      "epoch": 29.02,
+      "learning_rate": 2.2322322322322323e-05,
+      "loss": 0.0094,
+      "step": 6640
+    },
+    {
+      "epoch": 29.02,
+      "learning_rate": 2.2272272272272273e-05,
+      "loss": 0.0009,
+      "step": 6650
+    },
+    {
+      "epoch": 29.02,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.1326,
+      "step": 6660
+    },
+    {
+      "epoch": 29.02,
+      "eval_accuracy": 0.8417061611374408,
+      "eval_loss": 0.8443952798843384,
+      "eval_runtime": 245.0797,
+      "eval_samples_per_second": 4.305,
+      "eval_steps_per_second": 2.154,
+      "step": 6660
+    },
+    {
+      "epoch": 30.0,
+      "learning_rate": 2.2172172172172173e-05,
+      "loss": 0.0767,
+      "step": 6670
+    },
+    {
+      "epoch": 30.0,
+      "learning_rate": 2.2122122122122123e-05,
+      "loss": 0.3648,
+      "step": 6680
+    },
+    {
+      "epoch": 30.0,
+      "learning_rate": 2.2072072072072073e-05,
+      "loss": 0.02,
+      "step": 6690
+    },
+    {
+      "epoch": 30.0,
+      "learning_rate": 2.2022022022022024e-05,
+      "loss": 0.3898,
+      "step": 6700
+    },
+    {
+      "epoch": 30.0,
+      "learning_rate": 2.197197197197197e-05,
+      "loss": 0.2153,
+      "step": 6710
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.1921921921921924e-05,
+      "loss": 0.1903,
+      "step": 6720
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.1871871871871874e-05,
+      "loss": 0.254,
+      "step": 6730
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.1821821821821824e-05,
+      "loss": 0.4832,
+      "step": 6740
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.1771771771771774e-05,
+      "loss": 0.8542,
+      "step": 6750
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.172172172172172e-05,
+      "loss": 0.0256,
+      "step": 6760
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.1671671671671674e-05,
+      "loss": 0.0053,
+      "step": 6770
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.1621621621621624e-05,
+      "loss": 1.0811,
+      "step": 6780
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.157157157157157e-05,
+      "loss": 0.1894,
+      "step": 6790
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.1521521521521524e-05,
+      "loss": 0.0044,
+      "step": 6800
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.147147147147147e-05,
+      "loss": 0.0658,
+      "step": 6810
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 2.142142142142142e-05,
+      "loss": 0.0026,
+      "step": 6820
+    },
+    {
+      "epoch": 30.02,
+      "learning_rate": 2.1371371371371374e-05,
+      "loss": 0.0024,
+      "step": 6830
+    },
+    {
+      "epoch": 30.02,
+      "learning_rate": 2.132132132132132e-05,
+      "loss": 0.0137,
+      "step": 6840
+    },
+    {
+      "epoch": 30.02,
+      "learning_rate": 2.1271271271271275e-05,
+      "loss": 0.9223,
+      "step": 6850
+    },
+    {
+      "epoch": 30.02,
+      "learning_rate": 2.122122122122122e-05,
+      "loss": 0.014,
+      "step": 6860
+    },
+    {
+      "epoch": 30.02,
+      "learning_rate": 2.117117117117117e-05,
+      "loss": 0.3707,
+      "step": 6870
+    },
+    {
+      "epoch": 30.02,
+      "learning_rate": 2.1121121121121125e-05,
+      "loss": 0.0436,
+      "step": 6880
+    },
+    {
+      "epoch": 30.02,
+      "eval_accuracy": 0.8322274881516588,
+      "eval_loss": 0.8615030646324158,
+      "eval_runtime": 242.4706,
+      "eval_samples_per_second": 4.351,
+      "eval_steps_per_second": 2.178,
+      "step": 6882
+    },
+    {
+      "epoch": 31.0,
+      "learning_rate": 2.107107107107107e-05,
+      "loss": 0.23,
+      "step": 6890
+    },
+    {
+      "epoch": 31.0,
+      "learning_rate": 2.102102102102102e-05,
+      "loss": 0.0124,
+      "step": 6900
+    },
+    {
+      "epoch": 31.0,
+      "learning_rate": 2.097097097097097e-05,
+      "loss": 0.0386,
+      "step": 6910
+    },
+    {
+      "epoch": 31.0,
+      "learning_rate": 2.0920920920920922e-05,
+      "loss": 0.4349,
+      "step": 6920
+    },
+    {
+      "epoch": 31.0,
+      "learning_rate": 2.0870870870870872e-05,
+      "loss": 0.0056,
+      "step": 6930
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0820820820820822e-05,
+      "loss": 0.1028,
+      "step": 6940
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0770770770770772e-05,
+      "loss": 0.0028,
+      "step": 6950
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0720720720720722e-05,
+      "loss": 0.7829,
+      "step": 6960
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0670670670670672e-05,
+      "loss": 0.0068,
+      "step": 6970
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.062062062062062e-05,
+      "loss": 0.0489,
+      "step": 6980
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0570570570570572e-05,
+      "loss": 0.2531,
+      "step": 6990
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0520520520520522e-05,
+      "loss": 0.3751,
+      "step": 7000
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0470470470470472e-05,
+      "loss": 0.0199,
+      "step": 7010
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0420420420420422e-05,
+      "loss": 0.2208,
+      "step": 7020
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.2553,
+      "step": 7030
+    },
+    {
+      "epoch": 31.01,
+      "learning_rate": 2.0320320320320323e-05,
+      "loss": 0.3365,
+      "step": 7040
+    },
+    {
+      "epoch": 31.02,
+      "learning_rate": 2.0270270270270273e-05,
+      "loss": 0.3933,
+      "step": 7050
+    },
+    {
+      "epoch": 31.02,
+      "learning_rate": 2.022022022022022e-05,
+      "loss": 0.0046,
+      "step": 7060
+    },
+    {
+      "epoch": 31.02,
+      "learning_rate": 2.0170170170170173e-05,
+      "loss": 0.3302,
+      "step": 7070
+    },
+    {
+      "epoch": 31.02,
+      "learning_rate": 2.012012012012012e-05,
+      "loss": 0.0033,
+      "step": 7080
+    },
+    {
+      "epoch": 31.02,
+      "learning_rate": 2.007007007007007e-05,
+      "loss": 0.0267,
+      "step": 7090
+    },
+    {
+      "epoch": 31.02,
+      "learning_rate": 2.0020020020020023e-05,
+      "loss": 0.0408,
+      "step": 7100
+    },
+    {
+      "epoch": 31.02,
+      "eval_accuracy": 0.833175355450237,
+      "eval_loss": 0.8074727654457092,
+      "eval_runtime": 247.5031,
+      "eval_samples_per_second": 4.263,
+      "eval_steps_per_second": 2.133,
+      "step": 7104
+    },
+    {
+      "epoch": 32.0,
+      "learning_rate": 1.996996996996997e-05,
+      "loss": 0.0056,
+      "step": 7110
+    },
+    {
+      "epoch": 32.0,
+      "learning_rate": 1.9919919919919923e-05,
+      "loss": 0.0013,
+      "step": 7120
+    },
+    {
+      "epoch": 32.0,
+      "learning_rate": 1.986986986986987e-05,
+      "loss": 0.4676,
+      "step": 7130
+    },
+    {
+      "epoch": 32.0,
+      "learning_rate": 1.981981981981982e-05,
+      "loss": 0.227,
+      "step": 7140
+    },
+    {
+      "epoch": 32.0,
+      "learning_rate": 1.9769769769769773e-05,
+      "loss": 0.1267,
+      "step": 7150
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.971971971971972e-05,
+      "loss": 0.2614,
+      "step": 7160
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.966966966966967e-05,
+      "loss": 0.2624,
+      "step": 7170
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.961961961961962e-05,
+      "loss": 0.24,
+      "step": 7180
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.956956956956957e-05,
+      "loss": 0.6802,
+      "step": 7190
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.951951951951952e-05,
+      "loss": 0.377,
+      "step": 7200
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.946946946946947e-05,
+      "loss": 0.0394,
+      "step": 7210
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.941941941941942e-05,
+      "loss": 0.1165,
+      "step": 7220
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.936936936936937e-05,
+      "loss": 0.0021,
+      "step": 7230
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.931931931931932e-05,
+      "loss": 0.2327,
+      "step": 7240
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.9269269269269267e-05,
+      "loss": 0.3187,
+      "step": 7250
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.921921921921922e-05,
+      "loss": 0.1452,
+      "step": 7260
+    },
+    {
+      "epoch": 32.01,
+      "learning_rate": 1.916916916916917e-05,
+      "loss": 0.1138,
+      "step": 7270
+    },
+    {
+      "epoch": 32.02,
+      "learning_rate": 1.911911911911912e-05,
+      "loss": 0.0011,
+      "step": 7280
+    },
+    {
+      "epoch": 32.02,
+      "learning_rate": 1.906906906906907e-05,
+      "loss": 0.4207,
+      "step": 7290
+    },
+    {
+      "epoch": 32.02,
+      "learning_rate": 1.9019019019019018e-05,
+      "loss": 0.1415,
+      "step": 7300
+    },
+    {
+      "epoch": 32.02,
+      "learning_rate": 1.896896896896897e-05,
+      "loss": 0.0101,
+      "step": 7310
+    },
+    {
+      "epoch": 32.02,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 0.0316,
+      "step": 7320
+    },
+    {
+      "epoch": 32.02,
+      "eval_accuracy": 0.8341232227488151,
+      "eval_loss": 0.8699161410331726,
+      "eval_runtime": 247.2353,
+      "eval_samples_per_second": 4.267,
+      "eval_steps_per_second": 2.136,
+      "step": 7326
+    },
+    {
+      "epoch": 33.0,
+      "learning_rate": 1.8868868868868868e-05,
+      "loss": 0.0064,
+      "step": 7330
+    },
+    {
+      "epoch": 33.0,
+      "learning_rate": 1.881881881881882e-05,
+      "loss": 0.1905,
+      "step": 7340
+    },
+    {
+      "epoch": 33.0,
+      "learning_rate": 1.8768768768768768e-05,
+      "loss": 0.2165,
+      "step": 7350
+    },
+    {
+      "epoch": 33.0,
+      "learning_rate": 1.8718718718718718e-05,
+      "loss": 0.0013,
+      "step": 7360
+    },
+    {
+      "epoch": 33.0,
+      "learning_rate": 1.866866866866867e-05,
+      "loss": 0.0683,
+      "step": 7370
+    },
+    {
+      "epoch": 33.0,
+      "learning_rate": 1.8618618618618618e-05,
+      "loss": 0.6117,
+      "step": 7380
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.856856856856857e-05,
+      "loss": 0.0273,
+      "step": 7390
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.0644,
+      "step": 7400
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.846846846846847e-05,
+      "loss": 0.3223,
+      "step": 7410
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.8418418418418422e-05,
+      "loss": 0.1092,
+      "step": 7420
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.836836836836837e-05,
+      "loss": 0.2037,
+      "step": 7430
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.831831831831832e-05,
+      "loss": 0.0145,
+      "step": 7440
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.826826826826827e-05,
+      "loss": 0.5524,
+      "step": 7450
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.821821821821822e-05,
+      "loss": 0.0722,
+      "step": 7460
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.816816816816817e-05,
+      "loss": 0.4103,
+      "step": 7470
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.811811811811812e-05,
+      "loss": 0.1167,
+      "step": 7480
+    },
+    {
+      "epoch": 33.01,
+      "learning_rate": 1.806806806806807e-05,
+      "loss": 0.0438,
+      "step": 7490
+    },
+    {
+      "epoch": 33.02,
+      "learning_rate": 1.801801801801802e-05,
+      "loss": 0.2624,
+      "step": 7500
+    },
+    {
+      "epoch": 33.02,
+      "learning_rate": 1.796796796796797e-05,
+      "loss": 0.2069,
+      "step": 7510
+    },
+    {
+      "epoch": 33.02,
+      "learning_rate": 1.7917917917917916e-05,
+      "loss": 0.0942,
+      "step": 7520
+    },
+    {
+      "epoch": 33.02,
+      "learning_rate": 1.786786786786787e-05,
+      "loss": 0.0071,
+      "step": 7530
+    },
+    {
+      "epoch": 33.02,
+      "learning_rate": 1.781781781781782e-05,
+      "loss": 0.2235,
+      "step": 7540
+    },
+    {
+      "epoch": 33.02,
+      "eval_accuracy": 0.8454976303317535,
+      "eval_loss": 0.8150666952133179,
+      "eval_runtime": 242.6132,
+      "eval_samples_per_second": 4.348,
+      "eval_steps_per_second": 2.176,
+      "step": 7548
+    },
+    {
+      "epoch": 34.0,
+      "learning_rate": 1.776776776776777e-05,
+      "loss": 0.1875,
+      "step": 7550
+    },
+    {
+      "epoch": 34.0,
+      "learning_rate": 1.771771771771772e-05,
+      "loss": 0.8216,
+      "step": 7560
+    },
+    {
+      "epoch": 34.0,
+      "learning_rate": 1.7667667667667666e-05,
+      "loss": 0.0497,
+      "step": 7570
+    },
+    {
+      "epoch": 34.0,
+      "learning_rate": 1.761761761761762e-05,
+      "loss": 0.0786,
+      "step": 7580
+    },
+    {
+      "epoch": 34.0,
+      "learning_rate": 1.756756756756757e-05,
+      "loss": 0.0391,
+      "step": 7590
+    },
+    {
+      "epoch": 34.0,
+      "learning_rate": 1.7517517517517516e-05,
+      "loss": 0.0136,
+      "step": 7600
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.746746746746747e-05,
+      "loss": 0.0021,
+      "step": 7610
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.7417417417417416e-05,
+      "loss": 0.0011,
+      "step": 7620
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.7367367367367366e-05,
+      "loss": 0.0285,
+      "step": 7630
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.731731731731732e-05,
+      "loss": 0.3956,
+      "step": 7640
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.7267267267267267e-05,
+      "loss": 0.1449,
+      "step": 7650
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.721721721721722e-05,
+      "loss": 0.455,
+      "step": 7660
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.7167167167167167e-05,
+      "loss": 0.0012,
+      "step": 7670
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.7117117117117117e-05,
+      "loss": 0.01,
+      "step": 7680
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.706706706706707e-05,
+      "loss": 0.1873,
+      "step": 7690
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.7017017017017017e-05,
+      "loss": 0.1313,
+      "step": 7700
+    },
+    {
+      "epoch": 34.01,
+      "learning_rate": 1.6966966966966967e-05,
+      "loss": 0.0641,
+      "step": 7710
+    },
+    {
+      "epoch": 34.02,
+      "learning_rate": 1.6916916916916917e-05,
+      "loss": 0.1805,
+      "step": 7720
+    },
+    {
+      "epoch": 34.02,
+      "learning_rate": 1.6866866866866867e-05,
+      "loss": 0.0006,
+      "step": 7730
+    },
+    {
+      "epoch": 34.02,
+      "learning_rate": 1.6816816816816817e-05,
+      "loss": 0.0574,
+      "step": 7740
+    },
+    {
+      "epoch": 34.02,
+      "learning_rate": 1.6766766766766767e-05,
+      "loss": 0.0014,
+      "step": 7750
+    },
+    {
+      "epoch": 34.02,
+      "learning_rate": 1.6716716716716717e-05,
+      "loss": 0.2973,
+      "step": 7760
+    },
+    {
+      "epoch": 34.02,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.0079,
+      "step": 7770
+    },
+    {
+      "epoch": 34.02,
+      "eval_accuracy": 0.8549763033175355,
+      "eval_loss": 0.8099485635757446,
+      "eval_runtime": 246.6955,
+      "eval_samples_per_second": 4.277,
+      "eval_steps_per_second": 2.14,
+      "step": 7770
+    },
+    {
+      "epoch": 35.0,
+      "learning_rate": 1.6616616616616618e-05,
+      "loss": 0.0035,
+      "step": 7780
+    },
+    {
+      "epoch": 35.0,
+      "learning_rate": 1.6566566566566568e-05,
+      "loss": 0.396,
+      "step": 7790
+    },
+    {
+      "epoch": 35.0,
+      "learning_rate": 1.6516516516516518e-05,
+      "loss": 0.2153,
+      "step": 7800
+    },
+    {
+      "epoch": 35.0,
+      "learning_rate": 1.6466466466466468e-05,
+      "loss": 0.0435,
+      "step": 7810
+    },
+    {
+      "epoch": 35.0,
+      "learning_rate": 1.6416416416416418e-05,
+      "loss": 0.238,
+      "step": 7820
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.6366366366366368e-05,
+      "loss": 0.211,
+      "step": 7830
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.6316316316316315e-05,
+      "loss": 0.1291,
+      "step": 7840
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.6266266266266268e-05,
+      "loss": 0.4311,
+      "step": 7850
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 0.383,
+      "step": 7860
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.6166166166166165e-05,
+      "loss": 0.0428,
+      "step": 7870
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.6116116116116118e-05,
+      "loss": 0.2092,
+      "step": 7880
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.6066066066066065e-05,
+      "loss": 0.0047,
+      "step": 7890
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.6016016016016015e-05,
+      "loss": 0.3449,
+      "step": 7900
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.596596596596597e-05,
+      "loss": 0.0567,
+      "step": 7910
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.5915915915915915e-05,
+      "loss": 0.246,
+      "step": 7920
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.586586586586587e-05,
+      "loss": 0.014,
+      "step": 7930
+    },
+    {
+      "epoch": 35.02,
+      "learning_rate": 1.5815815815815815e-05,
+      "loss": 0.293,
+      "step": 7940
+    },
+    {
+      "epoch": 35.02,
+      "learning_rate": 1.5765765765765765e-05,
+      "loss": 0.7073,
+      "step": 7950
+    },
+    {
+      "epoch": 35.02,
+      "learning_rate": 1.571571571571572e-05,
+      "loss": 0.6755,
+      "step": 7960
+    },
+    {
+      "epoch": 35.02,
+      "learning_rate": 1.5665665665665665e-05,
+      "loss": 0.1562,
+      "step": 7970
+    },
+    {
+      "epoch": 35.02,
+      "learning_rate": 1.5615615615615616e-05,
+      "loss": 0.418,
+      "step": 7980
+    },
+    {
+      "epoch": 35.02,
+      "learning_rate": 1.5565565565565566e-05,
+      "loss": 0.001,
+      "step": 7990
+    },
+    {
+      "epoch": 35.02,
+      "eval_accuracy": 0.8369668246445497,
+      "eval_loss": 0.8639578819274902,
+      "eval_runtime": 245.9861,
+      "eval_samples_per_second": 4.289,
+      "eval_steps_per_second": 2.146,
+      "step": 7992
+    },
+    {
+      "epoch": 36.0,
+      "learning_rate": 1.5515515515515516e-05,
+      "loss": 0.0044,
+      "step": 8000
+    },
+    {
+      "epoch": 36.0,
+      "learning_rate": 1.5465465465465466e-05,
+      "loss": 0.0388,
+      "step": 8010
+    },
+    {
+      "epoch": 36.0,
+      "learning_rate": 1.5415415415415416e-05,
+      "loss": 0.102,
+      "step": 8020
+    },
+    {
+      "epoch": 36.0,
+      "learning_rate": 1.5365365365365366e-05,
+      "loss": 0.2231,
+      "step": 8030
+    },
+    {
+      "epoch": 36.0,
+      "learning_rate": 1.5315315315315316e-05,
+      "loss": 0.1012,
+      "step": 8040
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.5265265265265266e-05,
+      "loss": 0.0153,
+      "step": 8050
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.5215215215215218e-05,
+      "loss": 0.0204,
+      "step": 8060
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.5165165165165166e-05,
+      "loss": 0.0011,
+      "step": 8070
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.5115115115115116e-05,
+      "loss": 0.0618,
+      "step": 8080
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.5065065065065065e-05,
+      "loss": 0.1385,
+      "step": 8090
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.5015015015015016e-05,
+      "loss": 0.2235,
+      "step": 8100
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.4964964964964966e-05,
+      "loss": 0.0051,
+      "step": 8110
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.4914914914914915e-05,
+      "loss": 0.7476,
+      "step": 8120
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.4864864864864867e-05,
+      "loss": 0.0007,
+      "step": 8130
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.5397,
+      "step": 8140
+    },
+    {
+      "epoch": 36.01,
+      "learning_rate": 1.4764764764764765e-05,
+      "loss": 0.0006,
+      "step": 8150
+    },
+    {
+      "epoch": 36.02,
+      "learning_rate": 1.4714714714714713e-05,
+      "loss": 0.0406,
+      "step": 8160
+    },
+    {
+      "epoch": 36.02,
+      "learning_rate": 1.4664664664664665e-05,
+      "loss": 0.0312,
+      "step": 8170
+    },
+    {
+      "epoch": 36.02,
+      "learning_rate": 1.4614614614614617e-05,
+      "loss": 0.2613,
+      "step": 8180
+    },
+    {
+      "epoch": 36.02,
+      "learning_rate": 1.4564564564564565e-05,
+      "loss": 0.0041,
+      "step": 8190
+    },
+    {
+      "epoch": 36.02,
+      "learning_rate": 1.4514514514514515e-05,
+      "loss": 0.0728,
+      "step": 8200
+    },
+    {
+      "epoch": 36.02,
+      "learning_rate": 1.4464464464464464e-05,
+      "loss": 0.0007,
+      "step": 8210
+    },
+    {
+      "epoch": 36.02,
+      "eval_accuracy": 0.8483412322274881,
+      "eval_loss": 0.7145898342132568,
+      "eval_runtime": 243.3375,
+      "eval_samples_per_second": 4.336,
+      "eval_steps_per_second": 2.17,
+      "step": 8214
+    },
+    {
+      "epoch": 37.0,
+      "learning_rate": 1.4414414414414416e-05,
+      "loss": 0.067,
+      "step": 8220
+    },
+    {
+      "epoch": 37.0,
+      "learning_rate": 1.4364364364364366e-05,
+      "loss": 0.4277,
+      "step": 8230
+    },
+    {
+      "epoch": 37.0,
+      "learning_rate": 1.4314314314314314e-05,
+      "loss": 0.0042,
+      "step": 8240
+    },
+    {
+      "epoch": 37.0,
+      "learning_rate": 1.4264264264264266e-05,
+      "loss": 0.1741,
+      "step": 8250
+    },
+    {
+      "epoch": 37.0,
+      "learning_rate": 1.4214214214214214e-05,
+      "loss": 0.0817,
+      "step": 8260
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.4164164164164164e-05,
+      "loss": 0.2151,
+      "step": 8270
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.4114114114114116e-05,
+      "loss": 0.0008,
+      "step": 8280
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.4064064064064064e-05,
+      "loss": 0.4043,
+      "step": 8290
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.4014014014014016e-05,
+      "loss": 0.0009,
+      "step": 8300
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.3963963963963963e-05,
+      "loss": 0.1504,
+      "step": 8310
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.3913913913913915e-05,
+      "loss": 0.051,
+      "step": 8320
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.3863863863863866e-05,
+      "loss": 0.3575,
+      "step": 8330
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.3813813813813815e-05,
+      "loss": 0.3269,
+      "step": 8340
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.3763763763763765e-05,
+      "loss": 0.0218,
+      "step": 8350
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.3713713713713713e-05,
+      "loss": 0.0009,
+      "step": 8360
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.3663663663663665e-05,
+      "loss": 0.0006,
+      "step": 8370
+    },
+    {
+      "epoch": 37.01,
+      "learning_rate": 1.3613613613613615e-05,
+      "loss": 0.2492,
+      "step": 8380
+    },
+    {
+      "epoch": 37.02,
+      "learning_rate": 1.3563563563563563e-05,
+      "loss": 0.0008,
+      "step": 8390
+    },
+    {
+      "epoch": 37.02,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 0.0086,
+      "step": 8400
+    },
+    {
+      "epoch": 37.02,
+      "learning_rate": 1.3463463463463463e-05,
+      "loss": 0.0013,
+      "step": 8410
+    },
+    {
+      "epoch": 37.02,
+      "learning_rate": 1.3413413413413414e-05,
+      "loss": 0.1867,
+      "step": 8420
+    },
+    {
+      "epoch": 37.02,
+      "learning_rate": 1.3363363363363365e-05,
+      "loss": 0.464,
+      "step": 8430
+    },
+    {
+      "epoch": 37.02,
+      "eval_accuracy": 0.8464454976303317,
+      "eval_loss": 0.7916568517684937,
+      "eval_runtime": 245.7907,
+      "eval_samples_per_second": 4.292,
+      "eval_steps_per_second": 2.148,
+      "step": 8436
+    },
+    {
+      "epoch": 38.0,
+      "learning_rate": 1.3313313313313314e-05,
+      "loss": 0.4968,
+      "step": 8440
+    },
+    {
+      "epoch": 38.0,
+      "learning_rate": 1.3263263263263265e-05,
+      "loss": 0.4161,
+      "step": 8450
+    },
+    {
+      "epoch": 38.0,
+      "learning_rate": 1.3213213213213214e-05,
+      "loss": 0.1849,
+      "step": 8460
+    },
+    {
+      "epoch": 38.0,
+      "learning_rate": 1.3163163163163164e-05,
+      "loss": 0.0809,
+      "step": 8470
+    },
+    {
+      "epoch": 38.0,
+      "learning_rate": 1.3113113113113112e-05,
+      "loss": 0.0009,
+      "step": 8480
+    },
+    {
+      "epoch": 38.0,
+      "learning_rate": 1.3063063063063064e-05,
+      "loss": 0.3014,
+      "step": 8490
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.3013013013013014e-05,
+      "loss": 0.0018,
+      "step": 8500
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.0703,
+      "step": 8510
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2912912912912914e-05,
+      "loss": 0.0031,
+      "step": 8520
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2862862862862863e-05,
+      "loss": 0.0024,
+      "step": 8530
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2812812812812813e-05,
+      "loss": 0.1886,
+      "step": 8540
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2762762762762764e-05,
+      "loss": 0.4053,
+      "step": 8550
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2712712712712713e-05,
+      "loss": 0.0014,
+      "step": 8560
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2662662662662665e-05,
+      "loss": 0.0026,
+      "step": 8570
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2612612612612611e-05,
+      "loss": 0.7475,
+      "step": 8580
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2562562562562563e-05,
+      "loss": 0.001,
+      "step": 8590
+    },
+    {
+      "epoch": 38.01,
+      "learning_rate": 1.2512512512512515e-05,
+      "loss": 0.5007,
+      "step": 8600
+    },
+    {
+      "epoch": 38.02,
+      "learning_rate": 1.2462462462462463e-05,
+      "loss": 0.6596,
+      "step": 8610
+    },
+    {
+      "epoch": 38.02,
+      "learning_rate": 1.2412412412412413e-05,
+      "loss": 0.0074,
+      "step": 8620
+    },
+    {
+      "epoch": 38.02,
+      "learning_rate": 1.2362362362362363e-05,
+      "loss": 0.045,
+      "step": 8630
+    },
+    {
+      "epoch": 38.02,
+      "learning_rate": 1.2312312312312313e-05,
+      "loss": 0.2084,
+      "step": 8640
+    },
+    {
+      "epoch": 38.02,
+      "learning_rate": 1.2262262262262263e-05,
+      "loss": 0.0005,
+      "step": 8650
+    },
+    {
+      "epoch": 38.02,
+      "eval_accuracy": 0.8530805687203792,
+      "eval_loss": 0.7239028811454773,
+      "eval_runtime": 247.9988,
+      "eval_samples_per_second": 4.254,
+      "eval_steps_per_second": 2.129,
+      "step": 8658
+    },
+    {
+      "epoch": 39.0,
+      "learning_rate": 1.2212212212212212e-05,
+      "loss": 0.0022,
+      "step": 8660
+    },
+    {
+      "epoch": 39.0,
+      "learning_rate": 1.2162162162162164e-05,
+      "loss": 0.2912,
+      "step": 8670
+    },
+    {
+      "epoch": 39.0,
+      "learning_rate": 1.2112112112112114e-05,
+      "loss": 0.2647,
+      "step": 8680
+    },
+    {
+      "epoch": 39.0,
+      "learning_rate": 1.2062062062062062e-05,
+      "loss": 0.021,
+      "step": 8690
+    },
+    {
+      "epoch": 39.0,
+      "learning_rate": 1.2012012012012012e-05,
+      "loss": 0.001,
+      "step": 8700
+    },
+    {
+      "epoch": 39.0,
+      "learning_rate": 1.1961961961961962e-05,
+      "loss": 0.0012,
+      "step": 8710
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1911911911911912e-05,
+      "loss": 0.3884,
+      "step": 8720
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1861861861861862e-05,
+      "loss": 0.3583,
+      "step": 8730
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1811811811811812e-05,
+      "loss": 0.0033,
+      "step": 8740
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1761761761761762e-05,
+      "loss": 0.1403,
+      "step": 8750
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1711711711711713e-05,
+      "loss": 0.0703,
+      "step": 8760
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1661661661661661e-05,
+      "loss": 0.0699,
+      "step": 8770
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1611611611611613e-05,
+      "loss": 0.2121,
+      "step": 8780
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1561561561561563e-05,
+      "loss": 0.2261,
+      "step": 8790
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1511511511511513e-05,
+      "loss": 0.0029,
+      "step": 8800
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1461461461461461e-05,
+      "loss": 0.3778,
+      "step": 8810
+    },
+    {
+      "epoch": 39.01,
+      "learning_rate": 1.1411411411411411e-05,
+      "loss": 0.381,
+      "step": 8820
+    },
+    {
+      "epoch": 39.02,
+      "learning_rate": 1.1361361361361363e-05,
+      "loss": 0.4107,
+      "step": 8830
+    },
+    {
+      "epoch": 39.02,
+      "learning_rate": 1.1311311311311313e-05,
+      "loss": 0.008,
+      "step": 8840
+    },
+    {
+      "epoch": 39.02,
+      "learning_rate": 1.1261261261261261e-05,
+      "loss": 0.1066,
+      "step": 8850
+    },
+    {
+      "epoch": 39.02,
+      "learning_rate": 1.1211211211211212e-05,
+      "loss": 0.0665,
+      "step": 8860
+    },
+    {
+      "epoch": 39.02,
+      "learning_rate": 1.1161161161161162e-05,
+      "loss": 0.1089,
+      "step": 8870
+    },
+    {
+      "epoch": 39.02,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.0004,
+      "step": 8880
+    },
+    {
+      "epoch": 39.02,
+      "eval_accuracy": 0.8701421800947867,
+      "eval_loss": 0.7702277302742004,
+      "eval_runtime": 246.605,
+      "eval_samples_per_second": 4.278,
+      "eval_steps_per_second": 2.141,
+      "step": 8880
+    },
+    {
+      "epoch": 40.0,
+      "learning_rate": 1.1061061061061062e-05,
+      "loss": 0.0014,
+      "step": 8890
+    },
+    {
+      "epoch": 40.0,
+      "learning_rate": 1.1011011011011012e-05,
+      "loss": 0.0042,
+      "step": 8900
+    },
+    {
+      "epoch": 40.0,
+      "learning_rate": 1.0960960960960962e-05,
+      "loss": 0.1076,
+      "step": 8910
+    },
+    {
+      "epoch": 40.0,
+      "learning_rate": 1.0910910910910912e-05,
+      "loss": 0.3793,
+      "step": 8920
+    },
+    {
+      "epoch": 40.0,
+      "learning_rate": 1.086086086086086e-05,
+      "loss": 0.4586,
+      "step": 8930
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 0.6425,
+      "step": 8940
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.0760760760760762e-05,
+      "loss": 0.0047,
+      "step": 8950
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.071071071071071e-05,
+      "loss": 0.1311,
+      "step": 8960
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.066066066066066e-05,
+      "loss": 0.0072,
+      "step": 8970
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.061061061061061e-05,
+      "loss": 0.0011,
+      "step": 8980
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.0560560560560562e-05,
+      "loss": 0.0205,
+      "step": 8990
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.051051051051051e-05,
+      "loss": 0.2733,
+      "step": 9000
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.0460460460460461e-05,
+      "loss": 0.0007,
+      "step": 9010
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.0410410410410411e-05,
+      "loss": 0.1359,
+      "step": 9020
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.0360360360360361e-05,
+      "loss": 0.0124,
+      "step": 9030
+    },
+    {
+      "epoch": 40.01,
+      "learning_rate": 1.031031031031031e-05,
+      "loss": 0.0689,
+      "step": 9040
+    },
+    {
+      "epoch": 40.02,
+      "learning_rate": 1.0260260260260261e-05,
+      "loss": 0.001,
+      "step": 9050
+    },
+    {
+      "epoch": 40.02,
+      "learning_rate": 1.0210210210210211e-05,
+      "loss": 0.0011,
+      "step": 9060
+    },
+    {
+      "epoch": 40.02,
+      "learning_rate": 1.0160160160160161e-05,
+      "loss": 0.0014,
+      "step": 9070
+    },
+    {
+      "epoch": 40.02,
+      "learning_rate": 1.011011011011011e-05,
+      "loss": 0.3321,
+      "step": 9080
+    },
+    {
+      "epoch": 40.02,
+      "learning_rate": 1.006006006006006e-05,
+      "loss": 0.1678,
+      "step": 9090
+    },
+    {
+      "epoch": 40.02,
+      "learning_rate": 1.0010010010010011e-05,
+      "loss": 0.1705,
+      "step": 9100
+    },
+    {
+      "epoch": 40.02,
+      "eval_accuracy": 0.852132701421801,
+      "eval_loss": 0.754317581653595,
+      "eval_runtime": 246.865,
+      "eval_samples_per_second": 4.274,
+      "eval_steps_per_second": 2.139,
+      "step": 9102
+    },
+    {
+      "epoch": 41.0,
+      "learning_rate": 9.959959959959962e-06,
+      "loss": 0.4683,
+      "step": 9110
+    },
+    {
+      "epoch": 41.0,
+      "learning_rate": 9.90990990990991e-06,
+      "loss": 0.2015,
+      "step": 9120
+    },
+    {
+      "epoch": 41.0,
+      "learning_rate": 9.85985985985986e-06,
+      "loss": 0.3734,
+      "step": 9130
+    },
+    {
+      "epoch": 41.0,
+      "learning_rate": 9.80980980980981e-06,
+      "loss": 0.0012,
+      "step": 9140
+    },
+    {
+      "epoch": 41.0,
+      "learning_rate": 9.75975975975976e-06,
+      "loss": 0.0008,
+      "step": 9150
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.70970970970971e-06,
+      "loss": 0.0008,
+      "step": 9160
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.65965965965966e-06,
+      "loss": 0.2823,
+      "step": 9170
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.60960960960961e-06,
+      "loss": 0.021,
+      "step": 9180
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.55955955955956e-06,
+      "loss": 0.1154,
+      "step": 9190
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.509509509509509e-06,
+      "loss": 0.1189,
+      "step": 9200
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.2888,
+      "step": 9210
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.40940940940941e-06,
+      "loss": 0.004,
+      "step": 9220
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.359359359359359e-06,
+      "loss": 0.0107,
+      "step": 9230
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.309309309309309e-06,
+      "loss": 0.023,
+      "step": 9240
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.0516,
+      "step": 9250
+    },
+    {
+      "epoch": 41.01,
+      "learning_rate": 9.209209209209211e-06,
+      "loss": 0.5651,
+      "step": 9260
+    },
+    {
+      "epoch": 41.02,
+      "learning_rate": 9.15915915915916e-06,
+      "loss": 0.0015,
+      "step": 9270
+    },
+    {
+      "epoch": 41.02,
+      "learning_rate": 9.10910910910911e-06,
+      "loss": 0.137,
+      "step": 9280
+    },
+    {
+      "epoch": 41.02,
+      "learning_rate": 9.05905905905906e-06,
+      "loss": 0.0049,
+      "step": 9290
+    },
+    {
+      "epoch": 41.02,
+      "learning_rate": 9.00900900900901e-06,
+      "loss": 0.0089,
+      "step": 9300
+    },
+    {
+      "epoch": 41.02,
+      "learning_rate": 8.958958958958958e-06,
+      "loss": 0.1895,
+      "step": 9310
+    },
+    {
+      "epoch": 41.02,
+      "learning_rate": 8.90890890890891e-06,
+      "loss": 0.0039,
+      "step": 9320
+    },
+    {
+      "epoch": 41.02,
+      "eval_accuracy": 0.8672985781990521,
+      "eval_loss": 0.745583713054657,
+      "eval_runtime": 244.1188,
+      "eval_samples_per_second": 4.322,
+      "eval_steps_per_second": 2.163,
+      "step": 9324
+    },
+    {
+      "epoch": 42.0,
+      "learning_rate": 8.85885885885886e-06,
+      "loss": 0.2742,
+      "step": 9330
+    },
+    {
+      "epoch": 42.0,
+      "learning_rate": 8.80880880880881e-06,
+      "loss": 0.0005,
+      "step": 9340
+    },
+    {
+      "epoch": 42.0,
+      "learning_rate": 8.758758758758758e-06,
+      "loss": 0.0014,
+      "step": 9350
+    },
+    {
+      "epoch": 42.0,
+      "learning_rate": 8.708708708708708e-06,
+      "loss": 0.2073,
+      "step": 9360
+    },
+    {
+      "epoch": 42.0,
+      "learning_rate": 8.65865865865866e-06,
+      "loss": 0.0024,
+      "step": 9370
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.60860860860861e-06,
+      "loss": 0.1651,
+      "step": 9380
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.558558558558558e-06,
+      "loss": 0.2201,
+      "step": 9390
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.508508508508508e-06,
+      "loss": 0.0086,
+      "step": 9400
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.458458458458459e-06,
+      "loss": 0.0475,
+      "step": 9410
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.408408408408409e-06,
+      "loss": 0.5435,
+      "step": 9420
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.358358358358359e-06,
+      "loss": 0.011,
+      "step": 9430
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.308308308308309e-06,
+      "loss": 0.003,
+      "step": 9440
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.258258258258259e-06,
+      "loss": 0.1767,
+      "step": 9450
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.208208208208209e-06,
+      "loss": 0.8353,
+      "step": 9460
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.158158158158157e-06,
+      "loss": 0.2716,
+      "step": 9470
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.1919,
+      "step": 9480
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 8.058058058058059e-06,
+      "loss": 0.1059,
+      "step": 9490
+    },
+    {
+      "epoch": 42.02,
+      "learning_rate": 8.008008008008007e-06,
+      "loss": 0.0036,
+      "step": 9500
+    },
+    {
+      "epoch": 42.02,
+      "learning_rate": 7.957957957957958e-06,
+      "loss": 0.0007,
+      "step": 9510
+    },
+    {
+      "epoch": 42.02,
+      "learning_rate": 7.907907907907908e-06,
+      "loss": 0.1438,
+      "step": 9520
+    },
+    {
+      "epoch": 42.02,
+      "learning_rate": 7.85785785785786e-06,
+      "loss": 0.0017,
+      "step": 9530
+    },
+    {
+      "epoch": 42.02,
+      "learning_rate": 7.807807807807808e-06,
+      "loss": 0.0168,
+      "step": 9540
+    },
+    {
+      "epoch": 42.02,
+      "eval_accuracy": 0.8729857819905213,
+      "eval_loss": 0.7254510521888733,
+      "eval_runtime": 241.2967,
+      "eval_samples_per_second": 4.372,
+      "eval_steps_per_second": 2.188,
+      "step": 9546
+    },
+    {
+      "epoch": 43.0,
+      "learning_rate": 7.757757757757758e-06,
+      "loss": 0.0014,
+      "step": 9550
+    },
+    {
+      "epoch": 43.0,
+      "learning_rate": 7.707707707707708e-06,
+      "loss": 0.0094,
+      "step": 9560
+    },
+    {
+      "epoch": 43.0,
+      "learning_rate": 7.657657657657658e-06,
+      "loss": 0.0006,
+      "step": 9570
+    },
+    {
+      "epoch": 43.0,
+      "learning_rate": 7.607607607607609e-06,
+      "loss": 0.2758,
+      "step": 9580
+    },
+    {
+      "epoch": 43.0,
+      "learning_rate": 7.557557557557558e-06,
+      "loss": 0.037,
+      "step": 9590
+    },
+    {
+      "epoch": 43.0,
+      "learning_rate": 7.507507507507508e-06,
+      "loss": 0.0006,
+      "step": 9600
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.457457457457457e-06,
+      "loss": 0.0019,
+      "step": 9610
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.0015,
+      "step": 9620
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.357357357357357e-06,
+      "loss": 0.0028,
+      "step": 9630
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.3073073073073085e-06,
+      "loss": 0.3271,
+      "step": 9640
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.257257257257258e-06,
+      "loss": 0.03,
+      "step": 9650
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.207207207207208e-06,
+      "loss": 0.0021,
+      "step": 9660
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.157157157157157e-06,
+      "loss": 0.0014,
+      "step": 9670
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.107107107107107e-06,
+      "loss": 0.1468,
+      "step": 9680
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.057057057057058e-06,
+      "loss": 0.0237,
+      "step": 9690
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 7.007007007007008e-06,
+      "loss": 0.6728,
+      "step": 9700
+    },
+    {
+      "epoch": 43.01,
+      "learning_rate": 6.956956956956957e-06,
+      "loss": 0.0013,
+      "step": 9710
+    },
+    {
+      "epoch": 43.02,
+      "learning_rate": 6.906906906906907e-06,
+      "loss": 0.1682,
+      "step": 9720
+    },
+    {
+      "epoch": 43.02,
+      "learning_rate": 6.8568568568568566e-06,
+      "loss": 0.0048,
+      "step": 9730
+    },
+    {
+      "epoch": 43.02,
+      "learning_rate": 6.8068068068068075e-06,
+      "loss": 0.0015,
+      "step": 9740
+    },
+    {
+      "epoch": 43.02,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.0035,
+      "step": 9750
+    },
+    {
+      "epoch": 43.02,
+      "learning_rate": 6.706706706706707e-06,
+      "loss": 0.2615,
+      "step": 9760
+    },
+    {
+      "epoch": 43.02,
+      "eval_accuracy": 0.875829383886256,
+      "eval_loss": 0.7453275918960571,
+      "eval_runtime": 246.2851,
+      "eval_samples_per_second": 4.284,
+      "eval_steps_per_second": 2.144,
+      "step": 9768
+    },
+    {
+      "epoch": 44.0,
+      "learning_rate": 6.656656656656657e-06,
+      "loss": 0.0054,
+      "step": 9770
+    },
+    {
+      "epoch": 44.0,
+      "learning_rate": 6.606606606606607e-06,
+      "loss": 0.0009,
+      "step": 9780
+    },
+    {
+      "epoch": 44.0,
+      "learning_rate": 6.556556556556556e-06,
+      "loss": 0.0156,
+      "step": 9790
+    },
+    {
+      "epoch": 44.0,
+      "learning_rate": 6.506506506506507e-06,
+      "loss": 0.8669,
+      "step": 9800
+    },
+    {
+      "epoch": 44.0,
+      "learning_rate": 6.456456456456457e-06,
+      "loss": 0.0007,
+      "step": 9810
+    },
+    {
+      "epoch": 44.0,
+      "learning_rate": 6.406406406406406e-06,
+      "loss": 0.3862,
+      "step": 9820
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 6.356356356356356e-06,
+      "loss": 0.481,
+      "step": 9830
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 6.306306306306306e-06,
+      "loss": 0.0194,
+      "step": 9840
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 6.256256256256257e-06,
+      "loss": 0.0013,
+      "step": 9850
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 6.206206206206207e-06,
+      "loss": 0.3295,
+      "step": 9860
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 6.156156156156157e-06,
+      "loss": 0.0008,
+      "step": 9870
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 6.106106106106106e-06,
+      "loss": 0.0036,
+      "step": 9880
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 6.056056056056057e-06,
+      "loss": 0.0007,
+      "step": 9890
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 6.006006006006006e-06,
+      "loss": 0.0035,
+      "step": 9900
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 5.955955955955956e-06,
+      "loss": 0.0006,
+      "step": 9910
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 5.905905905905906e-06,
+      "loss": 0.4467,
+      "step": 9920
+    },
+    {
+      "epoch": 44.01,
+      "learning_rate": 5.855855855855856e-06,
+      "loss": 0.275,
+      "step": 9930
+    },
+    {
+      "epoch": 44.02,
+      "learning_rate": 5.805805805805806e-06,
+      "loss": 0.0064,
+      "step": 9940
+    },
+    {
+      "epoch": 44.02,
+      "learning_rate": 5.755755755755756e-06,
+      "loss": 0.0347,
+      "step": 9950
+    },
+    {
+      "epoch": 44.02,
+      "learning_rate": 5.705705705705706e-06,
+      "loss": 0.1461,
+      "step": 9960
+    },
+    {
+      "epoch": 44.02,
+      "learning_rate": 5.6556556556556565e-06,
+      "loss": 0.0011,
+      "step": 9970
+    },
+    {
+      "epoch": 44.02,
+      "learning_rate": 5.605605605605606e-06,
+      "loss": 0.235,
+      "step": 9980
+    },
+    {
+      "epoch": 44.02,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.0004,
+      "step": 9990
+    },
+    {
+      "epoch": 44.02,
+      "eval_accuracy": 0.8805687203791469,
+      "eval_loss": 0.6823952198028564,
+      "eval_runtime": 247.5072,
+      "eval_samples_per_second": 4.263,
+      "eval_steps_per_second": 2.133,
+      "step": 9990
+    },
+    {
+      "epoch": 45.0,
+      "learning_rate": 5.505505505505506e-06,
+      "loss": 0.2194,
+      "step": 10000
+    },
+    {
+      "epoch": 45.0,
+      "learning_rate": 5.455455455455456e-06,
+      "loss": 0.0106,
+      "step": 10010
+    },
+    {
+      "epoch": 45.0,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.2476,
+      "step": 10020
+    },
+    {
+      "epoch": 45.0,
+      "learning_rate": 5.355355355355355e-06,
+      "loss": 0.0026,
+      "step": 10030
+    },
+    {
+      "epoch": 45.0,
+      "learning_rate": 5.305305305305305e-06,
+      "loss": 0.0022,
+      "step": 10040
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 5.255255255255255e-06,
+      "loss": 0.0074,
+      "step": 10050
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 5.2052052052052055e-06,
+      "loss": 0.2722,
+      "step": 10060
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 5.155155155155155e-06,
+      "loss": 0.056,
+      "step": 10070
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 5.105105105105106e-06,
+      "loss": 0.001,
+      "step": 10080
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 5.055055055055055e-06,
+      "loss": 0.183,
+      "step": 10090
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 5.005005005005006e-06,
+      "loss": 0.0007,
+      "step": 10100
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 4.954954954954955e-06,
+      "loss": 0.0023,
+      "step": 10110
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 4.904904904904905e-06,
+      "loss": 0.0013,
+      "step": 10120
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 4.854854854854855e-06,
+      "loss": 0.1421,
+      "step": 10130
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 4.804804804804805e-06,
+      "loss": 0.1778,
+      "step": 10140
+    },
+    {
+      "epoch": 45.01,
+      "learning_rate": 4.754754754754754e-06,
+      "loss": 0.0007,
+      "step": 10150
+    },
+    {
+      "epoch": 45.02,
+      "learning_rate": 4.704704704704705e-06,
+      "loss": 0.0045,
+      "step": 10160
+    },
+    {
+      "epoch": 45.02,
+      "learning_rate": 4.6546546546546545e-06,
+      "loss": 0.0102,
+      "step": 10170
+    },
+    {
+      "epoch": 45.02,
+      "learning_rate": 4.6046046046046055e-06,
+      "loss": 0.0019,
+      "step": 10180
+    },
+    {
+      "epoch": 45.02,
+      "learning_rate": 4.554554554554555e-06,
+      "loss": 0.0024,
+      "step": 10190
+    },
+    {
+      "epoch": 45.02,
+      "learning_rate": 4.504504504504505e-06,
+      "loss": 0.3791,
+      "step": 10200
+    },
+    {
+      "epoch": 45.02,
+      "learning_rate": 4.454454454454455e-06,
+      "loss": 0.236,
+      "step": 10210
+    },
+    {
+      "epoch": 45.02,
+      "eval_accuracy": 0.8824644549763033,
+      "eval_loss": 0.6623644828796387,
+      "eval_runtime": 247.6484,
+      "eval_samples_per_second": 4.26,
+      "eval_steps_per_second": 2.132,
+      "step": 10212
+    },
+    {
+      "epoch": 46.0,
+      "learning_rate": 4.404404404404405e-06,
+      "loss": 0.0367,
+      "step": 10220
+    },
+    {
+      "epoch": 46.0,
+      "learning_rate": 4.354354354354354e-06,
+      "loss": 0.1582,
+      "step": 10230
+    },
+    {
+      "epoch": 46.0,
+      "learning_rate": 4.304304304304305e-06,
+      "loss": 0.3461,
+      "step": 10240
+    },
+    {
+      "epoch": 46.0,
+      "learning_rate": 4.254254254254254e-06,
+      "loss": 0.003,
+      "step": 10250
+    },
+    {
+      "epoch": 46.0,
+      "learning_rate": 4.204204204204204e-06,
+      "loss": 0.172,
+      "step": 10260
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 4.154154154154154e-06,
+      "loss": 0.2162,
+      "step": 10270
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 4.1041041041041045e-06,
+      "loss": 0.0031,
+      "step": 10280
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.0017,
+      "step": 10290
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 4.004004004004004e-06,
+      "loss": 0.0007,
+      "step": 10300
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 3.953953953953954e-06,
+      "loss": 0.0016,
+      "step": 10310
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 3.903903903903904e-06,
+      "loss": 0.001,
+      "step": 10320
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 3.853853853853854e-06,
+      "loss": 0.0007,
+      "step": 10330
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 3.8038038038038044e-06,
+      "loss": 0.2318,
+      "step": 10340
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 3.753753753753754e-06,
+      "loss": 0.0045,
+      "step": 10350
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.0006,
+      "step": 10360
+    },
+    {
+      "epoch": 46.01,
+      "learning_rate": 3.6536536536536542e-06,
+      "loss": 0.0017,
+      "step": 10370
+    },
+    {
+      "epoch": 46.02,
+      "learning_rate": 3.603603603603604e-06,
+      "loss": 0.3637,
+      "step": 10380
+    },
+    {
+      "epoch": 46.02,
+      "learning_rate": 3.5535535535535535e-06,
+      "loss": 0.0161,
+      "step": 10390
+    },
+    {
+      "epoch": 46.02,
+      "learning_rate": 3.503503503503504e-06,
+      "loss": 0.0026,
+      "step": 10400
+    },
+    {
+      "epoch": 46.02,
+      "learning_rate": 3.4534534534534537e-06,
+      "loss": 0.0016,
+      "step": 10410
+    },
+    {
+      "epoch": 46.02,
+      "learning_rate": 3.4034034034034037e-06,
+      "loss": 0.0006,
+      "step": 10420
+    },
+    {
+      "epoch": 46.02,
+      "learning_rate": 3.3533533533533534e-06,
+      "loss": 0.0007,
+      "step": 10430
+    },
+    {
+      "epoch": 46.02,
+      "eval_accuracy": 0.8815165876777251,
+      "eval_loss": 0.6727480292320251,
+      "eval_runtime": 246.5992,
+      "eval_samples_per_second": 4.278,
+      "eval_steps_per_second": 2.141,
+      "step": 10434
+    },
+    {
+      "epoch": 47.0,
+      "learning_rate": 3.3033033033033035e-06,
+      "loss": 0.2267,
+      "step": 10440
+    },
+    {
+      "epoch": 47.0,
+      "learning_rate": 3.2532532532532535e-06,
+      "loss": 0.008,
+      "step": 10450
+    },
+    {
+      "epoch": 47.0,
+      "learning_rate": 3.203203203203203e-06,
+      "loss": 0.0008,
+      "step": 10460
+    },
+    {
+      "epoch": 47.0,
+      "learning_rate": 3.153153153153153e-06,
+      "loss": 0.0004,
+      "step": 10470
+    },
+    {
+      "epoch": 47.0,
+      "learning_rate": 3.1031031031031033e-06,
+      "loss": 0.0013,
+      "step": 10480
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 3.053053053053053e-06,
+      "loss": 0.3982,
+      "step": 10490
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 3.003003003003003e-06,
+      "loss": 0.0003,
+      "step": 10500
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.952952952952953e-06,
+      "loss": 0.0036,
+      "step": 10510
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.902902902902903e-06,
+      "loss": 0.004,
+      "step": 10520
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.852852852852853e-06,
+      "loss": 0.001,
+      "step": 10530
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.802802802802803e-06,
+      "loss": 0.2879,
+      "step": 10540
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.752752752752753e-06,
+      "loss": 0.001,
+      "step": 10550
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.2931,
+      "step": 10560
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.6526526526526527e-06,
+      "loss": 0.0084,
+      "step": 10570
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.6026026026026027e-06,
+      "loss": 0.2518,
+      "step": 10580
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.552552552552553e-06,
+      "loss": 0.0006,
+      "step": 10590
+    },
+    {
+      "epoch": 47.01,
+      "learning_rate": 2.502502502502503e-06,
+      "loss": 0.0017,
+      "step": 10600
+    },
+    {
+      "epoch": 47.02,
+      "learning_rate": 2.4524524524524525e-06,
+      "loss": 0.023,
+      "step": 10610
+    },
+    {
+      "epoch": 47.02,
+      "learning_rate": 2.4024024024024026e-06,
+      "loss": 0.0017,
+      "step": 10620
+    },
+    {
+      "epoch": 47.02,
+      "learning_rate": 2.3523523523523527e-06,
+      "loss": 0.3939,
+      "step": 10630
+    },
+    {
+      "epoch": 47.02,
+      "learning_rate": 2.3023023023023027e-06,
+      "loss": 0.874,
+      "step": 10640
+    },
+    {
+      "epoch": 47.02,
+      "learning_rate": 2.2522522522522524e-06,
+      "loss": 0.0004,
+      "step": 10650
+    },
+    {
+      "epoch": 47.02,
+      "eval_accuracy": 0.8862559241706162,
+      "eval_loss": 0.6478492021560669,
+      "eval_runtime": 248.2655,
+      "eval_samples_per_second": 4.249,
+      "eval_steps_per_second": 2.127,
+      "step": 10656
+    },
+    {
+      "epoch": 48.0,
+      "learning_rate": 2.2022022022022024e-06,
+      "loss": 0.543,
+      "step": 10660
+    },
+    {
+      "epoch": 48.0,
+      "learning_rate": 2.1521521521521525e-06,
+      "loss": 0.0335,
+      "step": 10670
+    },
+    {
+      "epoch": 48.0,
+      "learning_rate": 2.102102102102102e-06,
+      "loss": 0.0964,
+      "step": 10680
+    },
+    {
+      "epoch": 48.0,
+      "learning_rate": 2.0520520520520522e-06,
+      "loss": 0.001,
+      "step": 10690
+    },
+    {
+      "epoch": 48.0,
+      "learning_rate": 2.002002002002002e-06,
+      "loss": 0.0016,
+      "step": 10700
+    },
+    {
+      "epoch": 48.0,
+      "learning_rate": 1.951951951951952e-06,
+      "loss": 0.0025,
+      "step": 10710
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.9019019019019022e-06,
+      "loss": 0.0011,
+      "step": 10720
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.0053,
+      "step": 10730
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.801801801801802e-06,
+      "loss": 0.0018,
+      "step": 10740
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.751751751751752e-06,
+      "loss": 0.3022,
+      "step": 10750
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.7017017017017019e-06,
+      "loss": 0.0011,
+      "step": 10760
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.6516516516516517e-06,
+      "loss": 0.204,
+      "step": 10770
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.6016016016016016e-06,
+      "loss": 0.1849,
+      "step": 10780
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.5515515515515517e-06,
+      "loss": 0.4035,
+      "step": 10790
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.5015015015015015e-06,
+      "loss": 0.0031,
+      "step": 10800
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.4514514514514516e-06,
+      "loss": 0.1468,
+      "step": 10810
+    },
+    {
+      "epoch": 48.01,
+      "learning_rate": 1.4014014014014014e-06,
+      "loss": 0.0321,
+      "step": 10820
+    },
+    {
+      "epoch": 48.02,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.0004,
+      "step": 10830
+    },
+    {
+      "epoch": 48.02,
+      "learning_rate": 1.3013013013013014e-06,
+      "loss": 0.0008,
+      "step": 10840
+    },
+    {
+      "epoch": 48.02,
+      "learning_rate": 1.2512512512512514e-06,
+      "loss": 0.0009,
+      "step": 10850
+    },
+    {
+      "epoch": 48.02,
+      "learning_rate": 1.2012012012012013e-06,
+      "loss": 0.0003,
+      "step": 10860
+    },
+    {
+      "epoch": 48.02,
+      "learning_rate": 1.1511511511511514e-06,
+      "loss": 0.268,
+      "step": 10870
+    },
+    {
+      "epoch": 48.02,
+      "eval_accuracy": 0.8900473933649289,
+      "eval_loss": 0.6309255361557007,
+      "eval_runtime": 247.9172,
+      "eval_samples_per_second": 4.255,
+      "eval_steps_per_second": 2.13,
+      "step": 10878
+    },
+    {
+      "epoch": 49.0,
+      "learning_rate": 1.1011011011011012e-06,
+      "loss": 0.3578,
+      "step": 10880
+    },
+    {
+      "epoch": 49.0,
+      "learning_rate": 1.051051051051051e-06,
+      "loss": 0.2747,
+      "step": 10890
+    },
+    {
+      "epoch": 49.0,
+      "learning_rate": 1.001001001001001e-06,
+      "loss": 0.0309,
+      "step": 10900
+    },
+    {
+      "epoch": 49.0,
+      "learning_rate": 9.509509509509511e-07,
+      "loss": 0.2472,
+      "step": 10910
+    },
+    {
+      "epoch": 49.0,
+      "learning_rate": 9.00900900900901e-07,
+      "loss": 0.0008,
+      "step": 10920
+    },
+    {
+      "epoch": 49.0,
+      "learning_rate": 8.508508508508509e-07,
+      "loss": 0.0011,
+      "step": 10930
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 8.008008008008008e-07,
+      "loss": 0.0023,
+      "step": 10940
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 7.507507507507508e-07,
+      "loss": 0.0062,
+      "step": 10950
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 7.007007007007007e-07,
+      "loss": 0.0005,
+      "step": 10960
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 6.506506506506507e-07,
+      "loss": 0.0008,
+      "step": 10970
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 6.006006006006006e-07,
+      "loss": 0.0004,
+      "step": 10980
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 5.505505505505506e-07,
+      "loss": 0.0008,
+      "step": 10990
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 5.005005005005005e-07,
+      "loss": 0.2702,
+      "step": 11000
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 4.504504504504505e-07,
+      "loss": 0.0017,
+      "step": 11010
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 4.004004004004004e-07,
+      "loss": 0.2218,
+      "step": 11020
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 3.5035035035035036e-07,
+      "loss": 0.0178,
+      "step": 11030
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 3.003003003003003e-07,
+      "loss": 0.0018,
+      "step": 11040
+    },
+    {
+      "epoch": 49.02,
+      "learning_rate": 2.5025025025025023e-07,
+      "loss": 0.026,
+      "step": 11050
+    },
+    {
+      "epoch": 49.02,
+      "learning_rate": 2.002002002002002e-07,
+      "loss": 0.0038,
+      "step": 11060
+    },
+    {
+      "epoch": 49.02,
+      "learning_rate": 1.5015015015015016e-07,
+      "loss": 0.0311,
+      "step": 11070
+    },
+    {
+      "epoch": 49.02,
+      "learning_rate": 1.001001001001001e-07,
+      "loss": 0.0016,
+      "step": 11080
+    },
+    {
+      "epoch": 49.02,
+      "learning_rate": 5.005005005005005e-08,
+      "loss": 0.2836,
+      "step": 11090
+    },
+    {
+      "epoch": 49.02,
+      "learning_rate": 0.0,
+      "loss": 0.0025,
+      "step": 11100
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.8900473933649289,
+      "eval_loss": 0.6283966302871704,
+      "eval_runtime": 249.5538,
+      "eval_samples_per_second": 4.228,
+      "eval_steps_per_second": 2.116,
+      "step": 11100
+    },
+    {
+      "epoch": 49.02,
+      "step": 11100,
+      "total_flos": 9.749360839040041e+19,
+      "train_loss": 0.4945326050639002,
+      "train_runtime": 25011.5744,
+      "train_samples_per_second": 0.888,
+      "train_steps_per_second": 0.444
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.8900473933649289,
+      "eval_loss": 0.6309255361557007,
+      "eval_runtime": 248.4841,
+      "eval_samples_per_second": 4.246,
+      "eval_steps_per_second": 2.125,
+      "step": 11100
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.8900473933649289,
+      "eval_loss": 0.6309255361557007,
+      "eval_runtime": 248.2618,
+      "eval_samples_per_second": 4.25,
+      "eval_steps_per_second": 2.127,
+      "step": 11100
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 11100,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "total_flos": 9.749360839040041e+19,
+  "trial_name": null,
+  "trial_params": null
+}