diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5839 @@
+{
+  "best_metric": 0.730299667036626,
+  "best_model_checkpoint": "videomae-base-finetuned-inhard-finetuned-inhard2-finetuned-inhard3\\checkpoint-9580",
+  "epoch": 4.2,
+  "eval_steps": 500,
+  "global_step": 9580,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.219206680584552e-07,
+      "loss": 0.5692,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0438413361169103e-06,
+      "loss": 0.7418,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5657620041753654e-06,
+      "loss": 1.5564,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0876826722338207e-06,
+      "loss": 0.297,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6096033402922757e-06,
+      "loss": 0.5621,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1315240083507308e-06,
+      "loss": 0.5965,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.653444676409186e-06,
+      "loss": 1.1183,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175365344467641e-06,
+      "loss": 0.2349,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697286012526096e-06,
+      "loss": 0.1659,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.2192066805845514e-06,
+      "loss": 0.4711,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.741127348643006e-06,
+      "loss": 0.0735,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.2630480167014616e-06,
+      "loss": 1.0543,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.784968684759917e-06,
+      "loss": 0.5362,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.306889352818372e-06,
+      "loss": 0.4345,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.828810020876827e-06,
+      "loss": 0.0477,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.350730688935283e-06,
+      "loss": 0.3437,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.872651356993736e-06,
+      "loss": 0.3358,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.394572025052192e-06,
+      "loss": 0.8122,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.916492693110647e-06,
+      "loss": 0.3556,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0438413361169103e-05,
+      "loss": 0.3334,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0960334029227558e-05,
+      "loss": 0.3254,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1482254697286012e-05,
+      "loss": 0.851,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2004175365344468e-05,
+      "loss": 0.6903,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.2526096033402923e-05,
+      "loss": 0.764,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.304801670146138e-05,
+      "loss": 0.5178,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.3569937369519834e-05,
+      "loss": 1.2913,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.409185803757829e-05,
+      "loss": 0.3391,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.4613778705636743e-05,
+      "loss": 0.1099,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.51356993736952e-05,
+      "loss": 0.7243,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5657620041753654e-05,
+      "loss": 0.3011,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.617954070981211e-05,
+      "loss": 1.3619,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.6701461377870565e-05,
+      "loss": 0.5962,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.722338204592902e-05,
+      "loss": 1.4289,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.7745302713987473e-05,
+      "loss": 1.2534,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.826722338204593e-05,
+      "loss": 0.048,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.8789144050104384e-05,
+      "loss": 0.7844,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.931106471816284e-05,
+      "loss": 0.4099,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9832985386221295e-05,
+      "loss": 0.5046,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0354906054279752e-05,
+      "loss": 0.66,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0876826722338206e-05,
+      "loss": 2.3522,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.1398747390396663e-05,
+      "loss": 0.4574,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.1920668058455117e-05,
+      "loss": 0.6253,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.244258872651357e-05,
+      "loss": 1.2977,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.2964509394572024e-05,
+      "loss": 1.2197,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.348643006263048e-05,
+      "loss": 0.6024,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.4008350730688935e-05,
+      "loss": 0.5704,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.4530271398747392e-05,
+      "loss": 0.4268,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.5052192066805846e-05,
+      "loss": 0.3987,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.5574112734864303e-05,
+      "loss": 1.0774,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.609603340292276e-05,
+      "loss": 0.413,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.661795407098121e-05,
+      "loss": 0.4409,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7139874739039668e-05,
+      "loss": 0.5428,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.7661795407098122e-05,
+      "loss": 0.6046,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.818371607515658e-05,
+      "loss": 0.4175,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.870563674321503e-05,
+      "loss": 1.0152,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9227557411273487e-05,
+      "loss": 0.1657,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9749478079331944e-05,
+      "loss": 1.3343,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.02713987473904e-05,
+      "loss": 0.4839,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.079331941544885e-05,
+      "loss": 0.8434,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.131524008350731e-05,
+      "loss": 0.8133,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.1837160751565766e-05,
+      "loss": 0.208,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.235908141962422e-05,
+      "loss": 0.7799,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.2881002087682673e-05,
+      "loss": 0.6431,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.340292275574113e-05,
+      "loss": 0.6077,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.392484342379959e-05,
+      "loss": 0.5476,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.444676409185804e-05,
+      "loss": 1.0333,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.4968684759916495e-05,
+      "loss": 0.9309,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.5490605427974946e-05,
+      "loss": 0.2783,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.60125260960334e-05,
+      "loss": 0.307,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.653444676409186e-05,
+      "loss": 0.779,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.705636743215031e-05,
+      "loss": 0.8502,
+      "step": 710
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.757828810020877e-05,
+      "loss": 0.6186,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.8100208768267225e-05,
+      "loss": 0.673,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.862212943632568e-05,
+      "loss": 1.1561,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.914405010438413e-05,
+      "loss": 0.7157,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.966597077244259e-05,
+      "loss": 0.3736,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.018789144050105e-05,
+      "loss": 0.6284,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.0709812108559504e-05,
+      "loss": 0.2488,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.1231732776617954e-05,
+      "loss": 0.6222,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.175365344467641e-05,
+      "loss": 0.2867,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.227557411273487e-05,
+      "loss": 0.5589,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.2797494780793326e-05,
+      "loss": 0.5885,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.3319415448851776e-05,
+      "loss": 0.8908,
+      "step": 830
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.3841336116910233e-05,
+      "loss": 0.9186,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.436325678496869e-05,
+      "loss": 1.02,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.488517745302714e-05,
+      "loss": 0.2607,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.54070981210856e-05,
+      "loss": 0.8235,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.592901878914405e-05,
+      "loss": 0.5277,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6450939457202506e-05,
+      "loss": 0.7247,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.697286012526096e-05,
+      "loss": 0.3373,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.749478079331941e-05,
+      "loss": 0.7955,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.801670146137787e-05,
+      "loss": 0.1768,
+      "step": 920
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.853862212943633e-05,
+      "loss": 0.5384,
+      "step": 930
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9060542797494785e-05,
+      "loss": 0.3238,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9582463465553235e-05,
+      "loss": 0.8556,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9988401762932035e-05,
+      "loss": 0.642,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.993041057759221e-05,
+      "loss": 0.6531,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.987241939225238e-05,
+      "loss": 0.8796,
+      "step": 980
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.981442820691255e-05,
+      "loss": 0.8854,
+      "step": 990
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.975643702157272e-05,
+      "loss": 0.3387,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.969844583623289e-05,
+      "loss": 0.156,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9640454650893064e-05,
+      "loss": 0.4904,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9582463465553235e-05,
+      "loss": 0.1614,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.952447228021341e-05,
+      "loss": 1.6102,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.946648109487358e-05,
+      "loss": 0.6737,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.940848990953375e-05,
+      "loss": 0.4079,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.935049872419393e-05,
+      "loss": 0.6687,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.92925075388541e-05,
+      "loss": 1.3816,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.923451635351427e-05,
+      "loss": 0.5562,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.917652516817444e-05,
+      "loss": 0.3573,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9118533982834613e-05,
+      "loss": 0.9735,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9060542797494785e-05,
+      "loss": 0.9598,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9002551612154956e-05,
+      "loss": 0.7512,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.894456042681513e-05,
+      "loss": 0.9617,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.88865692414753e-05,
+      "loss": 0.4649,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.882857805613547e-05,
+      "loss": 0.1344,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.877058687079564e-05,
+      "loss": 0.7923,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8712595685455813e-05,
+      "loss": 0.6594,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8654604500115985e-05,
+      "loss": 1.2267,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8596613314776156e-05,
+      "loss": 1.0072,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.853862212943633e-05,
+      "loss": 0.6871,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.84806309440965e-05,
+      "loss": 0.1132,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.842263975875667e-05,
+      "loss": 0.5345,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.836464857341684e-05,
+      "loss": 1.14,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8306657388077013e-05,
+      "loss": 0.8623,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8248666202737185e-05,
+      "loss": 0.34,
+      "step": 1260
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8190675017397356e-05,
+      "loss": 1.0127,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.813268383205753e-05,
+      "loss": 0.434,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.80746926467177e-05,
+      "loss": 0.9594,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.801670146137787e-05,
+      "loss": 1.0121,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.795871027603804e-05,
+      "loss": 1.4359,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7900719090698213e-05,
+      "loss": 1.5377,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7842727905358385e-05,
+      "loss": 0.4244,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7784736720018556e-05,
+      "loss": 1.089,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.772674553467873e-05,
+      "loss": 0.9388,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.76687543493389e-05,
+      "loss": 0.9226,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.761076316399907e-05,
+      "loss": 0.759,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.755277197865924e-05,
+      "loss": 0.7712,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.749478079331941e-05,
+      "loss": 1.5533,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7436789607979585e-05,
+      "loss": 1.1887,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7378798422639756e-05,
+      "loss": 0.4889,
+      "step": 1410
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.732080723729993e-05,
+      "loss": 1.5862,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7262816051960106e-05,
+      "loss": 2.2534,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.720482486662028e-05,
+      "loss": 2.2412,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.714683368128045e-05,
+      "loss": 2.7346,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.708884249594062e-05,
+      "loss": 1.0058,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.703085131060079e-05,
+      "loss": 0.7523,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.697286012526096e-05,
+      "loss": 1.0709,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6914868939921134e-05,
+      "loss": 1.3392,
+      "step": 1490
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6856877754581306e-05,
+      "loss": 1.0161,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.679888656924148e-05,
+      "loss": 1.4733,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.674089538390165e-05,
+      "loss": 0.5984,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.668290419856182e-05,
+      "loss": 1.2994,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.662491301322199e-05,
+      "loss": 1.619,
+      "step": 1540
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.656692182788216e-05,
+      "loss": 1.692,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6508930642542334e-05,
+      "loss": 1.6839,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6450939457202506e-05,
+      "loss": 0.7313,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.639294827186268e-05,
+      "loss": 1.7281,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.633495708652285e-05,
+      "loss": 0.7177,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.627696590118302e-05,
+      "loss": 1.2609,
+      "step": 1600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.621897471584319e-05,
+      "loss": 0.071,
+      "step": 1610
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.616098353050336e-05,
+      "loss": 0.5907,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6102992345163534e-05,
+      "loss": 0.7179,
+      "step": 1630
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6045001159823706e-05,
+      "loss": 0.9342,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.598700997448388e-05,
+      "loss": 0.7552,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.592901878914405e-05,
+      "loss": 1.373,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.587102760380422e-05,
+      "loss": 1.317,
+      "step": 1670
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.581303641846439e-05,
+      "loss": 0.5541,
+      "step": 1680
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.575504523312456e-05,
+      "loss": 1.4798,
+      "step": 1690
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5697054047784734e-05,
+      "loss": 1.0714,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5639062862444906e-05,
+      "loss": 0.8297,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.558107167710508e-05,
+      "loss": 0.5601,
+      "step": 1720
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5523080491765255e-05,
+      "loss": 1.1746,
+      "step": 1730
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.546508930642543e-05,
+      "loss": 0.9592,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.54070981210856e-05,
+      "loss": 1.7718,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.534910693574577e-05,
+      "loss": 0.5513,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.529111575040594e-05,
+      "loss": 0.7822,
+      "step": 1770
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.523312456506611e-05,
+      "loss": 0.9223,
+      "step": 1780
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5175133379726284e-05,
+      "loss": 1.8303,
+      "step": 1790
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5117142194386455e-05,
+      "loss": 1.0465,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.505915100904663e-05,
+      "loss": 0.9001,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.50011598237068e-05,
+      "loss": 1.099,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.494316863836697e-05,
+      "loss": 1.656,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.488517745302714e-05,
+      "loss": 1.4548,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.482718626768731e-05,
+      "loss": 0.8735,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.4769195082347484e-05,
+      "loss": 1.2003,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.4711203897007655e-05,
+      "loss": 1.0928,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.465321271166783e-05,
+      "loss": 0.8571,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.4595221526328e-05,
+      "loss": 1.244,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.4537230340988176e-05,
+      "loss": 1.0048,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.447923915564835e-05,
+      "loss": 1.3379,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2,
+      "eval_accuracy": 0.6337402885682575,
+      "eval_loss": 1.381054162979126,
+      "eval_runtime": 737.5675,
+      "eval_samples_per_second": 1.222,
+      "eval_steps_per_second": 0.611,
+      "step": 1916
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.442124797030852e-05,
+      "loss": 1.0315,
+      "step": 1920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.436325678496869e-05,
+      "loss": 0.8126,
+      "step": 1930
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.430526559962886e-05,
+      "loss": 1.5155,
+      "step": 1940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4247274414289033e-05,
+      "loss": 1.2548,
+      "step": 1950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4189283228949205e-05,
+      "loss": 1.4572,
+      "step": 1960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4131292043609376e-05,
+      "loss": 1.3292,
+      "step": 1970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.407330085826955e-05,
+      "loss": 0.5441,
+      "step": 1980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.401530967292972e-05,
+      "loss": 0.6136,
+      "step": 1990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.395731848758989e-05,
+      "loss": 1.4784,
+      "step": 2000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.389932730225006e-05,
+      "loss": 0.9272,
+      "step": 2010
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3841336116910233e-05,
+      "loss": 1.5402,
+      "step": 2020
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3783344931570405e-05,
+      "loss": 1.278,
+      "step": 2030
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3725353746230576e-05,
+      "loss": 1.4288,
+      "step": 2040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.366736256089075e-05,
+      "loss": 1.6097,
+      "step": 2050
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.360937137555092e-05,
+      "loss": 1.1715,
+      "step": 2060
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.355138019021109e-05,
+      "loss": 0.4054,
+      "step": 2070
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.349338900487126e-05,
+      "loss": 1.506,
+      "step": 2080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3435397819531433e-05,
+      "loss": 1.4441,
+      "step": 2090
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3377406634191605e-05,
+      "loss": 0.8217,
+      "step": 2100
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3319415448851776e-05,
+      "loss": 1.2342,
+      "step": 2110
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.326142426351195e-05,
+      "loss": 1.4318,
+      "step": 2120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.320343307817212e-05,
+      "loss": 1.3119,
+      "step": 2130
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.314544189283229e-05,
+      "loss": 0.7289,
+      "step": 2140
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.308745070749246e-05,
+      "loss": 0.5309,
+      "step": 2150
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.302945952215264e-05,
+      "loss": 1.5988,
+      "step": 2160
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.297146833681281e-05,
+      "loss": 0.4522,
+      "step": 2170
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.291347715147298e-05,
+      "loss": 0.8885,
+      "step": 2180
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.2855485966133154e-05,
+      "loss": 0.6568,
+      "step": 2190
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.2797494780793326e-05,
+      "loss": 0.8321,
+      "step": 2200
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.27395035954535e-05,
+      "loss": 0.6621,
+      "step": 2210
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.268151241011367e-05,
+      "loss": 1.6906,
+      "step": 2220
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.262352122477384e-05,
+      "loss": 1.6109,
+      "step": 2230
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.256553003943401e-05,
+      "loss": 1.2098,
+      "step": 2240
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.250753885409418e-05,
+      "loss": 1.5247,
+      "step": 2250
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.2449547668754354e-05,
+      "loss": 0.9587,
+      "step": 2260
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.2391556483414526e-05,
+      "loss": 0.7673,
+      "step": 2270
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.23335652980747e-05,
+      "loss": 1.353,
+      "step": 2280
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.227557411273487e-05,
+      "loss": 0.6869,
+      "step": 2290
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.221758292739504e-05,
+      "loss": 0.7011,
+      "step": 2300
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.215959174205521e-05,
+      "loss": 1.0921,
+      "step": 2310
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.210160055671538e-05,
+      "loss": 0.8425,
+      "step": 2320
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.2043609371375554e-05,
+      "loss": 1.7543,
+      "step": 2330
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.1985618186035726e-05,
+      "loss": 1.1886,
+      "step": 2340
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.19276270006959e-05,
+      "loss": 0.4083,
+      "step": 2350
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.186963581535607e-05,
+      "loss": 0.7883,
+      "step": 2360
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.181164463001624e-05,
+      "loss": 1.1404,
+      "step": 2370
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.175365344467641e-05,
+      "loss": 1.3592,
+      "step": 2380
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.169566225933658e-05,
+      "loss": 0.9854,
+      "step": 2390
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.1637671073996754e-05,
+      "loss": 1.5895,
+      "step": 2400
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.1579679888656926e-05,
+      "loss": 0.4981,
+      "step": 2410
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.15216887033171e-05,
+      "loss": 1.0196,
+      "step": 2420
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.146369751797727e-05,
+      "loss": 1.0117,
+      "step": 2430
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.140570633263744e-05,
+      "loss": 1.0544,
+      "step": 2440
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.134771514729761e-05,
+      "loss": 0.6361,
+      "step": 2450
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.128972396195778e-05,
+      "loss": 0.5927,
+      "step": 2460
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.1231732776617954e-05,
+      "loss": 1.4149,
+      "step": 2470
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.1173741591278126e-05,
+      "loss": 0.7116,
+      "step": 2480
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.11157504059383e-05,
+      "loss": 1.464,
+      "step": 2490
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.105775922059847e-05,
+      "loss": 0.8996,
+      "step": 2500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.099976803525864e-05,
+      "loss": 1.3305,
+      "step": 2510
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.094177684991882e-05,
+      "loss": 0.401,
+      "step": 2520
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.088378566457899e-05,
+      "loss": 1.195,
+      "step": 2530
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.082579447923916e-05,
+      "loss": 0.8744,
+      "step": 2540
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.076780329389933e-05,
+      "loss": 1.3899,
+      "step": 2550
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.0709812108559504e-05,
+      "loss": 1.2248,
+      "step": 2560
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.0651820923219675e-05,
+      "loss": 1.1285,
+      "step": 2570
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.059382973787985e-05,
+      "loss": 0.7362,
+      "step": 2580
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.053583855254002e-05,
+      "loss": 1.4619,
+      "step": 2590
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.047784736720019e-05,
+      "loss": 1.6452,
+      "step": 2600
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.041985618186036e-05,
+      "loss": 0.6238,
+      "step": 2610
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.036186499652053e-05,
+      "loss": 0.961,
+      "step": 2620
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.0303873811180704e-05,
+      "loss": 0.864,
+      "step": 2630
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.0245882625840875e-05,
+      "loss": 0.8283,
+      "step": 2640
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.018789144050105e-05,
+      "loss": 0.9317,
+      "step": 2650
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.012990025516122e-05,
+      "loss": 1.1365,
+      "step": 2660
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.007190906982139e-05,
+      "loss": 1.5571,
+      "step": 2670
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.001391788448156e-05,
+      "loss": 0.3946,
+      "step": 2680
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.995592669914173e-05,
+      "loss": 1.9972,
+      "step": 2690
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.9897935513801904e-05,
+      "loss": 1.3821,
+      "step": 2700
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.9839944328462075e-05,
+      "loss": 0.9511,
+      "step": 2710
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.978195314312225e-05,
+      "loss": 1.3779,
+      "step": 2720
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.972396195778242e-05,
+      "loss": 0.8965,
+      "step": 2730
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.966597077244259e-05,
+      "loss": 0.8617,
+      "step": 2740
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.960797958710276e-05,
+      "loss": 2.5366,
+      "step": 2750
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.954998840176293e-05,
+      "loss": 1.0511,
+      "step": 2760
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.9491997216423104e-05,
+      "loss": 0.9333,
+      "step": 2770
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.9434006031083275e-05,
+      "loss": 1.1408,
+      "step": 2780
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.937601484574345e-05,
+      "loss": 1.4736,
+      "step": 2790
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.931802366040362e-05,
+      "loss": 0.7494,
+      "step": 2800
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.926003247506379e-05,
+      "loss": 0.5487,
+      "step": 2810
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.920204128972396e-05,
+      "loss": 1.1126,
+      "step": 2820
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.914405010438413e-05,
+      "loss": 1.2111,
+      "step": 2830
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.9086058919044304e-05,
+      "loss": 0.8896,
+      "step": 2840
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.9028067733704475e-05,
+      "loss": 0.7173,
+      "step": 2850
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.897007654836465e-05,
+      "loss": 0.4547,
+      "step": 2860
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.891208536302482e-05,
+      "loss": 1.0931,
+      "step": 2870
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.8854094177684996e-05,
+      "loss": 0.8147,
+      "step": 2880
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.879610299234517e-05,
+      "loss": 1.5138,
+      "step": 2890
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.873811180700534e-05,
+      "loss": 1.8059,
+      "step": 2900
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.868012062166551e-05,
+      "loss": 1.4106,
+      "step": 2910
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.862212943632568e-05,
+      "loss": 1.7028,
+      "step": 2920
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.8564138250985853e-05,
+      "loss": 1.0171,
+      "step": 2930
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.8506147065646025e-05,
+      "loss": 1.512,
+      "step": 2940
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.8448155880306196e-05,
+      "loss": 0.9193,
+      "step": 2950
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.839016469496637e-05,
+      "loss": 0.5502,
+      "step": 2960
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.833217350962654e-05,
+      "loss": 1.1857,
+      "step": 2970
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.827418232428671e-05,
+      "loss": 0.2992,
+      "step": 2980
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.821619113894688e-05,
+      "loss": 0.2535,
+      "step": 2990
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.8158199953607053e-05,
+      "loss": 1.0116,
+      "step": 3000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.8100208768267225e-05,
+      "loss": 0.8097,
+      "step": 3010
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.8042217582927396e-05,
+      "loss": 0.6159,
+      "step": 3020
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.798422639758757e-05,
+      "loss": 1.9265,
+      "step": 3030
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.792623521224774e-05,
+      "loss": 0.9544,
+      "step": 3040
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.786824402690791e-05,
+      "loss": 0.7648,
+      "step": 3050
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.781025284156808e-05,
+      "loss": 1.2016,
+      "step": 3060
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.7752261656228253e-05,
+      "loss": 0.5182,
+      "step": 3070
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.7694270470888425e-05,
+      "loss": 0.9896,
+      "step": 3080
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.7636279285548596e-05,
+      "loss": 0.2928,
+      "step": 3090
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.757828810020877e-05,
+      "loss": 1.1354,
+      "step": 3100
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.752029691486894e-05,
+      "loss": 0.6988,
+      "step": 3110
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.746230572952911e-05,
+      "loss": 0.8192,
+      "step": 3120
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.740431454418928e-05,
+      "loss": 0.6796,
+      "step": 3130
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.734632335884945e-05,
+      "loss": 0.9249,
+      "step": 3140
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.7288332173509625e-05,
+      "loss": 1.2755,
+      "step": 3150
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.7230340988169796e-05,
+      "loss": 0.9935,
+      "step": 3160
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.717234980282997e-05,
+      "loss": 1.5134,
+      "step": 3170
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.711435861749014e-05,
+      "loss": 1.2758,
+      "step": 3180
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.705636743215031e-05,
+      "loss": 1.566,
+      "step": 3190
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.699837624681048e-05,
+      "loss": 1.3858,
+      "step": 3200
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.694038506147065e-05,
+      "loss": 0.5155,
+      "step": 3210
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.6882393876130825e-05,
+      "loss": 0.9117,
+      "step": 3220
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.6824402690790996e-05,
+      "loss": 0.2465,
+      "step": 3230
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.676641150545117e-05,
+      "loss": 1.6941,
+      "step": 3240
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.6708420320111346e-05,
+      "loss": 1.6799,
+      "step": 3250
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.665042913477152e-05,
+      "loss": 1.3417,
+      "step": 3260
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.659243794943169e-05,
+      "loss": 0.4048,
+      "step": 3270
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.653444676409186e-05,
+      "loss": 1.1219,
+      "step": 3280
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.647645557875203e-05,
+      "loss": 0.3509,
+      "step": 3290
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.64184643934122e-05,
+      "loss": 1.0548,
+      "step": 3300
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.6360473208072374e-05,
+      "loss": 1.186,
+      "step": 3310
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.6302482022732546e-05,
+      "loss": 1.7842,
+      "step": 3320
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.624449083739272e-05,
+      "loss": 1.4586,
+      "step": 3330
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.618649965205289e-05,
+      "loss": 0.6378,
+      "step": 3340
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.612850846671306e-05,
+      "loss": 1.6264,
+      "step": 3350
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.607051728137323e-05,
+      "loss": 1.2856,
+      "step": 3360
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.60125260960334e-05,
+      "loss": 0.2541,
+      "step": 3370
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.5954534910693574e-05,
+      "loss": 1.2627,
+      "step": 3380
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.5896543725353746e-05,
+      "loss": 0.3799,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.583855254001392e-05,
+      "loss": 1.2415,
+      "step": 3400
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.578056135467409e-05,
+      "loss": 1.7637,
+      "step": 3410
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.572257016933426e-05,
+      "loss": 0.7023,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.566457898399443e-05,
+      "loss": 1.1136,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.56065877986546e-05,
+      "loss": 1.3275,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.5548596613314774e-05,
+      "loss": 0.7746,
+      "step": 3450
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.5490605427974946e-05,
+      "loss": 1.4437,
+      "step": 3460
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.543261424263512e-05,
+      "loss": 0.8954,
+      "step": 3470
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.537462305729529e-05,
+      "loss": 1.1996,
+      "step": 3480
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.531663187195547e-05,
+      "loss": 0.5672,
+      "step": 3490
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.525864068661564e-05,
+      "loss": 0.427,
+      "step": 3500
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.520064950127581e-05,
+      "loss": 1.0325,
+      "step": 3510
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.514265831593598e-05,
+      "loss": 0.6108,
+      "step": 3520
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.508466713059615e-05,
+      "loss": 0.8856,
+      "step": 3530
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.5026675945256324e-05,
+      "loss": 1.0719,
+      "step": 3540
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.4968684759916495e-05,
+      "loss": 1.1319,
+      "step": 3550
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.491069357457667e-05,
+      "loss": 0.9198,
+      "step": 3560
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.485270238923684e-05,
+      "loss": 1.1559,
+      "step": 3570
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.479471120389701e-05,
+      "loss": 0.961,
+      "step": 3580
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.473672001855718e-05,
+      "loss": 0.6364,
+      "step": 3590
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.467872883321735e-05,
+      "loss": 0.1699,
+      "step": 3600
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.4620737647877524e-05,
+      "loss": 0.9015,
+      "step": 3610
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.4562746462537695e-05,
+      "loss": 0.7862,
+      "step": 3620
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.450475527719787e-05,
+      "loss": 0.5483,
+      "step": 3630
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.444676409185804e-05,
+      "loss": 1.2363,
+      "step": 3640
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.438877290651821e-05,
+      "loss": 1.4381,
+      "step": 3650
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.433078172117839e-05,
+      "loss": 0.1331,
+      "step": 3660
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.427279053583856e-05,
+      "loss": 0.1726,
+      "step": 3670
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.421479935049873e-05,
+      "loss": 0.7826,
+      "step": 3680
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.41568081651589e-05,
+      "loss": 0.7392,
+      "step": 3690
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.4098816979819073e-05,
+      "loss": 0.6422,
+      "step": 3700
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.4040825794479245e-05,
+      "loss": 0.3669,
+      "step": 3710
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.3982834609139416e-05,
+      "loss": 1.2428,
+      "step": 3720
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.392484342379959e-05,
+      "loss": 2.4697,
+      "step": 3730
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.386685223845976e-05,
+      "loss": 0.3838,
+      "step": 3740
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.380886105311993e-05,
+      "loss": 0.7432,
+      "step": 3750
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.37508698677801e-05,
+      "loss": 0.8939,
+      "step": 3760
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.3692878682440273e-05,
+      "loss": 0.2091,
+      "step": 3770
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.3634887497100445e-05,
+      "loss": 1.006,
+      "step": 3780
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.3576896311760616e-05,
+      "loss": 0.4391,
+      "step": 3790
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.351890512642079e-05,
+      "loss": 1.4439,
+      "step": 3800
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.346091394108096e-05,
+      "loss": 1.1189,
+      "step": 3810
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.340292275574113e-05,
+      "loss": 0.8565,
+      "step": 3820
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.33449315704013e-05,
+      "loss": 0.8188,
+      "step": 3830
+    },
+    {
+      "epoch": 1.2,
+      "eval_accuracy": 0.5271920088790233,
+      "eval_loss": 1.8434526920318604,
+      "eval_runtime": 885.482,
+      "eval_samples_per_second": 1.018,
+      "eval_steps_per_second": 0.509,
+      "step": 3832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.3286940385061473e-05,
+      "loss": 1.2624,
+      "step": 3840
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.3228949199721645e-05,
+      "loss": 0.9132,
+      "step": 3850
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.3170958014381816e-05,
+      "loss": 0.5228,
+      "step": 3860
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.311296682904199e-05,
+      "loss": 0.6028,
+      "step": 3870
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.305497564370216e-05,
+      "loss": 0.746,
+      "step": 3880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.299698445836233e-05,
+      "loss": 1.1898,
+      "step": 3890
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.29389932730225e-05,
+      "loss": 1.0801,
+      "step": 3900
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2881002087682673e-05,
+      "loss": 0.2822,
+      "step": 3910
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2823010902342845e-05,
+      "loss": 0.753,
+      "step": 3920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2765019717003016e-05,
+      "loss": 1.2804,
+      "step": 3930
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.270702853166319e-05,
+      "loss": 1.2124,
+      "step": 3940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.264903734632336e-05,
+      "loss": 0.6875,
+      "step": 3950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.259104616098353e-05,
+      "loss": 1.9392,
+      "step": 3960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.253305497564371e-05,
+      "loss": 1.163,
+      "step": 3970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.247506379030388e-05,
+      "loss": 1.1741,
+      "step": 3980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.241707260496405e-05,
+      "loss": 1.0555,
+      "step": 3990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.235908141962422e-05,
+      "loss": 1.1706,
+      "step": 4000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.2301090234284394e-05,
+      "loss": 0.597,
+      "step": 4010
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.2243099048944566e-05,
+      "loss": 1.038,
+      "step": 4020
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.218510786360474e-05,
+      "loss": 1.0297,
+      "step": 4030
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.212711667826491e-05,
+      "loss": 1.5967,
+      "step": 4040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.206912549292508e-05,
+      "loss": 1.2922,
+      "step": 4050
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.201113430758525e-05,
+      "loss": 0.8386,
+      "step": 4060
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.195314312224542e-05,
+      "loss": 1.5293,
+      "step": 4070
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.1895151936905594e-05,
+      "loss": 1.2095,
+      "step": 4080
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.1837160751565766e-05,
+      "loss": 0.9311,
+      "step": 4090
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.177916956622594e-05,
+      "loss": 1.4232,
+      "step": 4100
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.172117838088611e-05,
+      "loss": 0.548,
+      "step": 4110
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.166318719554628e-05,
+      "loss": 1.4026,
+      "step": 4120
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.160519601020645e-05,
+      "loss": 0.5013,
+      "step": 4130
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.154720482486662e-05,
+      "loss": 0.2867,
+      "step": 4140
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.1489213639526794e-05,
+      "loss": 1.0079,
+      "step": 4150
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.1431222454186966e-05,
+      "loss": 1.415,
+      "step": 4160
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.137323126884714e-05,
+      "loss": 0.6958,
+      "step": 4170
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.131524008350731e-05,
+      "loss": 0.9422,
+      "step": 4180
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.125724889816748e-05,
+      "loss": 0.8432,
+      "step": 4190
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.119925771282765e-05,
+      "loss": 1.1306,
+      "step": 4200
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.114126652748782e-05,
+      "loss": 1.8626,
+      "step": 4210
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.1083275342147994e-05,
+      "loss": 1.1103,
+      "step": 4220
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.1025284156808166e-05,
+      "loss": 1.1079,
+      "step": 4230
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.096729297146834e-05,
+      "loss": 1.2238,
+      "step": 4240
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.090930178612851e-05,
+      "loss": 1.8076,
+      "step": 4250
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.085131060078868e-05,
+      "loss": 0.8964,
+      "step": 4260
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.079331941544885e-05,
+      "loss": 0.6576,
+      "step": 4270
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.073532823010902e-05,
+      "loss": 1.6563,
+      "step": 4280
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.0677337044769194e-05,
+      "loss": 0.2634,
+      "step": 4290
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.0619345859429366e-05,
+      "loss": 0.1947,
+      "step": 4300
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.056135467408954e-05,
+      "loss": 0.9497,
+      "step": 4310
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.050336348874971e-05,
+      "loss": 0.3522,
+      "step": 4320
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.0445372303409887e-05,
+      "loss": 1.5495,
+      "step": 4330
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.0387381118070058e-05,
+      "loss": 0.4288,
+      "step": 4340
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.032938993273023e-05,
+      "loss": 0.2865,
+      "step": 4350
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.02713987473904e-05,
+      "loss": 0.9748,
+      "step": 4360
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.0213407562050572e-05,
+      "loss": 0.9998,
+      "step": 4370
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.0155416376710744e-05,
+      "loss": 0.9992,
+      "step": 4380
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.0097425191370915e-05,
+      "loss": 0.8725,
+      "step": 4390
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.0039434006031087e-05,
+      "loss": 0.7761,
+      "step": 4400
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.9981442820691258e-05,
+      "loss": 1.4317,
+      "step": 4410
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.992345163535143e-05,
+      "loss": 1.267,
+      "step": 4420
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.98654604500116e-05,
+      "loss": 1.0834,
+      "step": 4430
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.9807469264671772e-05,
+      "loss": 1.8092,
+      "step": 4440
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.9749478079331944e-05,
+      "loss": 1.37,
+      "step": 4450
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9691486893992115e-05,
+      "loss": 0.6262,
+      "step": 4460
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9633495708652287e-05,
+      "loss": 0.5533,
+      "step": 4470
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9575504523312458e-05,
+      "loss": 0.6107,
+      "step": 4480
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.951751333797263e-05,
+      "loss": 0.643,
+      "step": 4490
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.94595221526328e-05,
+      "loss": 1.1418,
+      "step": 4500
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9401530967292972e-05,
+      "loss": 0.7036,
+      "step": 4510
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9343539781953144e-05,
+      "loss": 1.5465,
+      "step": 4520
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9285548596613315e-05,
+      "loss": 0.5909,
+      "step": 4530
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9227557411273487e-05,
+      "loss": 0.7398,
+      "step": 4540
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9169566225933658e-05,
+      "loss": 1.5429,
+      "step": 4550
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.911157504059383e-05,
+      "loss": 0.5007,
+      "step": 4560
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.9053583855254e-05,
+      "loss": 0.8971,
+      "step": 4570
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8995592669914172e-05,
+      "loss": 0.7796,
+      "step": 4580
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8937601484574344e-05,
+      "loss": 1.4779,
+      "step": 4590
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8879610299234515e-05,
+      "loss": 0.9193,
+      "step": 4600
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8821619113894687e-05,
+      "loss": 0.9528,
+      "step": 4610
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8763627928554858e-05,
+      "loss": 0.7458,
+      "step": 4620
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.870563674321503e-05,
+      "loss": 0.516,
+      "step": 4630
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.86476455578752e-05,
+      "loss": 1.6264,
+      "step": 4640
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8589654372535372e-05,
+      "loss": 0.9986,
+      "step": 4650
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8531663187195544e-05,
+      "loss": 0.7311,
+      "step": 4660
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8473672001855715e-05,
+      "loss": 0.9719,
+      "step": 4670
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8415680816515887e-05,
+      "loss": 0.8531,
+      "step": 4680
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8357689631176065e-05,
+      "loss": 1.0132,
+      "step": 4690
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8299698445836236e-05,
+      "loss": 0.3517,
+      "step": 4700
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8241707260496408e-05,
+      "loss": 0.9351,
+      "step": 4710
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.818371607515658e-05,
+      "loss": 1.0301,
+      "step": 4720
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.812572488981675e-05,
+      "loss": 0.997,
+      "step": 4730
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8067733704476922e-05,
+      "loss": 0.9708,
+      "step": 4740
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.8009742519137093e-05,
+      "loss": 1.0523,
+      "step": 4750
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7951751333797265e-05,
+      "loss": 0.6657,
+      "step": 4760
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7893760148457436e-05,
+      "loss": 0.5143,
+      "step": 4770
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7835768963117608e-05,
+      "loss": 0.2768,
+      "step": 4780
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.6623,
+      "step": 4790
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.771978659243795e-05,
+      "loss": 0.7383,
+      "step": 4800
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7661795407098122e-05,
+      "loss": 1.1703,
+      "step": 4810
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7603804221758293e-05,
+      "loss": 0.9627,
+      "step": 4820
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7545813036418465e-05,
+      "loss": 0.9261,
+      "step": 4830
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7487821851078636e-05,
+      "loss": 0.8958,
+      "step": 4840
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7429830665738808e-05,
+      "loss": 0.7933,
+      "step": 4850
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7371839480398982e-05,
+      "loss": 1.0833,
+      "step": 4860
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7313848295059154e-05,
+      "loss": 0.76,
+      "step": 4870
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7255857109719325e-05,
+      "loss": 0.8355,
+      "step": 4880
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7197865924379497e-05,
+      "loss": 0.5321,
+      "step": 4890
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7139874739039668e-05,
+      "loss": 0.4399,
+      "step": 4900
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.708188355369984e-05,
+      "loss": 0.9979,
+      "step": 4910
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.702389236836001e-05,
+      "loss": 1.2286,
+      "step": 4920
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.6965901183020182e-05,
+      "loss": 1.4501,
+      "step": 4930
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6907909997680354e-05,
+      "loss": 1.0885,
+      "step": 4940
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6849918812340525e-05,
+      "loss": 1.1784,
+      "step": 4950
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6791927627000697e-05,
+      "loss": 1.17,
+      "step": 4960
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6733936441660868e-05,
+      "loss": 0.5863,
+      "step": 4970
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.667594525632104e-05,
+      "loss": 0.6242,
+      "step": 4980
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.661795407098121e-05,
+      "loss": 0.9164,
+      "step": 4990
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6559962885641382e-05,
+      "loss": 0.6847,
+      "step": 5000
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6501971700301554e-05,
+      "loss": 0.4725,
+      "step": 5010
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6443980514961725e-05,
+      "loss": 0.708,
+      "step": 5020
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6385989329621897e-05,
+      "loss": 0.7712,
+      "step": 5030
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6327998144282068e-05,
+      "loss": 0.7797,
+      "step": 5040
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.627000695894224e-05,
+      "loss": 1.2895,
+      "step": 5050
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6212015773602418e-05,
+      "loss": 0.8245,
+      "step": 5060
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.615402458826259e-05,
+      "loss": 0.9187,
+      "step": 5070
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.609603340292276e-05,
+      "loss": 0.5652,
+      "step": 5080
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6038042217582932e-05,
+      "loss": 0.9337,
+      "step": 5090
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.5980051032243103e-05,
+      "loss": 0.3152,
+      "step": 5100
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.5922059846903275e-05,
+      "loss": 0.4042,
+      "step": 5110
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.5864068661563446e-05,
+      "loss": 1.3384,
+      "step": 5120
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.5806077476223618e-05,
+      "loss": 1.2538,
+      "step": 5130
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.574808629088379e-05,
+      "loss": 0.818,
+      "step": 5140
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.569009510554396e-05,
+      "loss": 0.1941,
+      "step": 5150
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.5632103920204132e-05,
+      "loss": 0.2833,
+      "step": 5160
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.5574112734864303e-05,
+      "loss": 1.274,
+      "step": 5170
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.5516121549524475e-05,
+      "loss": 0.7784,
+      "step": 5180
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.5458130364184646e-05,
+      "loss": 0.9075,
+      "step": 5190
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.5400139178844818e-05,
+      "loss": 1.1002,
+      "step": 5200
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.534214799350499e-05,
+      "loss": 1.2428,
+      "step": 5210
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.528415680816516e-05,
+      "loss": 0.2979,
+      "step": 5220
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.5226165622825332e-05,
+      "loss": 0.8655,
+      "step": 5230
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.5168174437485503e-05,
+      "loss": 0.8901,
+      "step": 5240
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.5110183252145675e-05,
+      "loss": 0.2868,
+      "step": 5250
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.5052192066805846e-05,
+      "loss": 1.1536,
+      "step": 5260
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.4994200881466018e-05,
+      "loss": 1.4035,
+      "step": 5270
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.493620969612619e-05,
+      "loss": 0.8101,
+      "step": 5280
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.487821851078636e-05,
+      "loss": 1.0066,
+      "step": 5290
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.4820227325446532e-05,
+      "loss": 0.6675,
+      "step": 5300
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.4762236140106703e-05,
+      "loss": 1.4472,
+      "step": 5310
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.4704244954766875e-05,
+      "loss": 0.9359,
+      "step": 5320
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.464625376942705e-05,
+      "loss": 1.6417,
+      "step": 5330
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.458826258408722e-05,
+      "loss": 0.5859,
+      "step": 5340
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.4530271398747392e-05,
+      "loss": 0.9326,
+      "step": 5350
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.4472280213407564e-05,
+      "loss": 0.9925,
+      "step": 5360
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.4414289028067735e-05,
+      "loss": 1.0956,
+      "step": 5370
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.4356297842727907e-05,
+      "loss": 0.9185,
+      "step": 5380
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.4298306657388078e-05,
+      "loss": 0.4399,
+      "step": 5390
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.424031547204825e-05,
+      "loss": 0.7184,
+      "step": 5400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.418232428670842e-05,
+      "loss": 1.1189,
+      "step": 5410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.4124333101368592e-05,
+      "loss": 0.6928,
+      "step": 5420
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.4066341916028764e-05,
+      "loss": 0.8546,
+      "step": 5430
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.4008350730688935e-05,
+      "loss": 1.1813,
+      "step": 5440
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.3950359545349107e-05,
+      "loss": 1.7317,
+      "step": 5450
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.3892368360009278e-05,
+      "loss": 1.102,
+      "step": 5460
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.383437717466945e-05,
+      "loss": 0.8529,
+      "step": 5470
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.377638598932962e-05,
+      "loss": 1.0327,
+      "step": 5480
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.3718394803989792e-05,
+      "loss": 0.7323,
+      "step": 5490
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.3660403618649964e-05,
+      "loss": 1.0322,
+      "step": 5500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.360241243331014e-05,
+      "loss": 0.8035,
+      "step": 5510
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.354442124797031e-05,
+      "loss": 0.3781,
+      "step": 5520
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.348643006263048e-05,
+      "loss": 1.0022,
+      "step": 5530
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.3428438877290653e-05,
+      "loss": 0.7912,
+      "step": 5540
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.3370447691950824e-05,
+      "loss": 0.519,
+      "step": 5550
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.3312456506610996e-05,
+      "loss": 0.5316,
+      "step": 5560
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.3254465321271167e-05,
+      "loss": 0.529,
+      "step": 5570
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.319647413593134e-05,
+      "loss": 0.4809,
+      "step": 5580
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.313848295059151e-05,
+      "loss": 0.2599,
+      "step": 5590
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.308049176525168e-05,
+      "loss": 1.0648,
+      "step": 5600
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.3022500579911853e-05,
+      "loss": 1.9211,
+      "step": 5610
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.2964509394572024e-05,
+      "loss": 0.8381,
+      "step": 5620
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.2906518209232196e-05,
+      "loss": 0.658,
+      "step": 5630
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.2848527023892367e-05,
+      "loss": 1.0075,
+      "step": 5640
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.279053583855254e-05,
+      "loss": 1.0003,
+      "step": 5650
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.2732544653212713e-05,
+      "loss": 0.9389,
+      "step": 5660
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.2674553467872885e-05,
+      "loss": 0.9876,
+      "step": 5670
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.2616562282533056e-05,
+      "loss": 1.1354,
+      "step": 5680
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.2558571097193228e-05,
+      "loss": 0.6671,
+      "step": 5690
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.25005799118534e-05,
+      "loss": 0.8593,
+      "step": 5700
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.244258872651357e-05,
+      "loss": 0.5948,
+      "step": 5710
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.2384597541173742e-05,
+      "loss": 1.1783,
+      "step": 5720
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.2326606355833913e-05,
+      "loss": 0.5499,
+      "step": 5730
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.2268615170494088e-05,
+      "loss": 0.5813,
+      "step": 5740
+    },
+    {
+      "epoch": 2.2,
+      "eval_accuracy": 0.7203107658157603,
+      "eval_loss": 1.1628159284591675,
+      "eval_runtime": 820.3164,
+      "eval_samples_per_second": 1.098,
+      "eval_steps_per_second": 0.55,
+      "step": 5748
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.221062398515426e-05,
+      "loss": 1.2559,
+      "step": 5750
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.215263279981443e-05,
+      "loss": 1.9458,
+      "step": 5760
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.2094641614474602e-05,
+      "loss": 0.3845,
+      "step": 5770
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.2036650429134774e-05,
+      "loss": 0.4935,
+      "step": 5780
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.1978659243794945e-05,
+      "loss": 0.9517,
+      "step": 5790
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.1920668058455117e-05,
+      "loss": 0.7981,
+      "step": 5800
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.1862676873115288e-05,
+      "loss": 0.6797,
+      "step": 5810
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.180468568777546e-05,
+      "loss": 1.3161,
+      "step": 5820
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.174669450243563e-05,
+      "loss": 1.4221,
+      "step": 5830
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.1688703317095802e-05,
+      "loss": 0.5534,
+      "step": 5840
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.1630712131755974e-05,
+      "loss": 1.1825,
+      "step": 5850
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.1572720946416145e-05,
+      "loss": 1.1031,
+      "step": 5860
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.151472976107632e-05,
+      "loss": 0.687,
+      "step": 5870
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.145673857573649e-05,
+      "loss": 1.3767,
+      "step": 5880
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.1398747390396663e-05,
+      "loss": 1.2661,
+      "step": 5890
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.1340756205056834e-05,
+      "loss": 1.0882,
+      "step": 5900
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.1282765019717006e-05,
+      "loss": 0.1892,
+      "step": 5910
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.1224773834377177e-05,
+      "loss": 1.8277,
+      "step": 5920
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.116678264903735e-05,
+      "loss": 0.5879,
+      "step": 5930
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.110879146369752e-05,
+      "loss": 1.3157,
+      "step": 5940
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.105080027835769e-05,
+      "loss": 0.9038,
+      "step": 5950
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.0992809093017863e-05,
+      "loss": 1.0074,
+      "step": 5960
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.0934817907678034e-05,
+      "loss": 0.44,
+      "step": 5970
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.0876826722338206e-05,
+      "loss": 0.5945,
+      "step": 5980
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.0818835536998377e-05,
+      "loss": 1.517,
+      "step": 5990
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.076084435165855e-05,
+      "loss": 1.3077,
+      "step": 6000
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.070285316631872e-05,
+      "loss": 1.0437,
+      "step": 6010
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.064486198097889e-05,
+      "loss": 1.0558,
+      "step": 6020
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.0586870795639063e-05,
+      "loss": 0.8743,
+      "step": 6030
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.0528879610299234e-05,
+      "loss": 1.0091,
+      "step": 6040
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.047088842495941e-05,
+      "loss": 0.4611,
+      "step": 6050
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.041289723961958e-05,
+      "loss": 0.518,
+      "step": 6060
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.0354906054279752e-05,
+      "loss": 1.0173,
+      "step": 6070
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.0296914868939923e-05,
+      "loss": 2.2123,
+      "step": 6080
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.0238923683600095e-05,
+      "loss": 1.2601,
+      "step": 6090
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.0180932498260266e-05,
+      "loss": 0.6831,
+      "step": 6100
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.0122941312920438e-05,
+      "loss": 0.8897,
+      "step": 6110
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.006495012758061e-05,
+      "loss": 1.1531,
+      "step": 6120
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.000695894224078e-05,
+      "loss": 0.6148,
+      "step": 6130
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.9948967756900952e-05,
+      "loss": 0.5459,
+      "step": 6140
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.9890976571561123e-05,
+      "loss": 0.1905,
+      "step": 6150
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.9832985386221295e-05,
+      "loss": 0.6639,
+      "step": 6160
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.9774994200881466e-05,
+      "loss": 0.7672,
+      "step": 6170
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9717003015541638e-05,
+      "loss": 1.4976,
+      "step": 6180
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.965901183020181e-05,
+      "loss": 1.3514,
+      "step": 6190
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.960102064486198e-05,
+      "loss": 0.8602,
+      "step": 6200
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9543029459522152e-05,
+      "loss": 1.0355,
+      "step": 6210
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9485038274182323e-05,
+      "loss": 0.4425,
+      "step": 6220
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9427047088842498e-05,
+      "loss": 0.5906,
+      "step": 6230
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.936905590350267e-05,
+      "loss": 0.8006,
+      "step": 6240
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.931106471816284e-05,
+      "loss": 1.345,
+      "step": 6250
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9253073532823012e-05,
+      "loss": 0.8586,
+      "step": 6260
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9195082347483184e-05,
+      "loss": 0.7298,
+      "step": 6270
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.9137091162143355e-05,
+      "loss": 0.3302,
+      "step": 6280
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.9079099976803527e-05,
+      "loss": 0.7833,
+      "step": 6290
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.9021108791463698e-05,
+      "loss": 1.8514,
+      "step": 6300
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.896311760612387e-05,
+      "loss": 1.5884,
+      "step": 6310
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.890512642078404e-05,
+      "loss": 0.1516,
+      "step": 6320
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.8847135235444212e-05,
+      "loss": 1.0005,
+      "step": 6330
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.8789144050104384e-05,
+      "loss": 0.3564,
+      "step": 6340
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.8731152864764555e-05,
+      "loss": 1.3575,
+      "step": 6350
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.8673161679424727e-05,
+      "loss": 0.9023,
+      "step": 6360
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.8615170494084898e-05,
+      "loss": 0.6841,
+      "step": 6370
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.855717930874507e-05,
+      "loss": 1.158,
+      "step": 6380
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.849918812340524e-05,
+      "loss": 0.9449,
+      "step": 6390
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.8441196938065412e-05,
+      "loss": 0.6163,
+      "step": 6400
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.8383205752725584e-05,
+      "loss": 0.8521,
+      "step": 6410
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.832521456738576e-05,
+      "loss": 0.9943,
+      "step": 6420
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.826722338204593e-05,
+      "loss": 0.9589,
+      "step": 6430
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.82092321967061e-05,
+      "loss": 0.4541,
+      "step": 6440
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.8151241011366273e-05,
+      "loss": 1.9594,
+      "step": 6450
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.8093249826026444e-05,
+      "loss": 0.8652,
+      "step": 6460
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.8035258640686616e-05,
+      "loss": 0.9064,
+      "step": 6470
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.7977267455346787e-05,
+      "loss": 0.6425,
+      "step": 6480
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.791927627000696e-05,
+      "loss": 1.0613,
+      "step": 6490
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.786128508466713e-05,
+      "loss": 0.8762,
+      "step": 6500
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.78032938993273e-05,
+      "loss": 0.473,
+      "step": 6510
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.7745302713987473e-05,
+      "loss": 0.7356,
+      "step": 6520
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.7687311528647644e-05,
+      "loss": 1.1862,
+      "step": 6530
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.762932034330782e-05,
+      "loss": 0.2934,
+      "step": 6540
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.757132915796799e-05,
+      "loss": 0.4302,
+      "step": 6550
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.7513337972628162e-05,
+      "loss": 0.9309,
+      "step": 6560
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.7455346787288333e-05,
+      "loss": 0.9268,
+      "step": 6570
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.7397355601948505e-05,
+      "loss": 0.6812,
+      "step": 6580
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.7339364416608676e-05,
+      "loss": 0.2455,
+      "step": 6590
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.7281373231268848e-05,
+      "loss": 1.0947,
+      "step": 6600
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.722338204592902e-05,
+      "loss": 0.564,
+      "step": 6610
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.7165390860589194e-05,
+      "loss": 0.8476,
+      "step": 6620
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.7107399675249365e-05,
+      "loss": 0.8956,
+      "step": 6630
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.7049408489909537e-05,
+      "loss": 0.2584,
+      "step": 6640
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.6991417304569708e-05,
+      "loss": 0.3439,
+      "step": 6650
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.693342611922988e-05,
+      "loss": 0.6293,
+      "step": 6660
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.687543493389005e-05,
+      "loss": 1.0799,
+      "step": 6670
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.6817443748550222e-05,
+      "loss": 1.2609,
+      "step": 6680
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.6759452563210394e-05,
+      "loss": 0.3787,
+      "step": 6690
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.6701461377870565e-05,
+      "loss": 0.9644,
+      "step": 6700
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.6643470192530737e-05,
+      "loss": 0.8635,
+      "step": 6710
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.6585479007190908e-05,
+      "loss": 0.5651,
+      "step": 6720
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.652748782185108e-05,
+      "loss": 0.5858,
+      "step": 6730
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.646949663651125e-05,
+      "loss": 1.2277,
+      "step": 6740
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.6411505451171422e-05,
+      "loss": 0.821,
+      "step": 6750
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.6353514265831594e-05,
+      "loss": 1.0182,
+      "step": 6760
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.6295523080491765e-05,
+      "loss": 1.0026,
+      "step": 6770
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.623753189515194e-05,
+      "loss": 1.1589,
+      "step": 6780
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.617954070981211e-05,
+      "loss": 0.9319,
+      "step": 6790
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.6121549524472283e-05,
+      "loss": 1.019,
+      "step": 6800
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.6063558339132454e-05,
+      "loss": 0.4481,
+      "step": 6810
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.6005567153792626e-05,
+      "loss": 1.4031,
+      "step": 6820
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.5947575968452797e-05,
+      "loss": 0.6644,
+      "step": 6830
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.588958478311297e-05,
+      "loss": 1.3141,
+      "step": 6840
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.583159359777314e-05,
+      "loss": 0.6123,
+      "step": 6850
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.577360241243331e-05,
+      "loss": 1.1272,
+      "step": 6860
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5715611227093483e-05,
+      "loss": 1.3676,
+      "step": 6870
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5657620041753654e-05,
+      "loss": 0.6766,
+      "step": 6880
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5599628856413826e-05,
+      "loss": 0.7709,
+      "step": 6890
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5541637671073997e-05,
+      "loss": 1.1335,
+      "step": 6900
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.548364648573417e-05,
+      "loss": 0.9222,
+      "step": 6910
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.542565530039434e-05,
+      "loss": 0.0981,
+      "step": 6920
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.536766411505451e-05,
+      "loss": 0.8834,
+      "step": 6930
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5309672929714683e-05,
+      "loss": 0.4818,
+      "step": 6940
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.5251681744374854e-05,
+      "loss": 1.1843,
+      "step": 6950
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.5193690559035029e-05,
+      "loss": 1.1047,
+      "step": 6960
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.51356993736952e-05,
+      "loss": 0.4618,
+      "step": 6970
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.5077708188355372e-05,
+      "loss": 0.6285,
+      "step": 6980
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.5019717003015543e-05,
+      "loss": 0.8472,
+      "step": 6990
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.4961725817675715e-05,
+      "loss": 0.745,
+      "step": 7000
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.4903734632335886e-05,
+      "loss": 0.5418,
+      "step": 7010
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.4845743446996058e-05,
+      "loss": 0.6041,
+      "step": 7020
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.4787752261656229e-05,
+      "loss": 1.0065,
+      "step": 7030
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.47297610763164e-05,
+      "loss": 1.1153,
+      "step": 7040
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.4671769890976572e-05,
+      "loss": 0.838,
+      "step": 7050
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.4613778705636743e-05,
+      "loss": 1.3795,
+      "step": 7060
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.4555787520296915e-05,
+      "loss": 0.8657,
+      "step": 7070
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.4497796334957086e-05,
+      "loss": 1.1964,
+      "step": 7080
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.4439805149617258e-05,
+      "loss": 1.3427,
+      "step": 7090
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.4381813964277429e-05,
+      "loss": 0.6488,
+      "step": 7100
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.43238227789376e-05,
+      "loss": 1.2859,
+      "step": 7110
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.4265831593597772e-05,
+      "loss": 1.0861,
+      "step": 7120
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.4207840408257943e-05,
+      "loss": 0.1797,
+      "step": 7130
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.4149849222918118e-05,
+      "loss": 0.5945,
+      "step": 7140
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.409185803757829e-05,
+      "loss": 0.9491,
+      "step": 7150
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.4033866852238461e-05,
+      "loss": 0.5122,
+      "step": 7160
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.3975875666898632e-05,
+      "loss": 0.7074,
+      "step": 7170
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.3917884481558804e-05,
+      "loss": 0.8528,
+      "step": 7180
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.3859893296218975e-05,
+      "loss": 0.9367,
+      "step": 7190
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.3801902110879147e-05,
+      "loss": 0.8921,
+      "step": 7200
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.3743910925539318e-05,
+      "loss": 1.4556,
+      "step": 7210
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.3685919740199491e-05,
+      "loss": 0.7498,
+      "step": 7220
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.3627928554859663e-05,
+      "loss": 1.1415,
+      "step": 7230
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.3569937369519834e-05,
+      "loss": 0.862,
+      "step": 7240
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.3511946184180006e-05,
+      "loss": 0.5544,
+      "step": 7250
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.3453954998840177e-05,
+      "loss": 1.2962,
+      "step": 7260
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.3395963813500348e-05,
+      "loss": 1.706,
+      "step": 7270
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.333797262816052e-05,
+      "loss": 1.297,
+      "step": 7280
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.3279981442820691e-05,
+      "loss": 1.6328,
+      "step": 7290
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.3221990257480863e-05,
+      "loss": 1.5138,
+      "step": 7300
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.3163999072141034e-05,
+      "loss": 0.7456,
+      "step": 7310
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.3106007886801209e-05,
+      "loss": 1.9206,
+      "step": 7320
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.304801670146138e-05,
+      "loss": 1.7376,
+      "step": 7330
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2990025516121552e-05,
+      "loss": 0.2005,
+      "step": 7340
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2932034330781723e-05,
+      "loss": 0.407,
+      "step": 7350
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2874043145441895e-05,
+      "loss": 0.7395,
+      "step": 7360
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2816051960102066e-05,
+      "loss": 0.6776,
+      "step": 7370
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2758060774762237e-05,
+      "loss": 0.2717,
+      "step": 7380
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2700069589422409e-05,
+      "loss": 1.4556,
+      "step": 7390
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.264207840408258e-05,
+      "loss": 0.245,
+      "step": 7400
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2584087218742752e-05,
+      "loss": 1.8686,
+      "step": 7410
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2526096033402923e-05,
+      "loss": 0.8555,
+      "step": 7420
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2468104848063095e-05,
+      "loss": 1.0947,
+      "step": 7430
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2410113662723266e-05,
+      "loss": 0.9371,
+      "step": 7440
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2352122477383437e-05,
+      "loss": 1.1081,
+      "step": 7450
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.229413129204361e-05,
+      "loss": 0.8434,
+      "step": 7460
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2236140106703782e-05,
+      "loss": 1.2701,
+      "step": 7470
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2178148921363953e-05,
+      "loss": 1.2144,
+      "step": 7480
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2120157736024125e-05,
+      "loss": 1.0603,
+      "step": 7490
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2062166550684296e-05,
+      "loss": 1.0112,
+      "step": 7500
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2004175365344468e-05,
+      "loss": 1.1033,
+      "step": 7510
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.1946184180004639e-05,
+      "loss": 0.7898,
+      "step": 7520
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.188819299466481e-05,
+      "loss": 1.0465,
+      "step": 7530
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.1830201809324982e-05,
+      "loss": 0.9565,
+      "step": 7540
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.1772210623985155e-05,
+      "loss": 1.3477,
+      "step": 7550
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.1714219438645326e-05,
+      "loss": 1.5173,
+      "step": 7560
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.1656228253305498e-05,
+      "loss": 0.4567,
+      "step": 7570
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.159823706796567e-05,
+      "loss": 0.3476,
+      "step": 7580
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.154024588262584e-05,
+      "loss": 1.3395,
+      "step": 7590
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.1482254697286012e-05,
+      "loss": 1.0687,
+      "step": 7600
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.1424263511946184e-05,
+      "loss": 0.4325,
+      "step": 7610
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.1366272326606357e-05,
+      "loss": 1.0135,
+      "step": 7620
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.1308281141266528e-05,
+      "loss": 1.0043,
+      "step": 7630
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.12502899559267e-05,
+      "loss": 0.4354,
+      "step": 7640
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.1192298770586871e-05,
+      "loss": 0.93,
+      "step": 7650
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.1134307585247044e-05,
+      "loss": 0.5712,
+      "step": 7660
+    },
+    {
+      "epoch": 3.2,
+      "eval_accuracy": 0.6970033296337403,
+      "eval_loss": 1.1364027261734009,
+      "eval_runtime": 591.9957,
+      "eval_samples_per_second": 1.522,
+      "eval_steps_per_second": 0.762,
+      "step": 7664
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.1076316399907216e-05,
+      "loss": 0.8007,
+      "step": 7670
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.1018325214567387e-05,
+      "loss": 0.7158,
+      "step": 7680
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.0960334029227558e-05,
+      "loss": 0.6792,
+      "step": 7690
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.090234284388773e-05,
+      "loss": 0.8184,
+      "step": 7700
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.0844351658547901e-05,
+      "loss": 0.6741,
+      "step": 7710
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.0786360473208073e-05,
+      "loss": 1.6746,
+      "step": 7720
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.0728369287868246e-05,
+      "loss": 0.8906,
+      "step": 7730
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.0670378102528417e-05,
+      "loss": 0.5721,
+      "step": 7740
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.0612386917188589e-05,
+      "loss": 0.5313,
+      "step": 7750
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.055439573184876e-05,
+      "loss": 0.3419,
+      "step": 7760
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.0496404546508931e-05,
+      "loss": 0.7249,
+      "step": 7770
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.0438413361169103e-05,
+      "loss": 1.1336,
+      "step": 7780
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.0380422175829274e-05,
+      "loss": 0.6763,
+      "step": 7790
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.0322430990489446e-05,
+      "loss": 1.009,
+      "step": 7800
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.0264439805149617e-05,
+      "loss": 1.0939,
+      "step": 7810
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.020644861980979e-05,
+      "loss": 0.7127,
+      "step": 7820
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.0148457434469962e-05,
+      "loss": 1.543,
+      "step": 7830
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.0090466249130133e-05,
+      "loss": 0.5289,
+      "step": 7840
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.0032475063790305e-05,
+      "loss": 0.9912,
+      "step": 7850
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.974483878450476e-06,
+      "loss": 1.5618,
+      "step": 7860
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.916492693110647e-06,
+      "loss": 1.0647,
+      "step": 7870
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.858501507770819e-06,
+      "loss": 1.038,
+      "step": 7880
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.80051032243099e-06,
+      "loss": 0.8973,
+      "step": 7890
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.742519137091162e-06,
+      "loss": 0.2889,
+      "step": 7900
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.684527951751335e-06,
+      "loss": 0.7013,
+      "step": 7910
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.626536766411506e-06,
+      "loss": 1.1109,
+      "step": 7920
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.568545581071678e-06,
+      "loss": 1.1751,
+      "step": 7930
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.510554395731849e-06,
+      "loss": 0.4777,
+      "step": 7940
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.45256321039202e-06,
+      "loss": 1.1054,
+      "step": 7950
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.394572025052192e-06,
+      "loss": 0.8301,
+      "step": 7960
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.336580839712363e-06,
+      "loss": 1.0416,
+      "step": 7970
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.278589654372535e-06,
+      "loss": 1.1081,
+      "step": 7980
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.220598469032706e-06,
+      "loss": 0.1521,
+      "step": 7990
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.16260728369288e-06,
+      "loss": 0.4169,
+      "step": 8000
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.10461609835305e-06,
+      "loss": 0.3955,
+      "step": 8010
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.046624913013222e-06,
+      "loss": 0.5808,
+      "step": 8020
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.988633727673394e-06,
+      "loss": 0.4718,
+      "step": 8030
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.930642542333565e-06,
+      "loss": 0.2894,
+      "step": 8040
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.872651356993736e-06,
+      "loss": 0.7434,
+      "step": 8050
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.81466017165391e-06,
+      "loss": 0.6093,
+      "step": 8060
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.756668986314081e-06,
+      "loss": 0.9806,
+      "step": 8070
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.698677800974252e-06,
+      "loss": 0.7348,
+      "step": 8080
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.640686615634424e-06,
+      "loss": 0.5672,
+      "step": 8090
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.582695430294597e-06,
+      "loss": 0.8777,
+      "step": 8100
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.524704244954768e-06,
+      "loss": 0.6509,
+      "step": 8110
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.46671305961494e-06,
+      "loss": 0.3298,
+      "step": 8120
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.408721874275111e-06,
+      "loss": 1.5041,
+      "step": 8130
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.350730688935283e-06,
+      "loss": 0.4883,
+      "step": 8140
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.292739503595454e-06,
+      "loss": 0.5874,
+      "step": 8150
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.234748318255625e-06,
+      "loss": 0.4102,
+      "step": 8160
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.176757132915797e-06,
+      "loss": 1.0393,
+      "step": 8170
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.11876594757597e-06,
+      "loss": 0.2589,
+      "step": 8180
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.060774762236141e-06,
+      "loss": 0.9598,
+      "step": 8190
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.002783576896313e-06,
+      "loss": 0.8448,
+      "step": 8200
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.944792391556484e-06,
+      "loss": 0.5414,
+      "step": 8210
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.886801206216656e-06,
+      "loss": 1.0305,
+      "step": 8220
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.828810020876827e-06,
+      "loss": 1.217,
+      "step": 8230
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.770818835536999e-06,
+      "loss": 0.8523,
+      "step": 8240
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.71282765019717e-06,
+      "loss": 0.4902,
+      "step": 8250
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.654836464857341e-06,
+      "loss": 0.8131,
+      "step": 8260
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.5968452795175146e-06,
+      "loss": 0.1387,
+      "step": 8270
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.538854094177686e-06,
+      "loss": 1.4893,
+      "step": 8280
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.480862908837857e-06,
+      "loss": 1.372,
+      "step": 8290
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.422871723498029e-06,
+      "loss": 1.2179,
+      "step": 8300
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.3648805381582e-06,
+      "loss": 0.6482,
+      "step": 8310
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.306889352818372e-06,
+      "loss": 0.7181,
+      "step": 8320
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.248898167478543e-06,
+      "loss": 0.8082,
+      "step": 8330
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.1909069821387145e-06,
+      "loss": 0.816,
+      "step": 8340
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.132915796798886e-06,
+      "loss": 1.1753,
+      "step": 8350
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.074924611459059e-06,
+      "loss": 0.7997,
+      "step": 8360
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.0169334261192305e-06,
+      "loss": 0.9955,
+      "step": 8370
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 6.958942240779402e-06,
+      "loss": 0.3213,
+      "step": 8380
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.900951055439573e-06,
+      "loss": 1.0039,
+      "step": 8390
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.842959870099746e-06,
+      "loss": 0.3616,
+      "step": 8400
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.784968684759917e-06,
+      "loss": 0.5561,
+      "step": 8410
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.7269774994200885e-06,
+      "loss": 2.0041,
+      "step": 8420
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.66898631408026e-06,
+      "loss": 1.2475,
+      "step": 8430
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.610995128740431e-06,
+      "loss": 0.0255,
+      "step": 8440
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.5530039434006044e-06,
+      "loss": 1.1149,
+      "step": 8450
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.495012758060776e-06,
+      "loss": 0.9821,
+      "step": 8460
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.437021572720947e-06,
+      "loss": 1.2371,
+      "step": 8470
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.379030387381119e-06,
+      "loss": 1.5603,
+      "step": 8480
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.32103920204129e-06,
+      "loss": 0.786,
+      "step": 8490
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.2630480167014616e-06,
+      "loss": 0.2398,
+      "step": 8500
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.205056831361633e-06,
+      "loss": 0.7071,
+      "step": 8510
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.147065646021805e-06,
+      "loss": 0.6936,
+      "step": 8520
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.089074460681977e-06,
+      "loss": 1.1953,
+      "step": 8530
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.031083275342148e-06,
+      "loss": 1.5947,
+      "step": 8540
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 5.9730920900023195e-06,
+      "loss": 0.6217,
+      "step": 8550
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 5.915100904662491e-06,
+      "loss": 1.1441,
+      "step": 8560
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 5.857109719322663e-06,
+      "loss": 1.3417,
+      "step": 8570
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.799118533982835e-06,
+      "loss": 1.2174,
+      "step": 8580
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.741127348643006e-06,
+      "loss": 1.1166,
+      "step": 8590
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.683136163303178e-06,
+      "loss": 0.3256,
+      "step": 8600
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.62514497796335e-06,
+      "loss": 0.9201,
+      "step": 8610
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.567153792623522e-06,
+      "loss": 0.1479,
+      "step": 8620
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.5091626072836935e-06,
+      "loss": 0.3946,
+      "step": 8630
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.451171421943865e-06,
+      "loss": 0.6609,
+      "step": 8640
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.393180236604036e-06,
+      "loss": 0.4718,
+      "step": 8650
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.335189051264209e-06,
+      "loss": 1.0772,
+      "step": 8660
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.27719786592438e-06,
+      "loss": 0.53,
+      "step": 8670
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.2192066805845514e-06,
+      "loss": 0.8714,
+      "step": 8680
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.161215495244723e-06,
+      "loss": 1.7171,
+      "step": 8690
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.103224309904895e-06,
+      "loss": 0.3548,
+      "step": 8700
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.0452331245650666e-06,
+      "loss": 0.4943,
+      "step": 8710
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 4.987241939225238e-06,
+      "loss": 0.6686,
+      "step": 8720
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 4.929250753885409e-06,
+      "loss": 0.9975,
+      "step": 8730
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 4.871259568545581e-06,
+      "loss": 0.9102,
+      "step": 8740
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 4.813268383205753e-06,
+      "loss": 1.4125,
+      "step": 8750
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 4.7552771978659245e-06,
+      "loss": 1.0263,
+      "step": 8760
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.697286012526096e-06,
+      "loss": 0.5355,
+      "step": 8770
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.639294827186267e-06,
+      "loss": 0.3894,
+      "step": 8780
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.58130364184644e-06,
+      "loss": 1.3545,
+      "step": 8790
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.523312456506611e-06,
+      "loss": 0.7195,
+      "step": 8800
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.4653212711667825e-06,
+      "loss": 0.4793,
+      "step": 8810
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.407330085826955e-06,
+      "loss": 0.807,
+      "step": 8820
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.349338900487126e-06,
+      "loss": 1.0489,
+      "step": 8830
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.2913477151472985e-06,
+      "loss": 0.3796,
+      "step": 8840
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.23335652980747e-06,
+      "loss": 0.6786,
+      "step": 8850
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.175365344467641e-06,
+      "loss": 0.7063,
+      "step": 8860
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 4.117374159127813e-06,
+      "loss": 1.6069,
+      "step": 8870
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 4.059382973787985e-06,
+      "loss": 0.3546,
+      "step": 8880
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 4.0013917884481564e-06,
+      "loss": 0.4889,
+      "step": 8890
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.943400603108328e-06,
+      "loss": 0.2019,
+      "step": 8900
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.885409417768499e-06,
+      "loss": 0.8494,
+      "step": 8910
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.827418232428671e-06,
+      "loss": 0.2225,
+      "step": 8920
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.769427047088843e-06,
+      "loss": 0.2444,
+      "step": 8930
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.7114358617490144e-06,
+      "loss": 1.3131,
+      "step": 8940
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.653444676409186e-06,
+      "loss": 0.7759,
+      "step": 8950
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.5954534910693573e-06,
+      "loss": 0.5356,
+      "step": 8960
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.5374623057295295e-06,
+      "loss": 0.9673,
+      "step": 8970
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.479471120389701e-06,
+      "loss": 1.0627,
+      "step": 8980
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.421479935049873e-06,
+      "loss": 1.2619,
+      "step": 8990
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.3634887497100442e-06,
+      "loss": 0.2523,
+      "step": 9000
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.3054975643702157e-06,
+      "loss": 0.6017,
+      "step": 9010
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.247506379030388e-06,
+      "loss": 0.4459,
+      "step": 9020
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.1895151936905594e-06,
+      "loss": 0.3974,
+      "step": 9030
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.1315240083507308e-06,
+      "loss": 0.5499,
+      "step": 9040
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.0735328230109026e-06,
+      "loss": 2.1391,
+      "step": 9050
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 3.015541637671074e-06,
+      "loss": 0.9935,
+      "step": 9060
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.9575504523312455e-06,
+      "loss": 0.7154,
+      "step": 9070
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.8995592669914173e-06,
+      "loss": 1.0268,
+      "step": 9080
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.841568081651589e-06,
+      "loss": 0.6102,
+      "step": 9090
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.783576896311761e-06,
+      "loss": 0.5946,
+      "step": 9100
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.7255857109719324e-06,
+      "loss": 1.0439,
+      "step": 9110
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.6675945256321043e-06,
+      "loss": 0.4686,
+      "step": 9120
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.6096033402922757e-06,
+      "loss": 1.2022,
+      "step": 9130
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.5516121549524476e-06,
+      "loss": 0.235,
+      "step": 9140
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.493620969612619e-06,
+      "loss": 1.1548,
+      "step": 9150
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.4356297842727904e-06,
+      "loss": 0.6478,
+      "step": 9160
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.3776385989329623e-06,
+      "loss": 0.7001,
+      "step": 9170
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.3196474135931337e-06,
+      "loss": 0.1863,
+      "step": 9180
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.2616562282533055e-06,
+      "loss": 1.231,
+      "step": 9190
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.2036650429134774e-06,
+      "loss": 0.5608,
+      "step": 9200
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.1456738575736492e-06,
+      "loss": 1.201,
+      "step": 9210
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.0876826722338207e-06,
+      "loss": 0.2587,
+      "step": 9220
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.0296914868939925e-06,
+      "loss": 0.4604,
+      "step": 9230
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.971700301554164e-06,
+      "loss": 1.0893,
+      "step": 9240
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.9137091162143354e-06,
+      "loss": 1.1889,
+      "step": 9250
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.8557179308745072e-06,
+      "loss": 1.1303,
+      "step": 9260
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.7977267455346786e-06,
+      "loss": 0.8834,
+      "step": 9270
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.7397355601948505e-06,
+      "loss": 0.8237,
+      "step": 9280
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.6817443748550221e-06,
+      "loss": 1.0304,
+      "step": 9290
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.623753189515194e-06,
+      "loss": 0.4812,
+      "step": 9300
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.5657620041753654e-06,
+      "loss": 0.2212,
+      "step": 9310
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.507770818835537e-06,
+      "loss": 1.0802,
+      "step": 9320
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.4497796334957087e-06,
+      "loss": 0.5977,
+      "step": 9330
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.3917884481558805e-06,
+      "loss": 0.287,
+      "step": 9340
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.3337972628160521e-06,
+      "loss": 0.7585,
+      "step": 9350
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.2758060774762238e-06,
+      "loss": 1.169,
+      "step": 9360
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.2178148921363952e-06,
+      "loss": 1.048,
+      "step": 9370
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.1598237067965668e-06,
+      "loss": 0.0311,
+      "step": 9380
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.1018325214567387e-06,
+      "loss": 0.7598,
+      "step": 9390
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.0438413361169103e-06,
+      "loss": 0.4497,
+      "step": 9400
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 9.85850150777082e-07,
+      "loss": 0.7617,
+      "step": 9410
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 9.278589654372536e-07,
+      "loss": 0.7371,
+      "step": 9420
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.698677800974252e-07,
+      "loss": 0.8452,
+      "step": 9430
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.11876594757597e-07,
+      "loss": 0.7718,
+      "step": 9440
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 7.538854094177685e-07,
+      "loss": 1.1021,
+      "step": 9450
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 6.958942240779403e-07,
+      "loss": 1.4094,
+      "step": 9460
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 6.379030387381119e-07,
+      "loss": 0.5406,
+      "step": 9470
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 5.799118533982834e-07,
+      "loss": 1.8451,
+      "step": 9480
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 5.219206680584552e-07,
+      "loss": 0.783,
+      "step": 9490
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.639294827186268e-07,
+      "loss": 0.7202,
+      "step": 9500
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.059382973787985e-07,
+      "loss": 0.9798,
+      "step": 9510
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.4794711203897013e-07,
+      "loss": 0.2955,
+      "step": 9520
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.899559266991417e-07,
+      "loss": 0.4296,
+      "step": 9530
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 2.319647413593134e-07,
+      "loss": 1.2112,
+      "step": 9540
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.7397355601948506e-07,
+      "loss": 1.4026,
+      "step": 9550
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.159823706796567e-07,
+      "loss": 0.9493,
+      "step": 9560
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 5.799118533982835e-08,
+      "loss": 0.9242,
+      "step": 9570
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 0.0,
+      "loss": 0.723,
+      "step": 9580
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 0.730299667036626,
+      "eval_loss": 1.0293892621994019,
+      "eval_runtime": 561.1246,
+      "eval_samples_per_second": 1.606,
+      "eval_steps_per_second": 0.804,
+      "step": 9580
+    },
+    {
+      "epoch": 4.2,
+      "step": 9580,
+      "total_flos": 2.387715253845885e+19,
+      "train_loss": 0.8973092991679844,
+      "train_runtime": 68017.2603,
+      "train_samples_per_second": 0.282,
+      "train_steps_per_second": 0.141
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 0.6987104337631888,
+      "eval_loss": 1.181971549987793,
+      "eval_runtime": 581.8763,
+      "eval_samples_per_second": 1.466,
+      "eval_steps_per_second": 0.734,
+      "step": 9580
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 0.6987104337631888,
+      "eval_loss": 1.1819714307785034,
+      "eval_runtime": 573.3606,
+      "eval_samples_per_second": 1.488,
+      "eval_steps_per_second": 0.745,
+      "step": 9580
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 9580,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "total_flos": 2.387715253845885e+19,
+  "trial_name": null,
+  "trial_params": null
+}