diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7627 @@
+{
+  "best_metric": 0.8251666013328106,
+  "best_model_checkpoint": "videomae-large-cctv-brawl_extended_v1/checkpoint-12565",
+  "epoch": 4.199681655391962,
+  "eval_steps": 500,
+  "global_step": 12565,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.386634844868735e-09,
+      "loss": 0.7489,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.77326968973747e-09,
+      "loss": 0.724,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.159904534606205e-09,
+      "loss": 0.7649,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.54653937947494e-09,
+      "loss": 0.6728,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1933174224343675e-08,
+      "loss": 0.7368,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.431980906921241e-08,
+      "loss": 0.7369,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6706443914081144e-08,
+      "loss": 0.7463,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.909307875894988e-08,
+      "loss": 0.6903,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1479713603818616e-08,
+      "loss": 0.725,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.386634844868735e-08,
+      "loss": 0.7142,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6252983293556084e-08,
+      "loss": 0.7131,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.863961813842482e-08,
+      "loss": 0.7178,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.102625298329355e-08,
+      "loss": 0.6786,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341288782816229e-08,
+      "loss": 0.7279,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579952267303102e-08,
+      "loss": 0.6817,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818615751789976e-08,
+      "loss": 0.6895,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05727923627685e-08,
+      "loss": 0.6714,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.295942720763723e-08,
+      "loss": 0.7969,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5346062052505965e-08,
+      "loss": 0.6794,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77326968973747e-08,
+      "loss": 0.7054,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.0119331742243434e-08,
+      "loss": 0.7599,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.250596658711217e-08,
+      "loss": 0.7148,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.48926014319809e-08,
+      "loss": 0.7105,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.727923627684964e-08,
+      "loss": 0.7208,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.966587112171838e-08,
+      "loss": 0.6222,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.20525059665871e-08,
+      "loss": 0.6742,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.443914081145585e-08,
+      "loss": 0.661,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.682577565632457e-08,
+      "loss": 0.722,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.921241050119332e-08,
+      "loss": 0.7342,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.159904534606204e-08,
+      "loss": 0.6948,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.398568019093078e-08,
+      "loss": 0.7301,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.637231503579952e-08,
+      "loss": 0.6596,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.875894988066825e-08,
+      "loss": 0.6857,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.1145584725537e-08,
+      "loss": 0.7129,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.353221957040572e-08,
+      "loss": 0.6889,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.591885441527446e-08,
+      "loss": 0.6673,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.830548926014319e-08,
+      "loss": 0.6823,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.069212410501193e-08,
+      "loss": 0.6655,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.307875894988066e-08,
+      "loss": 0.6166,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.54653937947494e-08,
+      "loss": 0.6371,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.785202863961813e-08,
+      "loss": 0.704,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0023866348448687e-07,
+      "loss": 0.7303,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0262529832935561e-07,
+      "loss": 0.709,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.0501193317422434e-07,
+      "loss": 0.6405,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.0739856801909308e-07,
+      "loss": 0.6122,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.097852028639618e-07,
+      "loss": 0.6458,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1217183770883055e-07,
+      "loss": 0.6574,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1455847255369927e-07,
+      "loss": 0.6348,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1694510739856802e-07,
+      "loss": 0.6998,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1933174224343676e-07,
+      "loss": 0.6245,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.2171837708830548e-07,
+      "loss": 0.6499,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.241050119331742e-07,
+      "loss": 0.5856,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.2649164677804294e-07,
+      "loss": 0.7143,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.288782816229117e-07,
+      "loss": 0.6285,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.3126491646778042e-07,
+      "loss": 0.6485,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.3365155131264915e-07,
+      "loss": 0.6322,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.3603818615751788e-07,
+      "loss": 0.6228,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.3842482100238663e-07,
+      "loss": 0.6469,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.4081145584725536e-07,
+      "loss": 0.6374,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.4319809069212409e-07,
+      "loss": 0.6418,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.4558472553699284e-07,
+      "loss": 0.6212,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.4797136038186157e-07,
+      "loss": 0.66,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.503579952267303e-07,
+      "loss": 0.5956,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5274463007159905e-07,
+      "loss": 0.6426,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5513126491646775e-07,
+      "loss": 0.6602,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.575178997613365e-07,
+      "loss": 0.633,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5990453460620523e-07,
+      "loss": 0.6016,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.62291169451074e-07,
+      "loss": 0.6235,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.646778042959427e-07,
+      "loss": 0.579,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.6706443914081144e-07,
+      "loss": 0.5734,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.6945107398568017e-07,
+      "loss": 0.5904,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.7183770883054892e-07,
+      "loss": 0.5944,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.7422434367541765e-07,
+      "loss": 0.6764,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.7661097852028638e-07,
+      "loss": 0.5815,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.789976133651551e-07,
+      "loss": 0.5883,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.8138424821002386e-07,
+      "loss": 0.6033,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.837708830548926e-07,
+      "loss": 0.5302,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.8615751789976132e-07,
+      "loss": 0.5989,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.8854415274463004e-07,
+      "loss": 0.5534,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.909307875894988e-07,
+      "loss": 0.6403,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9331742243436753e-07,
+      "loss": 0.4963,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9570405727923625e-07,
+      "loss": 0.653,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.98090692124105e-07,
+      "loss": 0.606,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.0047732696897374e-07,
+      "loss": 0.6209,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.028639618138425e-07,
+      "loss": 0.5385,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.0525059665871122e-07,
+      "loss": 0.557,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.0763723150357995e-07,
+      "loss": 0.6703,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.1002386634844867e-07,
+      "loss": 0.5217,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.1241050119331743e-07,
+      "loss": 0.5697,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.1479713603818616e-07,
+      "loss": 0.5365,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.1718377088305488e-07,
+      "loss": 0.56,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.195704057279236e-07,
+      "loss": 0.4775,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.2195704057279237e-07,
+      "loss": 0.5328,
+      "step": 930
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.243436754176611e-07,
+      "loss": 0.6135,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.2673031026252982e-07,
+      "loss": 0.5732,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.2911694510739855e-07,
+      "loss": 0.5636,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.315035799522673e-07,
+      "loss": 0.6081,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.3389021479713603e-07,
+      "loss": 0.619,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.3627684964200478e-07,
+      "loss": 0.5916,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.386634844868735e-07,
+      "loss": 0.6113,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.4105011933174227e-07,
+      "loss": 0.6695,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.4343675417661097e-07,
+      "loss": 0.559,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.458233890214797e-07,
+      "loss": 0.5641,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.482100238663484e-07,
+      "loss": 0.5845,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.505966587112172e-07,
+      "loss": 0.6465,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.529832935560859e-07,
+      "loss": 0.5213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5536992840095463e-07,
+      "loss": 0.5663,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.577565632458234e-07,
+      "loss": 0.6348,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.6014319809069214e-07,
+      "loss": 0.5389,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.6252983293556084e-07,
+      "loss": 0.6299,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.649164677804296e-07,
+      "loss": 0.5185,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.673031026252983e-07,
+      "loss": 0.4967,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.6968973747016705e-07,
+      "loss": 0.5323,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.7207637231503575e-07,
+      "loss": 0.4797,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.744630071599045e-07,
+      "loss": 0.591,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.7684964200477326e-07,
+      "loss": 0.488,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.79236276849642e-07,
+      "loss": 0.4519,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.816229116945107e-07,
+      "loss": 0.5839,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.8400954653937947e-07,
+      "loss": 0.5682,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.8639618138424817e-07,
+      "loss": 0.4765,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.8878281622911693e-07,
+      "loss": 0.6187,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.911694510739857e-07,
+      "loss": 0.4894,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.935560859188544e-07,
+      "loss": 0.5424,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9594272076372314e-07,
+      "loss": 0.5351,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.983293556085919e-07,
+      "loss": 0.5059,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.999204103289706e-07,
+      "loss": 0.5178,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.996551114255394e-07,
+      "loss": 0.3589,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9938981252210824e-07,
+      "loss": 0.5448,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9912451361867705e-07,
+      "loss": 0.4891,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.988592147152458e-07,
+      "loss": 0.5095,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9859391581181463e-07,
+      "loss": 0.5336,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9832861690838345e-07,
+      "loss": 0.5037,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.980633180049522e-07,
+      "loss": 0.4463,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.97798019101521e-07,
+      "loss": 0.4545,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9753272019808984e-07,
+      "loss": 0.5409,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9726742129465865e-07,
+      "loss": 0.408,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9700212239122747e-07,
+      "loss": 0.5175,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9673682348779623e-07,
+      "loss": 0.4775,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9647152458436504e-07,
+      "loss": 0.6129,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9620622568093386e-07,
+      "loss": 0.366,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.959409267775026e-07,
+      "loss": 0.4697,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9567562787407143e-07,
+      "loss": 0.4706,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9541032897064025e-07,
+      "loss": 0.4555,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9514503006720906e-07,
+      "loss": 0.5475,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.948797311637778e-07,
+      "loss": 0.4358,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9461443226034664e-07,
+      "loss": 0.5115,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9434913335691545e-07,
+      "loss": 0.6776,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.940838344534842e-07,
+      "loss": 0.4034,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.938185355500531e-07,
+      "loss": 0.5232,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9355323664662185e-07,
+      "loss": 0.5778,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9328793774319066e-07,
+      "loss": 0.5217,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.930226388397595e-07,
+      "loss": 0.5378,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9275733993632824e-07,
+      "loss": 0.5131,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9249204103289705e-07,
+      "loss": 0.4615,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9222674212946587e-07,
+      "loss": 0.5084,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9196144322603463e-07,
+      "loss": 0.5639,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9169614432260344e-07,
+      "loss": 0.6176,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9143084541917226e-07,
+      "loss": 0.4819,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9116554651574107e-07,
+      "loss": 0.6385,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9090024761230983e-07,
+      "loss": 0.5515,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9063494870887865e-07,
+      "loss": 0.5406,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9036964980544746e-07,
+      "loss": 0.5813,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.901043509020162e-07,
+      "loss": 0.6192,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.898390519985851e-07,
+      "loss": 0.5759,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.8957375309515386e-07,
+      "loss": 0.4963,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.8930845419172267e-07,
+      "loss": 0.5021,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.890431552882915e-07,
+      "loss": 0.5574,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.8877785638486025e-07,
+      "loss": 0.4434,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.8851255748142906e-07,
+      "loss": 0.5213,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.882472585779979e-07,
+      "loss": 0.461,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.879819596745667e-07,
+      "loss": 0.4587,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.8771666077113545e-07,
+      "loss": 0.3879,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.8745136186770427e-07,
+      "loss": 0.48,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.871860629642731e-07,
+      "loss": 0.6378,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.8692076406084184e-07,
+      "loss": 0.5548,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.8665546515741066e-07,
+      "loss": 0.4741,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.8639016625397947e-07,
+      "loss": 0.4441,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.8612486735054823e-07,
+      "loss": 0.4345,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.858595684471171e-07,
+      "loss": 0.6156,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.8559426954368586e-07,
+      "loss": 0.4104,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.853289706402547e-07,
+      "loss": 0.5361,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.850636717368235e-07,
+      "loss": 0.4971,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.8479837283339226e-07,
+      "loss": 0.4066,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.8453307392996107e-07,
+      "loss": 0.4786,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.842677750265299e-07,
+      "loss": 0.4475,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.840024761230987e-07,
+      "loss": 0.596,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.8373717721966746e-07,
+      "loss": 0.4797,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.834718783162363e-07,
+      "loss": 0.6381,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.832065794128051e-07,
+      "loss": 0.5512,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.8294128050937385e-07,
+      "loss": 0.526,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.8267598160594267e-07,
+      "loss": 0.4208,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.824106827025115e-07,
+      "loss": 0.5019,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.821453837990803e-07,
+      "loss": 0.5601,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.818800848956491e-07,
+      "loss": 0.4532,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.816147859922179e-07,
+      "loss": 0.6079,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.813494870887867e-07,
+      "loss": 0.4444,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.810841881853555e-07,
+      "loss": 0.5132,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.8081888928192426e-07,
+      "loss": 0.5627,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.805535903784931e-07,
+      "loss": 0.4318,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.802882914750619e-07,
+      "loss": 0.5026,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.800229925716307e-07,
+      "loss": 0.5187,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.7975769366819947e-07,
+      "loss": 0.396,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.794923947647683e-07,
+      "loss": 0.4171,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.792270958613371e-07,
+      "loss": 0.4792,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.7896179695790586e-07,
+      "loss": 0.4526,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.7869649805447473e-07,
+      "loss": 0.4952,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.784311991510435e-07,
+      "loss": 0.5462,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.781659002476123e-07,
+      "loss": 0.6351,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.779006013441811e-07,
+      "loss": 0.5329,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.776353024407499e-07,
+      "loss": 0.3761,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.773700035373187e-07,
+      "loss": 0.4625,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.771047046338875e-07,
+      "loss": 0.5105,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.768394057304563e-07,
+      "loss": 0.4063,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.765741068270251e-07,
+      "loss": 0.4644,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.763088079235939e-07,
+      "loss": 0.4825,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.760435090201627e-07,
+      "loss": 0.5125,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.757782101167315e-07,
+      "loss": 0.5761,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.755129112133003e-07,
+      "loss": 0.4859,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.752476123098691e-07,
+      "loss": 0.4388,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.7498231340643787e-07,
+      "loss": 0.5507,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.7471701450300674e-07,
+      "loss": 0.4343,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.744517155995755e-07,
+      "loss": 0.4135,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.741864166961443e-07,
+      "loss": 0.4108,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.7392111779271313e-07,
+      "loss": 0.4274,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.736558188892819e-07,
+      "loss": 0.4261,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.733905199858507e-07,
+      "loss": 0.4734,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.731252210824195e-07,
+      "loss": 0.5553,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.7285992217898834e-07,
+      "loss": 0.4643,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.725946232755571e-07,
+      "loss": 0.458,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.723293243721259e-07,
+      "loss": 0.5908,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.7206402546869473e-07,
+      "loss": 0.533,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.717987265652635e-07,
+      "loss": 0.6152,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.715334276618323e-07,
+      "loss": 0.3958,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.712681287584011e-07,
+      "loss": 0.4243,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.710028298549699e-07,
+      "loss": 0.4743,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.7073753095153875e-07,
+      "loss": 0.603,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.704722320481075e-07,
+      "loss": 0.4256,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.702069331446763e-07,
+      "loss": 0.5295,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.6994163424124514e-07,
+      "loss": 0.4739,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.696763353378139e-07,
+      "loss": 0.3976,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.694110364343827e-07,
+      "loss": 0.5386,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.6914573753095153e-07,
+      "loss": 0.5893,
+      "step": 2420
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.6888043862752035e-07,
+      "loss": 0.4234,
+      "step": 2430
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.686151397240891e-07,
+      "loss": 0.411,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.683498408206579e-07,
+      "loss": 0.5547,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.6808454191722674e-07,
+      "loss": 0.4212,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.678192430137955e-07,
+      "loss": 0.3612,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.675539441103643e-07,
+      "loss": 0.7204,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.6728864520693313e-07,
+      "loss": 0.5413,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.6702334630350194e-07,
+      "loss": 0.4005,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.6675804740007076e-07,
+      "loss": 0.4193,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2,
+      "eval_accuracy": 0.6985495883967072,
+      "eval_loss": 0.48386532068252563,
+      "eval_runtime": 707.7642,
+      "eval_samples_per_second": 3.604,
+      "eval_steps_per_second": 0.901,
+      "step": 2514
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.664927484966395e-07,
+      "loss": 0.4189,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6622744959320833e-07,
+      "loss": 0.5329,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6596215068977715e-07,
+      "loss": 0.528,
+      "step": 2540
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.656968517863459e-07,
+      "loss": 0.4498,
+      "step": 2550
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.654315528829147e-07,
+      "loss": 0.4362,
+      "step": 2560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6516625397948354e-07,
+      "loss": 0.4513,
+      "step": 2570
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6490095507605236e-07,
+      "loss": 0.45,
+      "step": 2580
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6463565617262117e-07,
+      "loss": 0.5109,
+      "step": 2590
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6437035726918993e-07,
+      "loss": 0.454,
+      "step": 2600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6410505836575875e-07,
+      "loss": 0.5103,
+      "step": 2610
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6383975946232756e-07,
+      "loss": 0.521,
+      "step": 2620
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.635744605588964e-07,
+      "loss": 0.5211,
+      "step": 2630
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6330916165546514e-07,
+      "loss": 0.485,
+      "step": 2640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6304386275203395e-07,
+      "loss": 0.3995,
+      "step": 2650
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6277856384860277e-07,
+      "loss": 0.6925,
+      "step": 2660
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6251326494517153e-07,
+      "loss": 0.3811,
+      "step": 2670
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6224796604174034e-07,
+      "loss": 0.4942,
+      "step": 2680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6198266713830916e-07,
+      "loss": 0.4295,
+      "step": 2690
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.617173682348779e-07,
+      "loss": 0.6002,
+      "step": 2700
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.614520693314468e-07,
+      "loss": 0.3757,
+      "step": 2710
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6118677042801555e-07,
+      "loss": 0.502,
+      "step": 2720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6092147152458436e-07,
+      "loss": 0.2979,
+      "step": 2730
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.606561726211532e-07,
+      "loss": 0.3475,
+      "step": 2740
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6039087371772194e-07,
+      "loss": 0.5197,
+      "step": 2750
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6012557481429076e-07,
+      "loss": 0.379,
+      "step": 2760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5986027591085957e-07,
+      "loss": 0.4161,
+      "step": 2770
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.595949770074284e-07,
+      "loss": 0.561,
+      "step": 2780
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5932967810399715e-07,
+      "loss": 0.4509,
+      "step": 2790
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5906437920056596e-07,
+      "loss": 0.3427,
+      "step": 2800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.587990802971348e-07,
+      "loss": 0.315,
+      "step": 2810
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5853378139370354e-07,
+      "loss": 0.5377,
+      "step": 2820
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5826848249027235e-07,
+      "loss": 0.4747,
+      "step": 2830
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5800318358684117e-07,
+      "loss": 0.604,
+      "step": 2840
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5773788468341e-07,
+      "loss": 0.6197,
+      "step": 2850
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.574725857799788e-07,
+      "loss": 0.6246,
+      "step": 2860
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5720728687654756e-07,
+      "loss": 0.385,
+      "step": 2870
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.569419879731164e-07,
+      "loss": 0.3885,
+      "step": 2880
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.566766890696852e-07,
+      "loss": 0.3586,
+      "step": 2890
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5641139016625395e-07,
+      "loss": 0.4253,
+      "step": 2900
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5614609126282276e-07,
+      "loss": 0.3906,
+      "step": 2910
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.558807923593916e-07,
+      "loss": 0.5617,
+      "step": 2920
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.556154934559604e-07,
+      "loss": 0.5688,
+      "step": 2930
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5535019455252916e-07,
+      "loss": 0.5803,
+      "step": 2940
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5508489564909797e-07,
+      "loss": 0.4535,
+      "step": 2950
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.548195967456668e-07,
+      "loss": 0.4301,
+      "step": 2960
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.5455429784223555e-07,
+      "loss": 0.4844,
+      "step": 2970
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.542889989388044e-07,
+      "loss": 0.4359,
+      "step": 2980
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.540237000353732e-07,
+      "loss": 0.4562,
+      "step": 2990
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.53758401131942e-07,
+      "loss": 0.3864,
+      "step": 3000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.534931022285108e-07,
+      "loss": 0.5046,
+      "step": 3010
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.5322780332507957e-07,
+      "loss": 0.2887,
+      "step": 3020
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.529625044216484e-07,
+      "loss": 0.4083,
+      "step": 3030
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.526972055182172e-07,
+      "loss": 0.3877,
+      "step": 3040
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.5243190661478596e-07,
+      "loss": 0.5078,
+      "step": 3050
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.521666077113548e-07,
+      "loss": 0.4621,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.519013088079236e-07,
+      "loss": 0.4411,
+      "step": 3070
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.516360099044924e-07,
+      "loss": 0.5897,
+      "step": 3080
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.5137071100106117e-07,
+      "loss": 0.4126,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.5110541209763e-07,
+      "loss": 0.4725,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.508401131941988e-07,
+      "loss": 0.4072,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.5057481429076756e-07,
+      "loss": 0.5103,
+      "step": 3120
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.503095153873364e-07,
+      "loss": 0.4725,
+      "step": 3130
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.500442164839052e-07,
+      "loss": 0.4113,
+      "step": 3140
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.49778917580474e-07,
+      "loss": 0.3694,
+      "step": 3150
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.495136186770428e-07,
+      "loss": 0.4339,
+      "step": 3160
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.492483197736116e-07,
+      "loss": 0.4924,
+      "step": 3170
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.489830208701804e-07,
+      "loss": 0.4624,
+      "step": 3180
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.487177219667492e-07,
+      "loss": 0.406,
+      "step": 3190
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.4845242306331797e-07,
+      "loss": 0.4311,
+      "step": 3200
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.481871241598868e-07,
+      "loss": 0.3505,
+      "step": 3210
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.479218252564556e-07,
+      "loss": 0.5599,
+      "step": 3220
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.476565263530244e-07,
+      "loss": 0.4658,
+      "step": 3230
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.473912274495932e-07,
+      "loss": 0.3503,
+      "step": 3240
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.47125928546162e-07,
+      "loss": 0.4346,
+      "step": 3250
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.468606296427308e-07,
+      "loss": 0.4888,
+      "step": 3260
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.4659533073929957e-07,
+      "loss": 0.4938,
+      "step": 3270
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.4633003183586843e-07,
+      "loss": 0.3205,
+      "step": 3280
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.460647329324372e-07,
+      "loss": 0.3941,
+      "step": 3290
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.45799434029006e-07,
+      "loss": 0.447,
+      "step": 3300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.455341351255748e-07,
+      "loss": 0.2545,
+      "step": 3310
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.452688362221436e-07,
+      "loss": 0.4893,
+      "step": 3320
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.450035373187124e-07,
+      "loss": 0.4399,
+      "step": 3330
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.447382384152812e-07,
+      "loss": 0.4247,
+      "step": 3340
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.4447293951185003e-07,
+      "loss": 0.5176,
+      "step": 3350
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.442076406084188e-07,
+      "loss": 0.5129,
+      "step": 3360
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.439423417049876e-07,
+      "loss": 0.3134,
+      "step": 3370
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.436770428015564e-07,
+      "loss": 0.4207,
+      "step": 3380
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.434117438981252e-07,
+      "loss": 0.6662,
+      "step": 3390
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.43146444994694e-07,
+      "loss": 0.4619,
+      "step": 3400
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.428811460912628e-07,
+      "loss": 0.5505,
+      "step": 3410
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.426158471878316e-07,
+      "loss": 0.5201,
+      "step": 3420
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.4235054828440044e-07,
+      "loss": 0.3945,
+      "step": 3430
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.420852493809692e-07,
+      "loss": 0.4216,
+      "step": 3440
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.41819950477538e-07,
+      "loss": 0.4044,
+      "step": 3450
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.4155465157410683e-07,
+      "loss": 0.4147,
+      "step": 3460
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.412893526706756e-07,
+      "loss": 0.4367,
+      "step": 3470
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.410240537672444e-07,
+      "loss": 0.4442,
+      "step": 3480
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.407587548638132e-07,
+      "loss": 0.3516,
+      "step": 3490
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.4049345596038204e-07,
+      "loss": 0.4012,
+      "step": 3500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.402281570569508e-07,
+      "loss": 0.3966,
+      "step": 3510
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.399628581535196e-07,
+      "loss": 0.4944,
+      "step": 3520
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3969755925008843e-07,
+      "loss": 0.5081,
+      "step": 3530
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.394322603466572e-07,
+      "loss": 0.5445,
+      "step": 3540
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.39166961443226e-07,
+      "loss": 0.3361,
+      "step": 3550
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.389016625397948e-07,
+      "loss": 0.3811,
+      "step": 3560
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3863636363636364e-07,
+      "loss": 0.4163,
+      "step": 3570
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3837106473293243e-07,
+      "loss": 0.3715,
+      "step": 3580
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3810576582950121e-07,
+      "loss": 0.4421,
+      "step": 3590
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3784046692607003e-07,
+      "loss": 0.5121,
+      "step": 3600
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3757516802263884e-07,
+      "loss": 0.4881,
+      "step": 3610
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3730986911920763e-07,
+      "loss": 0.4282,
+      "step": 3620
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3704457021577642e-07,
+      "loss": 0.3844,
+      "step": 3630
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3677927131234523e-07,
+      "loss": 0.3984,
+      "step": 3640
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3651397240891402e-07,
+      "loss": 0.3395,
+      "step": 3650
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.362486735054828e-07,
+      "loss": 0.3538,
+      "step": 3660
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3598337460205165e-07,
+      "loss": 0.469,
+      "step": 3670
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3571807569862044e-07,
+      "loss": 0.4087,
+      "step": 3680
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3545277679518923e-07,
+      "loss": 0.5183,
+      "step": 3690
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3518747789175804e-07,
+      "loss": 0.4856,
+      "step": 3700
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3492217898832683e-07,
+      "loss": 0.3906,
+      "step": 3710
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3465688008489562e-07,
+      "loss": 0.4084,
+      "step": 3720
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3439158118146444e-07,
+      "loss": 0.4071,
+      "step": 3730
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3412628227803325e-07,
+      "loss": 0.4447,
+      "step": 3740
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3386098337460204e-07,
+      "loss": 0.4718,
+      "step": 3750
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3359568447117085e-07,
+      "loss": 0.4005,
+      "step": 3760
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3333038556773964e-07,
+      "loss": 0.4809,
+      "step": 3770
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3306508666430843e-07,
+      "loss": 0.3627,
+      "step": 3780
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3279978776087724e-07,
+      "loss": 0.4105,
+      "step": 3790
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3253448885744603e-07,
+      "loss": 0.4147,
+      "step": 3800
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3226918995401482e-07,
+      "loss": 0.5299,
+      "step": 3810
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3200389105058366e-07,
+      "loss": 0.4322,
+      "step": 3820
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.3173859214715245e-07,
+      "loss": 0.3699,
+      "step": 3830
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.3147329324372124e-07,
+      "loss": 0.3699,
+      "step": 3840
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.3120799434029005e-07,
+      "loss": 0.2841,
+      "step": 3850
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.3094269543685884e-07,
+      "loss": 0.5602,
+      "step": 3860
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.3067739653342763e-07,
+      "loss": 0.3621,
+      "step": 3870
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.3041209762999647e-07,
+      "loss": 0.3471,
+      "step": 3880
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.3014679872656526e-07,
+      "loss": 0.3144,
+      "step": 3890
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2988149982313405e-07,
+      "loss": 0.3478,
+      "step": 3900
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2961620091970286e-07,
+      "loss": 0.3689,
+      "step": 3910
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2935090201627165e-07,
+      "loss": 0.4915,
+      "step": 3920
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2908560311284044e-07,
+      "loss": 0.5927,
+      "step": 3930
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2882030420940925e-07,
+      "loss": 0.3461,
+      "step": 3940
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2855500530597804e-07,
+      "loss": 0.4687,
+      "step": 3950
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2828970640254686e-07,
+      "loss": 0.3873,
+      "step": 3960
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2802440749911567e-07,
+      "loss": 0.4167,
+      "step": 3970
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2775910859568446e-07,
+      "loss": 0.369,
+      "step": 3980
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2749380969225325e-07,
+      "loss": 0.3097,
+      "step": 3990
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2722851078882206e-07,
+      "loss": 0.4325,
+      "step": 4000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2696321188539085e-07,
+      "loss": 0.3702,
+      "step": 4010
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2669791298195964e-07,
+      "loss": 0.5312,
+      "step": 4020
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2643261407852848e-07,
+      "loss": 0.2592,
+      "step": 4030
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2616731517509727e-07,
+      "loss": 0.5263,
+      "step": 4040
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2590201627166606e-07,
+      "loss": 0.3491,
+      "step": 4050
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2563671736823487e-07,
+      "loss": 0.4487,
+      "step": 4060
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2537141846480366e-07,
+      "loss": 0.4003,
+      "step": 4070
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.2510611956137245e-07,
+      "loss": 0.554,
+      "step": 4080
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.248408206579413e-07,
+      "loss": 0.3426,
+      "step": 4090
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2457552175451008e-07,
+      "loss": 0.3495,
+      "step": 4100
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2431022285107887e-07,
+      "loss": 0.3407,
+      "step": 4110
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2404492394764768e-07,
+      "loss": 0.4444,
+      "step": 4120
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2377962504421647e-07,
+      "loss": 0.3747,
+      "step": 4130
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2351432614078526e-07,
+      "loss": 0.4015,
+      "step": 4140
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2324902723735407e-07,
+      "loss": 0.34,
+      "step": 4150
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2298372833392286e-07,
+      "loss": 0.4813,
+      "step": 4160
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2271842943049165e-07,
+      "loss": 0.4426,
+      "step": 4170
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.224531305270605e-07,
+      "loss": 0.314,
+      "step": 4180
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2218783162362928e-07,
+      "loss": 0.4372,
+      "step": 4190
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2192253272019807e-07,
+      "loss": 0.5747,
+      "step": 4200
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.2165723381676688e-07,
+      "loss": 0.3215,
+      "step": 4210
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.2139193491333567e-07,
+      "loss": 0.4074,
+      "step": 4220
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.2112663600990446e-07,
+      "loss": 0.3729,
+      "step": 4230
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.208613371064733e-07,
+      "loss": 0.4418,
+      "step": 4240
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.205960382030421e-07,
+      "loss": 0.4403,
+      "step": 4250
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.203307392996109e-07,
+      "loss": 0.4159,
+      "step": 4260
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.200654403961797e-07,
+      "loss": 0.4298,
+      "step": 4270
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.1980014149274848e-07,
+      "loss": 0.4128,
+      "step": 4280
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.195348425893173e-07,
+      "loss": 0.3948,
+      "step": 4290
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.1926954368588608e-07,
+      "loss": 0.3271,
+      "step": 4300
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.190042447824549e-07,
+      "loss": 0.4198,
+      "step": 4310
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.187389458790237e-07,
+      "loss": 0.4266,
+      "step": 4320
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.184736469755925e-07,
+      "loss": 0.4236,
+      "step": 4330
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.182083480721613e-07,
+      "loss": 0.4734,
+      "step": 4340
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.179430491687301e-07,
+      "loss": 0.4639,
+      "step": 4350
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.176777502652989e-07,
+      "loss": 0.3681,
+      "step": 4360
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.1741245136186768e-07,
+      "loss": 0.2595,
+      "step": 4370
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.1714715245843652e-07,
+      "loss": 0.4727,
+      "step": 4380
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.168818535550053e-07,
+      "loss": 0.3284,
+      "step": 4390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.166165546515741e-07,
+      "loss": 0.3289,
+      "step": 4400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.163512557481429e-07,
+      "loss": 0.4443,
+      "step": 4410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.160859568447117e-07,
+      "loss": 0.2787,
+      "step": 4420
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.158206579412805e-07,
+      "loss": 0.6671,
+      "step": 4430
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.1555535903784933e-07,
+      "loss": 0.3741,
+      "step": 4440
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.1529006013441812e-07,
+      "loss": 0.3775,
+      "step": 4450
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.150247612309869e-07,
+      "loss": 0.3939,
+      "step": 4460
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.1475946232755572e-07,
+      "loss": 0.2898,
+      "step": 4470
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.144941634241245e-07,
+      "loss": 0.535,
+      "step": 4480
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.142288645206933e-07,
+      "loss": 0.3356,
+      "step": 4490
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.139635656172621e-07,
+      "loss": 0.4034,
+      "step": 4500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.136982667138309e-07,
+      "loss": 0.3448,
+      "step": 4510
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.134329678103997e-07,
+      "loss": 0.4358,
+      "step": 4520
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.1316766890696853e-07,
+      "loss": 0.2735,
+      "step": 4530
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.1290237000353732e-07,
+      "loss": 0.4325,
+      "step": 4540
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.126370711001061e-07,
+      "loss": 0.3883,
+      "step": 4550
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.1237177219667492e-07,
+      "loss": 0.3338,
+      "step": 4560
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.121064732932437e-07,
+      "loss": 0.3989,
+      "step": 4570
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.118411743898125e-07,
+      "loss": 0.3722,
+      "step": 4580
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.1157587548638134e-07,
+      "loss": 0.3492,
+      "step": 4590
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.1131057658295013e-07,
+      "loss": 0.3775,
+      "step": 4600
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.1104527767951891e-07,
+      "loss": 0.3562,
+      "step": 4610
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.1077997877608773e-07,
+      "loss": 0.6652,
+      "step": 4620
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.1051467987265652e-07,
+      "loss": 0.3927,
+      "step": 4630
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.102493809692253e-07,
+      "loss": 0.3628,
+      "step": 4640
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0998408206579412e-07,
+      "loss": 0.5185,
+      "step": 4650
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0971878316236294e-07,
+      "loss": 0.3643,
+      "step": 4660
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0945348425893172e-07,
+      "loss": 0.4946,
+      "step": 4670
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0918818535550054e-07,
+      "loss": 0.4316,
+      "step": 4680
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0892288645206933e-07,
+      "loss": 0.3947,
+      "step": 4690
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0865758754863811e-07,
+      "loss": 0.3337,
+      "step": 4700
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0839228864520693e-07,
+      "loss": 0.3022,
+      "step": 4710
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0812698974177572e-07,
+      "loss": 0.3628,
+      "step": 4720
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.078616908383445e-07,
+      "loss": 0.3703,
+      "step": 4730
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0759639193491335e-07,
+      "loss": 0.275,
+      "step": 4740
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0733109303148214e-07,
+      "loss": 0.3861,
+      "step": 4750
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0706579412805092e-07,
+      "loss": 0.416,
+      "step": 4760
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0680049522461974e-07,
+      "loss": 0.3163,
+      "step": 4770
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0653519632118853e-07,
+      "loss": 0.371,
+      "step": 4780
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0626989741775732e-07,
+      "loss": 0.2409,
+      "step": 4790
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0600459851432616e-07,
+      "loss": 0.5185,
+      "step": 4800
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0573929961089494e-07,
+      "loss": 0.3575,
+      "step": 4810
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0547400070746373e-07,
+      "loss": 0.5805,
+      "step": 4820
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0520870180403255e-07,
+      "loss": 0.3353,
+      "step": 4830
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0494340290060134e-07,
+      "loss": 0.3381,
+      "step": 4840
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0467810399717012e-07,
+      "loss": 0.3048,
+      "step": 4850
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0441280509373894e-07,
+      "loss": 0.3347,
+      "step": 4860
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0414750619030773e-07,
+      "loss": 0.3736,
+      "step": 4870
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0388220728687654e-07,
+      "loss": 0.3378,
+      "step": 4880
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0361690838344536e-07,
+      "loss": 0.3411,
+      "step": 4890
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0335160948001414e-07,
+      "loss": 0.2934,
+      "step": 4900
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0308631057658293e-07,
+      "loss": 0.4118,
+      "step": 4910
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0282101167315175e-07,
+      "loss": 0.3951,
+      "step": 4920
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0255571276972054e-07,
+      "loss": 0.322,
+      "step": 4930
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0229041386628932e-07,
+      "loss": 0.3692,
+      "step": 4940
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0202511496285817e-07,
+      "loss": 0.3782,
+      "step": 4950
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.0175981605942695e-07,
+      "loss": 0.4777,
+      "step": 4960
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0149451715599574e-07,
+      "loss": 0.421,
+      "step": 4970
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0122921825256456e-07,
+      "loss": 0.3586,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0096391934913335e-07,
+      "loss": 0.3617,
+      "step": 4990
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0069862044570213e-07,
+      "loss": 0.2693,
+      "step": 5000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0043332154227095e-07,
+      "loss": 0.2447,
+      "step": 5010
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0016802263883976e-07,
+      "loss": 0.3542,
+      "step": 5020
+    },
+    {
+      "epoch": 1.2,
+      "eval_accuracy": 0.7593100744805958,
+      "eval_loss": 0.44453132152557373,
+      "eval_runtime": 710.6654,
+      "eval_samples_per_second": 3.59,
+      "eval_steps_per_second": 0.898,
+      "step": 5028
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9990272373540855e-07,
+      "loss": 0.3873,
+      "step": 5030
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9963742483197737e-07,
+      "loss": 0.3811,
+      "step": 5040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9937212592854615e-07,
+      "loss": 0.2772,
+      "step": 5050
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9910682702511494e-07,
+      "loss": 0.4806,
+      "step": 5060
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9884152812168376e-07,
+      "loss": 0.3124,
+      "step": 5070
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9857622921825255e-07,
+      "loss": 0.247,
+      "step": 5080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9831093031482133e-07,
+      "loss": 0.3793,
+      "step": 5090
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9804563141139017e-07,
+      "loss": 0.5119,
+      "step": 5100
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9778033250795896e-07,
+      "loss": 0.2919,
+      "step": 5110
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9751503360452775e-07,
+      "loss": 0.3725,
+      "step": 5120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9724973470109657e-07,
+      "loss": 0.4659,
+      "step": 5130
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9698443579766535e-07,
+      "loss": 0.3603,
+      "step": 5140
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9671913689423414e-07,
+      "loss": 0.3629,
+      "step": 5150
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9645383799080298e-07,
+      "loss": 0.4292,
+      "step": 5160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9618853908737177e-07,
+      "loss": 0.3256,
+      "step": 5170
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9592324018394056e-07,
+      "loss": 0.2833,
+      "step": 5180
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9565794128050938e-07,
+      "loss": 0.47,
+      "step": 5190
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9539264237707816e-07,
+      "loss": 0.2898,
+      "step": 5200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9512734347364695e-07,
+      "loss": 0.2773,
+      "step": 5210
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9486204457021577e-07,
+      "loss": 0.4663,
+      "step": 5220
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9459674566678455e-07,
+      "loss": 0.4463,
+      "step": 5230
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9433144676335337e-07,
+      "loss": 0.306,
+      "step": 5240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9406614785992218e-07,
+      "loss": 0.5645,
+      "step": 5250
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9380084895649097e-07,
+      "loss": 0.3439,
+      "step": 5260
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9353555005305976e-07,
+      "loss": 0.311,
+      "step": 5270
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9327025114962858e-07,
+      "loss": 0.4669,
+      "step": 5280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9300495224619736e-07,
+      "loss": 0.2812,
+      "step": 5290
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9273965334276615e-07,
+      "loss": 0.3005,
+      "step": 5300
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.92474354439335e-07,
+      "loss": 0.3576,
+      "step": 5310
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9220905553590378e-07,
+      "loss": 0.2867,
+      "step": 5320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9194375663247257e-07,
+      "loss": 0.2798,
+      "step": 5330
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9167845772904138e-07,
+      "loss": 0.3237,
+      "step": 5340
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.9141315882561017e-07,
+      "loss": 0.3129,
+      "step": 5350
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.9114785992217896e-07,
+      "loss": 0.3229,
+      "step": 5360
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.908825610187478e-07,
+      "loss": 0.4269,
+      "step": 5370
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.906172621153166e-07,
+      "loss": 0.4231,
+      "step": 5380
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.9035196321188538e-07,
+      "loss": 0.3866,
+      "step": 5390
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.900866643084542e-07,
+      "loss": 0.2466,
+      "step": 5400
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.8982136540502298e-07,
+      "loss": 0.4208,
+      "step": 5410
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.8955606650159177e-07,
+      "loss": 0.3165,
+      "step": 5420
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.8929076759816058e-07,
+      "loss": 0.3127,
+      "step": 5430
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.8902546869472937e-07,
+      "loss": 0.3945,
+      "step": 5440
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.8876016979129816e-07,
+      "loss": 0.381,
+      "step": 5450
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.88494870887867e-07,
+      "loss": 0.4172,
+      "step": 5460
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.882295719844358e-07,
+      "loss": 0.461,
+      "step": 5470
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.8796427308100458e-07,
+      "loss": 0.46,
+      "step": 5480
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.876989741775734e-07,
+      "loss": 0.2381,
+      "step": 5490
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.8743367527414218e-07,
+      "loss": 0.4672,
+      "step": 5500
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.8716837637071097e-07,
+      "loss": 0.2909,
+      "step": 5510
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.869030774672798e-07,
+      "loss": 0.4985,
+      "step": 5520
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.866377785638486e-07,
+      "loss": 0.3446,
+      "step": 5530
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.863724796604174e-07,
+      "loss": 0.3878,
+      "step": 5540
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.861071807569862e-07,
+      "loss": 0.2926,
+      "step": 5550
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.85841881853555e-07,
+      "loss": 0.2187,
+      "step": 5560
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.8557658295012378e-07,
+      "loss": 0.338,
+      "step": 5570
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.853112840466926e-07,
+      "loss": 0.2448,
+      "step": 5580
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.850459851432614e-07,
+      "loss": 0.2129,
+      "step": 5590
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.847806862398302e-07,
+      "loss": 0.3554,
+      "step": 5600
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.84515387336399e-07,
+      "loss": 0.444,
+      "step": 5610
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.842500884329678e-07,
+      "loss": 0.306,
+      "step": 5620
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.839847895295366e-07,
+      "loss": 0.4596,
+      "step": 5630
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.837194906261054e-07,
+      "loss": 0.4518,
+      "step": 5640
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.834541917226742e-07,
+      "loss": 0.2654,
+      "step": 5650
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.8318889281924298e-07,
+      "loss": 0.2054,
+      "step": 5660
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.8292359391581182e-07,
+      "loss": 0.2483,
+      "step": 5670
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.826582950123806e-07,
+      "loss": 0.3451,
+      "step": 5680
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.823929961089494e-07,
+      "loss": 0.3117,
+      "step": 5690
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.821276972055182e-07,
+      "loss": 0.2829,
+      "step": 5700
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.81862398302087e-07,
+      "loss": 0.3713,
+      "step": 5710
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.815970993986558e-07,
+      "loss": 0.2603,
+      "step": 5720
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.8133180049522463e-07,
+      "loss": 0.2378,
+      "step": 5730
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.8106650159179342e-07,
+      "loss": 0.4374,
+      "step": 5740
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.808012026883622e-07,
+      "loss": 0.4291,
+      "step": 5750
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.8053590378493102e-07,
+      "loss": 0.4139,
+      "step": 5760
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.802706048814998e-07,
+      "loss": 0.385,
+      "step": 5770
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.800053059780686e-07,
+      "loss": 0.3487,
+      "step": 5780
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.797400070746374e-07,
+      "loss": 0.316,
+      "step": 5790
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.794747081712062e-07,
+      "loss": 0.3861,
+      "step": 5800
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.7920940926777502e-07,
+      "loss": 0.3524,
+      "step": 5810
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.7894411036434383e-07,
+      "loss": 0.2247,
+      "step": 5820
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.7867881146091262e-07,
+      "loss": 0.3113,
+      "step": 5830
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.784135125574814e-07,
+      "loss": 0.5685,
+      "step": 5840
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7814821365405022e-07,
+      "loss": 0.4044,
+      "step": 5850
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.77882914750619e-07,
+      "loss": 0.3159,
+      "step": 5860
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.776176158471878e-07,
+      "loss": 0.372,
+      "step": 5870
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7735231694375664e-07,
+      "loss": 0.2586,
+      "step": 5880
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7708701804032543e-07,
+      "loss": 0.3496,
+      "step": 5890
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7682171913689422e-07,
+      "loss": 0.3252,
+      "step": 5900
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7655642023346303e-07,
+      "loss": 0.2223,
+      "step": 5910
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7629112133003182e-07,
+      "loss": 0.3757,
+      "step": 5920
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7602582242660063e-07,
+      "loss": 0.3672,
+      "step": 5930
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7576052352316945e-07,
+      "loss": 0.2711,
+      "step": 5940
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7549522461973824e-07,
+      "loss": 0.3355,
+      "step": 5950
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7522992571630705e-07,
+      "loss": 0.3543,
+      "step": 5960
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7496462681287584e-07,
+      "loss": 0.3793,
+      "step": 5970
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7469932790944463e-07,
+      "loss": 0.5126,
+      "step": 5980
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7443402900601344e-07,
+      "loss": 0.2448,
+      "step": 5990
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7416873010258223e-07,
+      "loss": 0.2939,
+      "step": 6000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7390343119915102e-07,
+      "loss": 0.246,
+      "step": 6010
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7363813229571986e-07,
+      "loss": 0.4639,
+      "step": 6020
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7337283339228865e-07,
+      "loss": 0.3554,
+      "step": 6030
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7310753448885744e-07,
+      "loss": 0.5248,
+      "step": 6040
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7284223558542625e-07,
+      "loss": 0.2787,
+      "step": 6050
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7257693668199504e-07,
+      "loss": 0.3388,
+      "step": 6060
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7231163777856383e-07,
+      "loss": 0.306,
+      "step": 6070
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7204633887513267e-07,
+      "loss": 0.3312,
+      "step": 6080
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7178103997170146e-07,
+      "loss": 0.3058,
+      "step": 6090
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.7151574106827025e-07,
+      "loss": 0.6773,
+      "step": 6100
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.7125044216483906e-07,
+      "loss": 0.2879,
+      "step": 6110
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.7098514326140785e-07,
+      "loss": 0.3534,
+      "step": 6120
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.7071984435797664e-07,
+      "loss": 0.3959,
+      "step": 6130
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.7045454545454545e-07,
+      "loss": 0.4433,
+      "step": 6140
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.7018924655111424e-07,
+      "loss": 0.3306,
+      "step": 6150
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6992394764768303e-07,
+      "loss": 0.2876,
+      "step": 6160
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6965864874425187e-07,
+      "loss": 0.2279,
+      "step": 6170
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6939334984082066e-07,
+      "loss": 0.47,
+      "step": 6180
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6912805093738945e-07,
+      "loss": 0.5794,
+      "step": 6190
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6886275203395826e-07,
+      "loss": 0.3706,
+      "step": 6200
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6859745313052705e-07,
+      "loss": 0.3674,
+      "step": 6210
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6833215422709584e-07,
+      "loss": 0.4053,
+      "step": 6220
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6806685532366468e-07,
+      "loss": 0.3724,
+      "step": 6230
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6780155642023347e-07,
+      "loss": 0.4831,
+      "step": 6240
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6753625751680225e-07,
+      "loss": 0.1894,
+      "step": 6250
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6727095861337107e-07,
+      "loss": 0.4633,
+      "step": 6260
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6700565970993986e-07,
+      "loss": 0.2882,
+      "step": 6270
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6674036080650865e-07,
+      "loss": 0.4873,
+      "step": 6280
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6647506190307746e-07,
+      "loss": 0.4315,
+      "step": 6290
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6620976299964628e-07,
+      "loss": 0.4249,
+      "step": 6300
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6594446409621506e-07,
+      "loss": 0.3353,
+      "step": 6310
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6567916519278388e-07,
+      "loss": 0.7332,
+      "step": 6320
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6541386628935267e-07,
+      "loss": 0.1961,
+      "step": 6330
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6514856738592146e-07,
+      "loss": 0.3474,
+      "step": 6340
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6488326848249027e-07,
+      "loss": 0.2356,
+      "step": 6350
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6461796957905906e-07,
+      "loss": 0.3085,
+      "step": 6360
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6435267067562785e-07,
+      "loss": 0.3192,
+      "step": 6370
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.640873717721967e-07,
+      "loss": 0.2222,
+      "step": 6380
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6382207286876548e-07,
+      "loss": 0.4065,
+      "step": 6390
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6355677396533426e-07,
+      "loss": 0.4045,
+      "step": 6400
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6329147506190308e-07,
+      "loss": 0.3183,
+      "step": 6410
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6302617615847187e-07,
+      "loss": 0.1826,
+      "step": 6420
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6276087725504066e-07,
+      "loss": 0.3059,
+      "step": 6430
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.624955783516095e-07,
+      "loss": 0.2374,
+      "step": 6440
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6223027944817828e-07,
+      "loss": 0.3193,
+      "step": 6450
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.6196498054474707e-07,
+      "loss": 0.2761,
+      "step": 6460
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.616996816413159e-07,
+      "loss": 0.3633,
+      "step": 6470
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.6143438273788468e-07,
+      "loss": 0.3956,
+      "step": 6480
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.6116908383445346e-07,
+      "loss": 0.2156,
+      "step": 6490
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.6090378493102228e-07,
+      "loss": 0.3898,
+      "step": 6500
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.6063848602759107e-07,
+      "loss": 0.2793,
+      "step": 6510
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.6037318712415988e-07,
+      "loss": 0.3232,
+      "step": 6520
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.601078882207287e-07,
+      "loss": 0.2486,
+      "step": 6530
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5984258931729749e-07,
+      "loss": 0.447,
+      "step": 6540
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5957729041386627e-07,
+      "loss": 0.368,
+      "step": 6550
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.593119915104351e-07,
+      "loss": 0.3826,
+      "step": 6560
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5904669260700388e-07,
+      "loss": 0.2456,
+      "step": 6570
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5878139370357266e-07,
+      "loss": 0.3172,
+      "step": 6580
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.585160948001415e-07,
+      "loss": 0.2154,
+      "step": 6590
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.582507958967103e-07,
+      "loss": 0.3329,
+      "step": 6600
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5798549699327908e-07,
+      "loss": 0.266,
+      "step": 6610
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.577201980898479e-07,
+      "loss": 0.3691,
+      "step": 6620
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5745489918641669e-07,
+      "loss": 0.2765,
+      "step": 6630
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5718960028298547e-07,
+      "loss": 0.2228,
+      "step": 6640
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5692430137955432e-07,
+      "loss": 0.1927,
+      "step": 6650
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.566590024761231e-07,
+      "loss": 0.4745,
+      "step": 6660
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.563937035726919e-07,
+      "loss": 0.1534,
+      "step": 6670
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.561284046692607e-07,
+      "loss": 0.1242,
+      "step": 6680
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.558631057658295e-07,
+      "loss": 0.4647,
+      "step": 6690
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5559780686239828e-07,
+      "loss": 0.5586,
+      "step": 6700
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.553325079589671e-07,
+      "loss": 0.4222,
+      "step": 6710
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5506720905553589e-07,
+      "loss": 0.206,
+      "step": 6720
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5480191015210467e-07,
+      "loss": 0.3112,
+      "step": 6730
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5453661124867352e-07,
+      "loss": 0.4122,
+      "step": 6740
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.542713123452423e-07,
+      "loss": 0.3272,
+      "step": 6750
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.540060134418111e-07,
+      "loss": 0.3635,
+      "step": 6760
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.537407145383799e-07,
+      "loss": 0.5525,
+      "step": 6770
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.534754156349487e-07,
+      "loss": 0.2226,
+      "step": 6780
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5321011673151748e-07,
+      "loss": 0.1621,
+      "step": 6790
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5294481782808632e-07,
+      "loss": 0.5783,
+      "step": 6800
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.526795189246551e-07,
+      "loss": 0.3406,
+      "step": 6810
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.524142200212239e-07,
+      "loss": 0.2203,
+      "step": 6820
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5214892111779272e-07,
+      "loss": 0.4043,
+      "step": 6830
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.518836222143615e-07,
+      "loss": 0.3353,
+      "step": 6840
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.516183233109303e-07,
+      "loss": 0.1902,
+      "step": 6850
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.513530244074991e-07,
+      "loss": 0.3804,
+      "step": 6860
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.5108772550406792e-07,
+      "loss": 0.3891,
+      "step": 6870
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.508224266006367e-07,
+      "loss": 0.2982,
+      "step": 6880
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.5055712769720552e-07,
+      "loss": 0.319,
+      "step": 6890
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.502918287937743e-07,
+      "loss": 0.3148,
+      "step": 6900
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.500265298903431e-07,
+      "loss": 0.3978,
+      "step": 6910
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4976123098691192e-07,
+      "loss": 0.2571,
+      "step": 6920
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.494959320834807e-07,
+      "loss": 0.4223,
+      "step": 6930
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4923063318004952e-07,
+      "loss": 0.4157,
+      "step": 6940
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.489653342766183e-07,
+      "loss": 0.4624,
+      "step": 6950
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4870003537318712e-07,
+      "loss": 0.2294,
+      "step": 6960
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.484347364697559e-07,
+      "loss": 0.2723,
+      "step": 6970
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4816943756632472e-07,
+      "loss": 0.3129,
+      "step": 6980
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.479041386628935e-07,
+      "loss": 0.2511,
+      "step": 6990
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4763883975946233e-07,
+      "loss": 0.4619,
+      "step": 7000
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4737354085603112e-07,
+      "loss": 0.3391,
+      "step": 7010
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4710824195259993e-07,
+      "loss": 0.3051,
+      "step": 7020
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4684294304916872e-07,
+      "loss": 0.2132,
+      "step": 7030
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.465776441457375e-07,
+      "loss": 0.3529,
+      "step": 7040
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4631234524230632e-07,
+      "loss": 0.411,
+      "step": 7050
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4604704633887514e-07,
+      "loss": 0.5373,
+      "step": 7060
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4578174743544393e-07,
+      "loss": 0.2255,
+      "step": 7070
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4551644853201271e-07,
+      "loss": 0.3849,
+      "step": 7080
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4525114962858153e-07,
+      "loss": 0.3697,
+      "step": 7090
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4498585072515032e-07,
+      "loss": 0.3,
+      "step": 7100
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4472055182171913e-07,
+      "loss": 0.2502,
+      "step": 7110
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4445525291828795e-07,
+      "loss": 0.2659,
+      "step": 7120
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4418995401485673e-07,
+      "loss": 0.2379,
+      "step": 7130
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4392465511142552e-07,
+      "loss": 0.3465,
+      "step": 7140
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4365935620799434e-07,
+      "loss": 0.2112,
+      "step": 7150
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4339405730456313e-07,
+      "loss": 0.2263,
+      "step": 7160
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4312875840113194e-07,
+      "loss": 0.2231,
+      "step": 7170
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4286345949770073e-07,
+      "loss": 0.3545,
+      "step": 7180
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4259816059426954e-07,
+      "loss": 0.3163,
+      "step": 7190
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4233286169083833e-07,
+      "loss": 0.4264,
+      "step": 7200
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4206756278740715e-07,
+      "loss": 0.1768,
+      "step": 7210
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4180226388397596e-07,
+      "loss": 0.3796,
+      "step": 7220
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4153696498054475e-07,
+      "loss": 0.3193,
+      "step": 7230
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4127166607711354e-07,
+      "loss": 0.531,
+      "step": 7240
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4100636717368235e-07,
+      "loss": 0.2121,
+      "step": 7250
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4074106827025114e-07,
+      "loss": 0.3917,
+      "step": 7260
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4047576936681996e-07,
+      "loss": 0.0989,
+      "step": 7270
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4021047046338874e-07,
+      "loss": 0.5345,
+      "step": 7280
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3994517155995753e-07,
+      "loss": 0.1947,
+      "step": 7290
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3967987265652635e-07,
+      "loss": 0.3083,
+      "step": 7300
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3941457375309516e-07,
+      "loss": 0.4358,
+      "step": 7310
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3914927484966395e-07,
+      "loss": 0.1397,
+      "step": 7320
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3888397594623274e-07,
+      "loss": 0.3461,
+      "step": 7330
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3861867704280155e-07,
+      "loss": 0.237,
+      "step": 7340
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3835337813937034e-07,
+      "loss": 0.3058,
+      "step": 7350
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3808807923593916e-07,
+      "loss": 0.2037,
+      "step": 7360
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3782278033250797e-07,
+      "loss": 0.3189,
+      "step": 7370
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3755748142907676e-07,
+      "loss": 0.2856,
+      "step": 7380
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3729218252564555e-07,
+      "loss": 0.3189,
+      "step": 7390
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3702688362221436e-07,
+      "loss": 0.27,
+      "step": 7400
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3676158471878315e-07,
+      "loss": 0.3278,
+      "step": 7410
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3649628581535196e-07,
+      "loss": 0.5432,
+      "step": 7420
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3623098691192075e-07,
+      "loss": 0.3994,
+      "step": 7430
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3596568800848954e-07,
+      "loss": 0.1706,
+      "step": 7440
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3570038910505836e-07,
+      "loss": 0.2554,
+      "step": 7450
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3543509020162717e-07,
+      "loss": 0.3261,
+      "step": 7460
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3516979129819596e-07,
+      "loss": 0.4152,
+      "step": 7470
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3490449239476477e-07,
+      "loss": 0.4484,
+      "step": 7480
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3463919349133356e-07,
+      "loss": 0.5551,
+      "step": 7490
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3437389458790235e-07,
+      "loss": 0.5414,
+      "step": 7500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3410859568447116e-07,
+      "loss": 0.3728,
+      "step": 7510
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3384329678103998e-07,
+      "loss": 0.576,
+      "step": 7520
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3357799787760877e-07,
+      "loss": 0.5139,
+      "step": 7530
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3331269897417756e-07,
+      "loss": 0.2673,
+      "step": 7540
+    },
+    {
+      "epoch": 2.2,
+      "eval_accuracy": 0.7992943943551548,
+      "eval_loss": 0.4343836009502411,
+      "eval_runtime": 667.7031,
+      "eval_samples_per_second": 3.821,
+      "eval_steps_per_second": 0.956,
+      "step": 7542
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3304740007074637e-07,
+      "loss": 0.3858,
+      "step": 7550
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3278210116731516e-07,
+      "loss": 0.4057,
+      "step": 7560
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3251680226388397e-07,
+      "loss": 0.2731,
+      "step": 7570
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.322515033604528e-07,
+      "loss": 0.3348,
+      "step": 7580
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3198620445702158e-07,
+      "loss": 0.1794,
+      "step": 7590
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3172090555359037e-07,
+      "loss": 0.2986,
+      "step": 7600
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3145560665015918e-07,
+      "loss": 0.3266,
+      "step": 7610
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3119030774672797e-07,
+      "loss": 0.3126,
+      "step": 7620
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3092500884329678e-07,
+      "loss": 0.3153,
+      "step": 7630
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3065970993986557e-07,
+      "loss": 0.3477,
+      "step": 7640
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3039441103643436e-07,
+      "loss": 0.3177,
+      "step": 7650
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3012911213300317e-07,
+      "loss": 0.2935,
+      "step": 7660
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.29863813229572e-07,
+      "loss": 0.3827,
+      "step": 7670
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.2959851432614078e-07,
+      "loss": 0.3086,
+      "step": 7680
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.293332154227096e-07,
+      "loss": 0.2633,
+      "step": 7690
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.2906791651927838e-07,
+      "loss": 0.2911,
+      "step": 7700
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.2880261761584717e-07,
+      "loss": 0.3615,
+      "step": 7710
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.2853731871241598e-07,
+      "loss": 0.4242,
+      "step": 7720
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.282720198089848e-07,
+      "loss": 0.2024,
+      "step": 7730
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.2800672090555359e-07,
+      "loss": 0.4539,
+      "step": 7740
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.2774142200212237e-07,
+      "loss": 0.2429,
+      "step": 7750
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.274761230986912e-07,
+      "loss": 0.2325,
+      "step": 7760
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.2721082419525998e-07,
+      "loss": 0.4162,
+      "step": 7770
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.269455252918288e-07,
+      "loss": 0.3089,
+      "step": 7780
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.2668022638839758e-07,
+      "loss": 0.4546,
+      "step": 7790
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.264149274849664e-07,
+      "loss": 0.2543,
+      "step": 7800
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.2614962858153518e-07,
+      "loss": 0.377,
+      "step": 7810
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.25884329678104e-07,
+      "loss": 0.3386,
+      "step": 7820
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.2561903077467279e-07,
+      "loss": 0.3492,
+      "step": 7830
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.253537318712416e-07,
+      "loss": 0.275,
+      "step": 7840
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.250884329678104e-07,
+      "loss": 0.3005,
+      "step": 7850
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.2482313406437918e-07,
+      "loss": 0.301,
+      "step": 7860
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.24557835160948e-07,
+      "loss": 0.3116,
+      "step": 7870
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.242925362575168e-07,
+      "loss": 0.6061,
+      "step": 7880
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.240272373540856e-07,
+      "loss": 0.2186,
+      "step": 7890
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.2376193845065438e-07,
+      "loss": 0.2526,
+      "step": 7900
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.234966395472232e-07,
+      "loss": 0.3165,
+      "step": 7910
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.2323134064379199e-07,
+      "loss": 0.1801,
+      "step": 7920
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.229660417403608e-07,
+      "loss": 0.4456,
+      "step": 7930
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.2270074283692962e-07,
+      "loss": 0.1817,
+      "step": 7940
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.224354439334984e-07,
+      "loss": 0.3027,
+      "step": 7950
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.221701450300672e-07,
+      "loss": 0.2641,
+      "step": 7960
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.21904846126636e-07,
+      "loss": 0.1637,
+      "step": 7970
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.216395472232048e-07,
+      "loss": 0.4759,
+      "step": 7980
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.213742483197736e-07,
+      "loss": 0.2518,
+      "step": 7990
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.211089494163424e-07,
+      "loss": 0.3353,
+      "step": 8000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.2084365051291121e-07,
+      "loss": 0.2325,
+      "step": 8010
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.2057835160948e-07,
+      "loss": 0.2963,
+      "step": 8020
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.2031305270604882e-07,
+      "loss": 0.3176,
+      "step": 8030
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.2004775380261763e-07,
+      "loss": 0.2076,
+      "step": 8040
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.1978245489918642e-07,
+      "loss": 0.3213,
+      "step": 8050
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.195171559957552e-07,
+      "loss": 0.2378,
+      "step": 8060
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.1925185709232402e-07,
+      "loss": 0.1516,
+      "step": 8070
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.1898655818889281e-07,
+      "loss": 0.2366,
+      "step": 8080
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.1872125928546161e-07,
+      "loss": 0.4461,
+      "step": 8090
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.1845596038203043e-07,
+      "loss": 0.402,
+      "step": 8100
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1819066147859922e-07,
+      "loss": 0.477,
+      "step": 8110
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1792536257516802e-07,
+      "loss": 0.331,
+      "step": 8120
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1766006367173683e-07,
+      "loss": 0.1779,
+      "step": 8130
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1739476476830562e-07,
+      "loss": 0.2126,
+      "step": 8140
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1712946586487442e-07,
+      "loss": 0.3229,
+      "step": 8150
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1686416696144322e-07,
+      "loss": 0.3967,
+      "step": 8160
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1659886805801201e-07,
+      "loss": 0.12,
+      "step": 8170
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1633356915458083e-07,
+      "loss": 0.2832,
+      "step": 8180
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1606827025114963e-07,
+      "loss": 0.2127,
+      "step": 8190
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1580297134771842e-07,
+      "loss": 0.2145,
+      "step": 8200
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1553767244428723e-07,
+      "loss": 0.2717,
+      "step": 8210
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1527237354085603e-07,
+      "loss": 0.2418,
+      "step": 8220
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1500707463742482e-07,
+      "loss": 0.2614,
+      "step": 8230
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1474177573399363e-07,
+      "loss": 0.1934,
+      "step": 8240
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1447647683056244e-07,
+      "loss": 0.4718,
+      "step": 8250
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1421117792713122e-07,
+      "loss": 0.2791,
+      "step": 8260
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1394587902370003e-07,
+      "loss": 0.3744,
+      "step": 8270
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1368058012026884e-07,
+      "loss": 0.4266,
+      "step": 8280
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1341528121683763e-07,
+      "loss": 0.2921,
+      "step": 8290
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1314998231340643e-07,
+      "loss": 0.3229,
+      "step": 8300
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1288468340997525e-07,
+      "loss": 0.3329,
+      "step": 8310
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1261938450654403e-07,
+      "loss": 0.1501,
+      "step": 8320
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1235408560311284e-07,
+      "loss": 0.3098,
+      "step": 8330
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1208878669968164e-07,
+      "loss": 0.3795,
+      "step": 8340
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1182348779625044e-07,
+      "loss": 0.2139,
+      "step": 8350
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.1155818889281924e-07,
+      "loss": 0.2172,
+      "step": 8360
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.1129288998938804e-07,
+      "loss": 0.4199,
+      "step": 8370
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.1102759108595683e-07,
+      "loss": 0.1432,
+      "step": 8380
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.1076229218252564e-07,
+      "loss": 0.2769,
+      "step": 8390
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.1049699327909445e-07,
+      "loss": 0.4188,
+      "step": 8400
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.1023169437566323e-07,
+      "loss": 0.5459,
+      "step": 8410
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0996639547223205e-07,
+      "loss": 0.5241,
+      "step": 8420
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0970109656880085e-07,
+      "loss": 0.573,
+      "step": 8430
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0943579766536964e-07,
+      "loss": 0.2892,
+      "step": 8440
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0917049876193844e-07,
+      "loss": 0.2946,
+      "step": 8450
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0890519985850725e-07,
+      "loss": 0.2701,
+      "step": 8460
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0863990095507604e-07,
+      "loss": 0.3341,
+      "step": 8470
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0837460205164484e-07,
+      "loss": 0.347,
+      "step": 8480
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0810930314821366e-07,
+      "loss": 0.3488,
+      "step": 8490
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0784400424478245e-07,
+      "loss": 0.2516,
+      "step": 8500
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0757870534135125e-07,
+      "loss": 0.4856,
+      "step": 8510
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0731340643792005e-07,
+      "loss": 0.3262,
+      "step": 8520
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0704810753448885e-07,
+      "loss": 0.4792,
+      "step": 8530
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0678280863105765e-07,
+      "loss": 0.3977,
+      "step": 8540
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0651750972762645e-07,
+      "loss": 0.247,
+      "step": 8550
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0625221082419524e-07,
+      "loss": 0.5944,
+      "step": 8560
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0598691192076406e-07,
+      "loss": 0.4538,
+      "step": 8570
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0572161301733286e-07,
+      "loss": 0.322,
+      "step": 8580
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0545631411390165e-07,
+      "loss": 0.8246,
+      "step": 8590
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0519101521047046e-07,
+      "loss": 0.2921,
+      "step": 8600
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0492571630703926e-07,
+      "loss": 0.2552,
+      "step": 8610
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0466041740360805e-07,
+      "loss": 0.2207,
+      "step": 8620
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0439511850017685e-07,
+      "loss": 0.136,
+      "step": 8630
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0412981959674567e-07,
+      "loss": 0.5574,
+      "step": 8640
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0386452069331446e-07,
+      "loss": 0.2432,
+      "step": 8650
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0359922178988326e-07,
+      "loss": 0.4907,
+      "step": 8660
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0333392288645207e-07,
+      "loss": 0.2984,
+      "step": 8670
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0306862398302086e-07,
+      "loss": 0.3477,
+      "step": 8680
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0280332507958966e-07,
+      "loss": 0.4005,
+      "step": 8690
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0253802617615848e-07,
+      "loss": 0.4083,
+      "step": 8700
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0227272727272727e-07,
+      "loss": 0.3129,
+      "step": 8710
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0200742836929607e-07,
+      "loss": 0.4127,
+      "step": 8720
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0174212946586487e-07,
+      "loss": 0.1239,
+      "step": 8730
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0147683056243366e-07,
+      "loss": 0.1863,
+      "step": 8740
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0121153165900247e-07,
+      "loss": 0.3504,
+      "step": 8750
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0094623275557127e-07,
+      "loss": 0.3059,
+      "step": 8760
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0068093385214006e-07,
+      "loss": 0.1524,
+      "step": 8770
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0041563494870888e-07,
+      "loss": 0.38,
+      "step": 8780
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0015033604527768e-07,
+      "loss": 0.3079,
+      "step": 8790
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.988503714184647e-08,
+      "loss": 0.3288,
+      "step": 8800
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.961973823841527e-08,
+      "loss": 0.554,
+      "step": 8810
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.935443933498408e-08,
+      "loss": 0.2589,
+      "step": 8820
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.908914043155288e-08,
+      "loss": 0.2241,
+      "step": 8830
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.882384152812167e-08,
+      "loss": 0.291,
+      "step": 8840
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.855854262469049e-08,
+      "loss": 0.1931,
+      "step": 8850
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.829324372125929e-08,
+      "loss": 0.2891,
+      "step": 8860
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.802794481782808e-08,
+      "loss": 0.301,
+      "step": 8870
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.776264591439689e-08,
+      "loss": 0.1596,
+      "step": 8880
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.749734701096569e-08,
+      "loss": 0.261,
+      "step": 8890
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.723204810753448e-08,
+      "loss": 0.2761,
+      "step": 8900
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.696674920410328e-08,
+      "loss": 0.2916,
+      "step": 8910
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.67014503006721e-08,
+      "loss": 0.4751,
+      "step": 8920
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.643615139724089e-08,
+      "loss": 0.2561,
+      "step": 8930
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.617085249380969e-08,
+      "loss": 0.1122,
+      "step": 8940
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.59055535903785e-08,
+      "loss": 0.1563,
+      "step": 8950
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.564025468694729e-08,
+      "loss": 0.2828,
+      "step": 8960
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.537495578351609e-08,
+      "loss": 0.3116,
+      "step": 8970
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.510965688008489e-08,
+      "loss": 0.1198,
+      "step": 8980
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.48443579766537e-08,
+      "loss": 0.2329,
+      "step": 8990
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.45790590732225e-08,
+      "loss": 0.2435,
+      "step": 9000
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.43137601697913e-08,
+      "loss": 0.2814,
+      "step": 9010
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.404846126636009e-08,
+      "loss": 0.4588,
+      "step": 9020
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.37831623629289e-08,
+      "loss": 0.3259,
+      "step": 9030
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.35178634594977e-08,
+      "loss": 0.2281,
+      "step": 9040
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.325256455606649e-08,
+      "loss": 0.1325,
+      "step": 9050
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.29872656526353e-08,
+      "loss": 0.1756,
+      "step": 9060
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.27219667492041e-08,
+      "loss": 0.4611,
+      "step": 9070
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.24566678457729e-08,
+      "loss": 0.3821,
+      "step": 9080
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.21913689423417e-08,
+      "loss": 0.4035,
+      "step": 9090
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.192607003891051e-08,
+      "loss": 0.2634,
+      "step": 9100
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.16607711354793e-08,
+      "loss": 0.5165,
+      "step": 9110
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.13954722320481e-08,
+      "loss": 0.2734,
+      "step": 9120
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.113017332861692e-08,
+      "loss": 0.3244,
+      "step": 9130
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.08648744251857e-08,
+      "loss": 0.5966,
+      "step": 9140
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.05995755217545e-08,
+      "loss": 0.2808,
+      "step": 9150
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.033427661832331e-08,
+      "loss": 0.0861,
+      "step": 9160
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.006897771489211e-08,
+      "loss": 0.5532,
+      "step": 9170
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.980367881146091e-08,
+      "loss": 0.1335,
+      "step": 9180
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.953837990802971e-08,
+      "loss": 0.2124,
+      "step": 9190
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.92730810045985e-08,
+      "loss": 0.3196,
+      "step": 9200
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.900778210116731e-08,
+      "loss": 0.3537,
+      "step": 9210
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.874248319773612e-08,
+      "loss": 0.2052,
+      "step": 9220
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.84771842943049e-08,
+      "loss": 0.2914,
+      "step": 9230
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.821188539087372e-08,
+      "loss": 0.201,
+      "step": 9240
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.794658648744252e-08,
+      "loss": 0.5735,
+      "step": 9250
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.768128758401131e-08,
+      "loss": 0.3856,
+      "step": 9260
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.741598868058011e-08,
+      "loss": 0.3567,
+      "step": 9270
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.715068977714892e-08,
+      "loss": 0.1276,
+      "step": 9280
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.688539087371771e-08,
+      "loss": 0.2726,
+      "step": 9290
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.662009197028651e-08,
+      "loss": 0.1824,
+      "step": 9300
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.635479306685533e-08,
+      "loss": 0.2436,
+      "step": 9310
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.608949416342412e-08,
+      "loss": 0.1757,
+      "step": 9320
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.582419525999292e-08,
+      "loss": 0.115,
+      "step": 9330
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.555889635656172e-08,
+      "loss": 0.3133,
+      "step": 9340
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.529359745313052e-08,
+      "loss": 0.3231,
+      "step": 9350
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.502829854969932e-08,
+      "loss": 0.3843,
+      "step": 9360
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.476299964626813e-08,
+      "loss": 0.3218,
+      "step": 9370
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.449770074283691e-08,
+      "loss": 0.2972,
+      "step": 9380
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.423240183940573e-08,
+      "loss": 0.1626,
+      "step": 9390
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.396710293597453e-08,
+      "loss": 0.3293,
+      "step": 9400
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.370180403254332e-08,
+      "loss": 0.1546,
+      "step": 9410
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.343650512911213e-08,
+      "loss": 0.2222,
+      "step": 9420
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.317120622568093e-08,
+      "loss": 0.5815,
+      "step": 9430
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.290590732224972e-08,
+      "loss": 0.4438,
+      "step": 9440
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.264060841881852e-08,
+      "loss": 0.1896,
+      "step": 9450
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.237530951538734e-08,
+      "loss": 0.5805,
+      "step": 9460
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.211001061195613e-08,
+      "loss": 0.5619,
+      "step": 9470
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.184471170852493e-08,
+      "loss": 0.256,
+      "step": 9480
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.157941280509374e-08,
+      "loss": 0.1082,
+      "step": 9490
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.131411390166253e-08,
+      "loss": 0.4024,
+      "step": 9500
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.104881499823133e-08,
+      "loss": 0.4838,
+      "step": 9510
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.078351609480015e-08,
+      "loss": 0.4411,
+      "step": 9520
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.051821719136894e-08,
+      "loss": 0.4502,
+      "step": 9530
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.025291828793774e-08,
+      "loss": 0.3072,
+      "step": 9540
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.998761938450654e-08,
+      "loss": 0.0834,
+      "step": 9550
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.972232048107533e-08,
+      "loss": 0.2351,
+      "step": 9560
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.945702157764414e-08,
+      "loss": 0.293,
+      "step": 9570
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.919172267421294e-08,
+      "loss": 0.4146,
+      "step": 9580
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.892642377078173e-08,
+      "loss": 0.2291,
+      "step": 9590
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.866112486735055e-08,
+      "loss": 0.1389,
+      "step": 9600
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.839582596391935e-08,
+      "loss": 0.2594,
+      "step": 9610
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.813052706048814e-08,
+      "loss": 0.5084,
+      "step": 9620
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.786522815705695e-08,
+      "loss": 0.4067,
+      "step": 9630
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.759992925362575e-08,
+      "loss": 0.3945,
+      "step": 9640
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.733463035019454e-08,
+      "loss": 0.3778,
+      "step": 9650
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.706933144676334e-08,
+      "loss": 0.2721,
+      "step": 9660
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.680403254333216e-08,
+      "loss": 0.304,
+      "step": 9670
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.653873363990096e-08,
+      "loss": 0.2496,
+      "step": 9680
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.627343473646975e-08,
+      "loss": 0.3475,
+      "step": 9690
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.600813583303856e-08,
+      "loss": 0.2106,
+      "step": 9700
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.574283692960736e-08,
+      "loss": 0.4637,
+      "step": 9710
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.547753802617615e-08,
+      "loss": 0.2955,
+      "step": 9720
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.521223912274495e-08,
+      "loss": 0.4538,
+      "step": 9730
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.494694021931375e-08,
+      "loss": 0.4784,
+      "step": 9740
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.468164131588256e-08,
+      "loss": 0.2853,
+      "step": 9750
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.441634241245136e-08,
+      "loss": 0.1619,
+      "step": 9760
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.415104350902016e-08,
+      "loss": 0.4363,
+      "step": 9770
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.388574460558896e-08,
+      "loss": 0.3628,
+      "step": 9780
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.362044570215776e-08,
+      "loss": 0.5235,
+      "step": 9790
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.335514679872656e-08,
+      "loss": 0.276,
+      "step": 9800
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.308984789529536e-08,
+      "loss": 0.3532,
+      "step": 9810
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.282454899186417e-08,
+      "loss": 0.3284,
+      "step": 9820
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.255925008843297e-08,
+      "loss": 0.305,
+      "step": 9830
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.229395118500176e-08,
+      "loss": 0.1895,
+      "step": 9840
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.202865228157057e-08,
+      "loss": 0.6018,
+      "step": 9850
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.176335337813937e-08,
+      "loss": 0.277,
+      "step": 9860
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 7.149805447470817e-08,
+      "loss": 0.1618,
+      "step": 9870
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 7.123275557127698e-08,
+      "loss": 0.1706,
+      "step": 9880
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 7.096745666784576e-08,
+      "loss": 0.1546,
+      "step": 9890
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 7.070215776441458e-08,
+      "loss": 0.2086,
+      "step": 9900
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 7.043685886098337e-08,
+      "loss": 0.2052,
+      "step": 9910
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 7.017155995755217e-08,
+      "loss": 0.2529,
+      "step": 9920
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.990626105412098e-08,
+      "loss": 0.2568,
+      "step": 9930
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.964096215068977e-08,
+      "loss": 0.5543,
+      "step": 9940
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.937566324725857e-08,
+      "loss": 0.4673,
+      "step": 9950
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.911036434382737e-08,
+      "loss": 0.2104,
+      "step": 9960
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.884506544039618e-08,
+      "loss": 0.1235,
+      "step": 9970
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.857976653696498e-08,
+      "loss": 0.3148,
+      "step": 9980
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.831446763353378e-08,
+      "loss": 0.2482,
+      "step": 9990
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.804916873010258e-08,
+      "loss": 0.2479,
+      "step": 10000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.778386982667138e-08,
+      "loss": 0.4603,
+      "step": 10010
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.751857092324018e-08,
+      "loss": 0.181,
+      "step": 10020
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.725327201980898e-08,
+      "loss": 0.0648,
+      "step": 10030
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.698797311637779e-08,
+      "loss": 0.1411,
+      "step": 10040
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.672267421294659e-08,
+      "loss": 0.3185,
+      "step": 10050
+    },
+    {
+      "epoch": 3.2,
+      "eval_accuracy": 0.8200705605644845,
+      "eval_loss": 0.4328227639198303,
+      "eval_runtime": 675.1569,
+      "eval_samples_per_second": 3.778,
+      "eval_steps_per_second": 0.945,
+      "step": 10056
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.645737530951539e-08,
+      "loss": 0.3281,
+      "step": 10060
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.619207640608418e-08,
+      "loss": 0.1367,
+      "step": 10070
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.592677750265299e-08,
+      "loss": 0.1347,
+      "step": 10080
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.566147859922178e-08,
+      "loss": 0.2287,
+      "step": 10090
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.539617969579058e-08,
+      "loss": 0.079,
+      "step": 10100
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.51308807923594e-08,
+      "loss": 0.4218,
+      "step": 10110
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.486558188892818e-08,
+      "loss": 0.1522,
+      "step": 10120
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.460028298549699e-08,
+      "loss": 0.2411,
+      "step": 10130
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.433498408206579e-08,
+      "loss": 0.2314,
+      "step": 10140
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.406968517863459e-08,
+      "loss": 0.4938,
+      "step": 10150
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.380438627520339e-08,
+      "loss": 0.1713,
+      "step": 10160
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.353908737177219e-08,
+      "loss": 0.3614,
+      "step": 10170
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.3273788468341e-08,
+      "loss": 0.2386,
+      "step": 10180
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.30084895649098e-08,
+      "loss": 0.4071,
+      "step": 10190
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.27431906614786e-08,
+      "loss": 0.3515,
+      "step": 10200
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.24778917580474e-08,
+      "loss": 0.3182,
+      "step": 10210
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.22125928546162e-08,
+      "loss": 0.395,
+      "step": 10220
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.1947293951185e-08,
+      "loss": 0.0892,
+      "step": 10230
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.16819950477538e-08,
+      "loss": 0.4639,
+      "step": 10240
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.141669614432259e-08,
+      "loss": 0.2954,
+      "step": 10250
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.11513972408914e-08,
+      "loss": 0.3138,
+      "step": 10260
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.088609833746021e-08,
+      "loss": 0.5433,
+      "step": 10270
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.062079943402901e-08,
+      "loss": 0.2787,
+      "step": 10280
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.035550053059781e-08,
+      "loss": 0.3296,
+      "step": 10290
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.00902016271666e-08,
+      "loss": 0.1484,
+      "step": 10300
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.982490272373541e-08,
+      "loss": 0.2677,
+      "step": 10310
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.955960382030421e-08,
+      "loss": 0.2695,
+      "step": 10320
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.9294304916873e-08,
+      "loss": 0.3003,
+      "step": 10330
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.902900601344181e-08,
+      "loss": 0.1844,
+      "step": 10340
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.876370711001061e-08,
+      "loss": 0.3414,
+      "step": 10350
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.849840820657941e-08,
+      "loss": 0.2815,
+      "step": 10360
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.8233109303148216e-08,
+      "loss": 0.2292,
+      "step": 10370
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.796781039971701e-08,
+      "loss": 0.3515,
+      "step": 10380
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.770251149628581e-08,
+      "loss": 0.3803,
+      "step": 10390
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.7437212592854614e-08,
+      "loss": 0.29,
+      "step": 10400
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.7171913689423415e-08,
+      "loss": 0.3232,
+      "step": 10410
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.690661478599221e-08,
+      "loss": 0.093,
+      "step": 10420
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.664131588256102e-08,
+      "loss": 0.2088,
+      "step": 10430
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.637601697912982e-08,
+      "loss": 0.487,
+      "step": 10440
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.6110718075698615e-08,
+      "loss": 0.2051,
+      "step": 10450
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.584541917226742e-08,
+      "loss": 0.5831,
+      "step": 10460
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.558012026883622e-08,
+      "loss": 0.4019,
+      "step": 10470
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.531482136540502e-08,
+      "loss": 0.3914,
+      "step": 10480
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.504952246197383e-08,
+      "loss": 0.1919,
+      "step": 10490
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.478422355854262e-08,
+      "loss": 0.2466,
+      "step": 10500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.451892465511142e-08,
+      "loss": 0.2837,
+      "step": 10510
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.4253625751680225e-08,
+      "loss": 0.3293,
+      "step": 10520
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.3988326848249027e-08,
+      "loss": 0.2261,
+      "step": 10530
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.372302794481782e-08,
+      "loss": 0.4414,
+      "step": 10540
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.345772904138663e-08,
+      "loss": 0.4747,
+      "step": 10550
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.3192430137955425e-08,
+      "loss": 0.297,
+      "step": 10560
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.2927131234524226e-08,
+      "loss": 0.4797,
+      "step": 10570
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.2661832331093034e-08,
+      "loss": 0.0799,
+      "step": 10580
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.239653342766183e-08,
+      "loss": 0.1772,
+      "step": 10590
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.213123452423063e-08,
+      "loss": 0.204,
+      "step": 10600
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.186593562079943e-08,
+      "loss": 0.2605,
+      "step": 10610
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.1600636717368234e-08,
+      "loss": 0.2314,
+      "step": 10620
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.133533781393703e-08,
+      "loss": 0.383,
+      "step": 10630
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.1070038910505837e-08,
+      "loss": 0.3802,
+      "step": 10640
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.080474000707463e-08,
+      "loss": 0.155,
+      "step": 10650
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.053944110364343e-08,
+      "loss": 0.4906,
+      "step": 10660
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.027414220021224e-08,
+      "loss": 0.3776,
+      "step": 10670
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.0008843296781036e-08,
+      "loss": 0.5581,
+      "step": 10680
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 4.974354439334984e-08,
+      "loss": 0.3237,
+      "step": 10690
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 4.947824548991864e-08,
+      "loss": 0.3431,
+      "step": 10700
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 4.921294658648744e-08,
+      "loss": 0.241,
+      "step": 10710
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 4.894764768305624e-08,
+      "loss": 0.389,
+      "step": 10720
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 4.8682348779625043e-08,
+      "loss": 0.4295,
+      "step": 10730
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 4.841704987619384e-08,
+      "loss": 0.2456,
+      "step": 10740
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.8151750972762646e-08,
+      "loss": 0.5222,
+      "step": 10750
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.788645206933145e-08,
+      "loss": 0.3246,
+      "step": 10760
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.762115316590024e-08,
+      "loss": 0.4138,
+      "step": 10770
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.735585426246905e-08,
+      "loss": 0.2455,
+      "step": 10780
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.7090555359037846e-08,
+      "loss": 0.1867,
+      "step": 10790
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.682525645560665e-08,
+      "loss": 0.3166,
+      "step": 10800
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.6559957552175455e-08,
+      "loss": 0.1877,
+      "step": 10810
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.629465864874425e-08,
+      "loss": 0.4553,
+      "step": 10820
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.6029359745313045e-08,
+      "loss": 0.1521,
+      "step": 10830
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.5764060841881853e-08,
+      "loss": 0.2166,
+      "step": 10840
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.5498761938450655e-08,
+      "loss": 0.1694,
+      "step": 10850
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.523346303501945e-08,
+      "loss": 0.1315,
+      "step": 10860
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 4.496816413158826e-08,
+      "loss": 0.4086,
+      "step": 10870
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.470286522815705e-08,
+      "loss": 0.3522,
+      "step": 10880
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.4437566324725854e-08,
+      "loss": 0.1301,
+      "step": 10890
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.417226742129466e-08,
+      "loss": 0.1597,
+      "step": 10900
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.390696851786346e-08,
+      "loss": 0.4818,
+      "step": 10910
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.364166961443226e-08,
+      "loss": 0.2074,
+      "step": 10920
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.337637071100106e-08,
+      "loss": 0.0484,
+      "step": 10930
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.311107180756986e-08,
+      "loss": 0.1093,
+      "step": 10940
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.2845772904138657e-08,
+      "loss": 0.3121,
+      "step": 10950
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.2580474000707465e-08,
+      "loss": 0.1872,
+      "step": 10960
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.231517509727626e-08,
+      "loss": 0.4645,
+      "step": 10970
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.204987619384506e-08,
+      "loss": 0.2252,
+      "step": 10980
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.178457729041387e-08,
+      "loss": 0.3651,
+      "step": 10990
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 4.1519278386982664e-08,
+      "loss": 0.272,
+      "step": 11000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 4.1253979483551466e-08,
+      "loss": 0.408,
+      "step": 11010
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 4.098868058012027e-08,
+      "loss": 0.2049,
+      "step": 11020
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 4.072338167668907e-08,
+      "loss": 0.1285,
+      "step": 11030
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 4.0458082773257864e-08,
+      "loss": 0.4743,
+      "step": 11040
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 4.019278386982667e-08,
+      "loss": 0.2603,
+      "step": 11050
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.9927484966395467e-08,
+      "loss": 0.5623,
+      "step": 11060
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.966218606296427e-08,
+      "loss": 0.2478,
+      "step": 11070
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.9396887159533076e-08,
+      "loss": 0.2938,
+      "step": 11080
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.913158825610187e-08,
+      "loss": 0.3001,
+      "step": 11090
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.886628935267067e-08,
+      "loss": 0.1743,
+      "step": 11100
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.8600990449239474e-08,
+      "loss": 0.3494,
+      "step": 11110
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.8335691545808276e-08,
+      "loss": 0.3701,
+      "step": 11120
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.8070392642377084e-08,
+      "loss": 0.6493,
+      "step": 11130
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.780509373894588e-08,
+      "loss": 0.2456,
+      "step": 11140
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.7539794835514673e-08,
+      "loss": 0.1634,
+      "step": 11150
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.727449593208348e-08,
+      "loss": 0.2219,
+      "step": 11160
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.700919702865228e-08,
+      "loss": 0.2627,
+      "step": 11170
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.674389812522108e-08,
+      "loss": 0.6152,
+      "step": 11180
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.647859922178988e-08,
+      "loss": 0.273,
+      "step": 11190
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.621330031835868e-08,
+      "loss": 0.2135,
+      "step": 11200
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.594800141492748e-08,
+      "loss": 0.2077,
+      "step": 11210
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.5682702511496284e-08,
+      "loss": 0.6408,
+      "step": 11220
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.5417403608065085e-08,
+      "loss": 0.1691,
+      "step": 11230
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.515210470463389e-08,
+      "loss": 0.4814,
+      "step": 11240
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.488680580120269e-08,
+      "loss": 0.1419,
+      "step": 11250
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.462150689777149e-08,
+      "loss": 0.3493,
+      "step": 11260
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.435620799434029e-08,
+      "loss": 0.5488,
+      "step": 11270
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.4090909090909086e-08,
+      "loss": 0.4774,
+      "step": 11280
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.382561018747789e-08,
+      "loss": 0.3423,
+      "step": 11290
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.356031128404669e-08,
+      "loss": 0.3831,
+      "step": 11300
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.329501238061549e-08,
+      "loss": 0.3576,
+      "step": 11310
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.302971347718429e-08,
+      "loss": 0.2333,
+      "step": 11320
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.2764414573753094e-08,
+      "loss": 0.2804,
+      "step": 11330
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.2499115670321895e-08,
+      "loss": 0.2403,
+      "step": 11340
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.22338167668907e-08,
+      "loss": 0.2754,
+      "step": 11350
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.19685178634595e-08,
+      "loss": 0.3769,
+      "step": 11360
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.170321896002829e-08,
+      "loss": 0.2209,
+      "step": 11370
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.1437920056597095e-08,
+      "loss": 0.338,
+      "step": 11380
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.11726211531659e-08,
+      "loss": 0.2673,
+      "step": 11390
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.09073222497347e-08,
+      "loss": 0.2406,
+      "step": 11400
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.06420233463035e-08,
+      "loss": 0.3665,
+      "step": 11410
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.03767244428723e-08,
+      "loss": 0.1279,
+      "step": 11420
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.01114255394411e-08,
+      "loss": 0.1948,
+      "step": 11430
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9846126636009904e-08,
+      "loss": 0.3244,
+      "step": 11440
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9580827732578705e-08,
+      "loss": 0.4653,
+      "step": 11450
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9315528829147507e-08,
+      "loss": 0.6083,
+      "step": 11460
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9050229925716305e-08,
+      "loss": 0.3703,
+      "step": 11470
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.8784931022285106e-08,
+      "loss": 0.4843,
+      "step": 11480
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.8519632118853908e-08,
+      "loss": 0.5002,
+      "step": 11490
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.8254333215422706e-08,
+      "loss": 0.1321,
+      "step": 11500
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.7989034311991508e-08,
+      "loss": 0.2233,
+      "step": 11510
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.7723735408560312e-08,
+      "loss": 0.3006,
+      "step": 11520
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.745843650512911e-08,
+      "loss": 0.5109,
+      "step": 11530
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.7193137601697912e-08,
+      "loss": 0.5902,
+      "step": 11540
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.6927838698266714e-08,
+      "loss": 0.4463,
+      "step": 11550
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.6662539794835512e-08,
+      "loss": 0.1674,
+      "step": 11560
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.6397240891404313e-08,
+      "loss": 0.3116,
+      "step": 11570
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.6131941987973115e-08,
+      "loss": 0.5825,
+      "step": 11580
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.5866643084541913e-08,
+      "loss": 0.2447,
+      "step": 11590
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.5601344181110718e-08,
+      "loss": 0.5195,
+      "step": 11600
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.533604527767952e-08,
+      "loss": 0.2391,
+      "step": 11610
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.5070746374248318e-08,
+      "loss": 0.2846,
+      "step": 11620
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.480544747081712e-08,
+      "loss": 0.1091,
+      "step": 11630
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.454014856738592e-08,
+      "loss": 0.2265,
+      "step": 11640
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.4274849663954722e-08,
+      "loss": 0.1529,
+      "step": 11650
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.400955076052352e-08,
+      "loss": 0.1775,
+      "step": 11660
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.3744251857092322e-08,
+      "loss": 0.2777,
+      "step": 11670
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.3478952953661127e-08,
+      "loss": 0.1498,
+      "step": 11680
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.3213654050229925e-08,
+      "loss": 0.0952,
+      "step": 11690
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.2948355146798726e-08,
+      "loss": 0.2005,
+      "step": 11700
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.2683056243367528e-08,
+      "loss": 0.4254,
+      "step": 11710
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.2417757339936326e-08,
+      "loss": 0.5729,
+      "step": 11720
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.2152458436505127e-08,
+      "loss": 0.2423,
+      "step": 11730
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.188715953307393e-08,
+      "loss": 0.4267,
+      "step": 11740
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.1621860629642727e-08,
+      "loss": 0.2208,
+      "step": 11750
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.1356561726211532e-08,
+      "loss": 0.2322,
+      "step": 11760
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.1091262822780334e-08,
+      "loss": 0.2109,
+      "step": 11770
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.0825963919349132e-08,
+      "loss": 0.3221,
+      "step": 11780
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.0560665015917933e-08,
+      "loss": 0.2768,
+      "step": 11790
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.0295366112486735e-08,
+      "loss": 0.161,
+      "step": 11800
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.0030067209055533e-08,
+      "loss": 0.4017,
+      "step": 11810
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.9764768305624334e-08,
+      "loss": 0.4653,
+      "step": 11820
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.9499469402193136e-08,
+      "loss": 0.3184,
+      "step": 11830
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.923417049876194e-08,
+      "loss": 0.2338,
+      "step": 11840
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.896887159533074e-08,
+      "loss": 0.5184,
+      "step": 11850
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.870357269189954e-08,
+      "loss": 0.3949,
+      "step": 11860
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.843827378846834e-08,
+      "loss": 0.2224,
+      "step": 11870
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.817297488503714e-08,
+      "loss": 0.2273,
+      "step": 11880
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.790767598160594e-08,
+      "loss": 0.4915,
+      "step": 11890
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.7642377078174743e-08,
+      "loss": 0.2633,
+      "step": 11900
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.7377078174743545e-08,
+      "loss": 0.4765,
+      "step": 11910
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.7111779271312346e-08,
+      "loss": 0.3887,
+      "step": 11920
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.6846480367881144e-08,
+      "loss": 0.4244,
+      "step": 11930
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.6581181464449946e-08,
+      "loss": 0.3747,
+      "step": 11940
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.6315882561018747e-08,
+      "loss": 0.2129,
+      "step": 11950
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.6050583657587546e-08,
+      "loss": 0.2404,
+      "step": 11960
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.578528475415635e-08,
+      "loss": 0.3692,
+      "step": 11970
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.551998585072515e-08,
+      "loss": 0.3378,
+      "step": 11980
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.525468694729395e-08,
+      "loss": 0.4417,
+      "step": 11990
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.498938804386275e-08,
+      "loss": 0.3159,
+      "step": 12000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.4724089140431551e-08,
+      "loss": 0.4654,
+      "step": 12010
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.4458790237000355e-08,
+      "loss": 0.2322,
+      "step": 12020
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.4193491333569154e-08,
+      "loss": 0.2092,
+      "step": 12030
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.3928192430137954e-08,
+      "loss": 0.2385,
+      "step": 12040
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.3662893526706756e-08,
+      "loss": 0.3575,
+      "step": 12050
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.3397594623275557e-08,
+      "loss": 0.2266,
+      "step": 12060
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.3132295719844357e-08,
+      "loss": 0.234,
+      "step": 12070
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.2866996816413159e-08,
+      "loss": 0.2282,
+      "step": 12080
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.2601697912981958e-08,
+      "loss": 0.2577,
+      "step": 12090
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.2336399009550762e-08,
+      "loss": 0.3597,
+      "step": 12100
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.2071100106119561e-08,
+      "loss": 0.4429,
+      "step": 12110
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.1805801202688361e-08,
+      "loss": 0.1702,
+      "step": 12120
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.1540502299257163e-08,
+      "loss": 0.2193,
+      "step": 12130
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.1275203395825964e-08,
+      "loss": 0.4391,
+      "step": 12140
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.1009904492394764e-08,
+      "loss": 0.2284,
+      "step": 12150
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.0744605588963566e-08,
+      "loss": 0.489,
+      "step": 12160
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.0479306685532365e-08,
+      "loss": 0.1444,
+      "step": 12170
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.0214007782101165e-08,
+      "loss": 0.2967,
+      "step": 12180
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 9.948708878669968e-09,
+      "loss": 0.3594,
+      "step": 12190
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 9.683409975238768e-09,
+      "loss": 0.5359,
+      "step": 12200
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 9.41811107180757e-09,
+      "loss": 0.2757,
+      "step": 12210
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 9.152812168376371e-09,
+      "loss": 0.3935,
+      "step": 12220
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.887513264945171e-09,
+      "loss": 0.14,
+      "step": 12230
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.622214361513971e-09,
+      "loss": 0.4049,
+      "step": 12240
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.356915458082773e-09,
+      "loss": 0.2653,
+      "step": 12250
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.091616554651574e-09,
+      "loss": 0.1909,
+      "step": 12260
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.826317651220374e-09,
+      "loss": 0.2591,
+      "step": 12270
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.561018747789175e-09,
+      "loss": 0.1013,
+      "step": 12280
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.295719844357976e-09,
+      "loss": 0.1527,
+      "step": 12290
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.0304209409267776e-09,
+      "loss": 0.1982,
+      "step": 12300
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 6.765122037495577e-09,
+      "loss": 0.3196,
+      "step": 12310
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 6.499823134064379e-09,
+      "loss": 0.2233,
+      "step": 12320
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 6.2345242306331796e-09,
+      "loss": 0.2557,
+      "step": 12330
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5.969225327201981e-09,
+      "loss": 0.1804,
+      "step": 12340
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5.703926423770781e-09,
+      "loss": 0.2618,
+      "step": 12350
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5.4386275203395824e-09,
+      "loss": 0.4291,
+      "step": 12360
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5.173328616908383e-09,
+      "loss": 0.2799,
+      "step": 12370
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 4.908029713477184e-09,
+      "loss": 0.1811,
+      "step": 12380
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.6427308100459845e-09,
+      "loss": 0.2275,
+      "step": 12390
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.377431906614786e-09,
+      "loss": 0.4325,
+      "step": 12400
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.112133003183587e-09,
+      "loss": 0.22,
+      "step": 12410
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.846834099752387e-09,
+      "loss": 0.22,
+      "step": 12420
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.5815351963211884e-09,
+      "loss": 0.199,
+      "step": 12430
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.3162362928899895e-09,
+      "loss": 0.4157,
+      "step": 12440
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.05093738945879e-09,
+      "loss": 0.4612,
+      "step": 12450
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.7856384860275913e-09,
+      "loss": 0.2144,
+      "step": 12460
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.520339582596392e-09,
+      "loss": 0.3051,
+      "step": 12470
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.2550406791651926e-09,
+      "loss": 0.4694,
+      "step": 12480
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.9897417757339937e-09,
+      "loss": 0.5988,
+      "step": 12490
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.7244428723027944e-09,
+      "loss": 0.1659,
+      "step": 12500
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.4591439688715953e-09,
+      "loss": 0.4332,
+      "step": 12510
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.1938450654403961e-09,
+      "loss": 0.2026,
+      "step": 12520
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 9.28546162009197e-10,
+      "loss": 0.3017,
+      "step": 12530
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 6.632472585779978e-10,
+      "loss": 0.2656,
+      "step": 12540
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 3.979483551467987e-10,
+      "loss": 0.5015,
+      "step": 12550
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.3264945171559958e-10,
+      "loss": 0.4006,
+      "step": 12560
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 0.8251666013328106,
+      "eval_loss": 0.44296565651893616,
+      "eval_runtime": 623.4351,
+      "eval_samples_per_second": 4.092,
+      "eval_steps_per_second": 1.023,
+      "step": 12565
+    },
+    {
+      "epoch": 4.2,
+      "step": 12565,
+      "total_flos": 2.2067541850155437e+20,
+      "train_loss": 0.39015588782197014,
+      "train_runtime": 24412.2348,
+      "train_samples_per_second": 2.059,
+      "train_steps_per_second": 0.515
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 0.8251666013328106,
+      "eval_loss": 0.44296565651893616,
+      "eval_runtime": 654.7979,
+      "eval_samples_per_second": 3.896,
+      "eval_steps_per_second": 0.974,
+      "step": 12565
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 0.8251666013328106,
+      "eval_loss": 0.44296565651893616,
+      "eval_runtime": 750.9501,
+      "eval_samples_per_second": 3.397,
+      "eval_steps_per_second": 0.85,
+      "step": 12565
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 12565,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "total_flos": 2.2067541850155437e+20,
+  "trial_name": null,
+  "trial_params": null
+}