Model save

all_results.json

@@ -1,8 +1,13 @@
 {
-    "epoch": 5.0,
-    "total_flos": 6.230614598311477e+18,
-    "train_loss": 0.0,
-    "train_runtime": 0.1997,
-    "train_samples_per_second": 1252007.126,
-    "train_steps_per_second": 19556.351
+    "epoch": 8.0,
+    "eval_accuracy": 0.8735,
+    "eval_loss": 0.42226287722587585,
+    "eval_runtime": 61.173,
+    "eval_samples_per_second": 163.471,
+    "eval_steps_per_second": 10.217,
+    "total_flos": 9.976400755629687e+18,
+    "train_loss": 0.05874636431585964,
+    "train_runtime": 2017.8469,
+    "train_samples_per_second": 495.578,
+    "train_steps_per_second": 7.741
 }

config.json

@@ -1,7 +1,7 @@
 {
-  "_name_or_path": "microsoft/swin-tiny-patch4-window7-224",
+  "_name_or_path": "/kaggle/working/swin-tiny-finetuned-cifar100/checkpoint-3905",
   "architectures": [
-    "SwinForImageClassification"
+    "SwinModel"
   ],
   "attention_probs_dropout_prob": 0.0,
   "depths": [
@@ -19,6 +19,14 @@
   "id2label": {
     "0": "apple",
     "1": "aquarium_fish",
+    "2": "baby",
+    "3": "bear",
+    "4": "beaver",
+    "5": "bed",
+    "6": "bee",
+    "7": "beetle",
+    "8": "bicycle",
+    "9": "bottle",
     "10": "bowl",
     "11": "boy",
     "12": "bridge",
@@ -29,7 +37,6 @@
     "17": "castle",
     "18": "caterpillar",
     "19": "cattle",
-    "2": "baby",
     "20": "chair",
     "21": "chimpanzee",
     "22": "clock",
@@ -40,7 +47,6 @@
     "27": "crocodile",
     "28": "cup",
     "29": "dinosaur",
-    "3": "bear",
     "30": "dolphin",
     "31": "elephant",
     "32": "flatfish",
@@ -51,7 +57,6 @@
     "37": "house",
     "38": "kangaroo",
     "39": "keyboard",
-    "4": "beaver",
     "40": "lamp",
     "41": "lawn_mower",
     "42": "leopard",
@@ -62,7 +67,6 @@
     "47": "maple_tree",
     "48": "motorcycle",
     "49": "mountain",
-    "5": "bed",
     "50": "mouse",
     "51": "mushroom",
     "52": "oak_tree",
@@ -73,7 +77,6 @@
     "57": "pear",
     "58": "pickup_truck",
     "59": "pine_tree",
-    "6": "bee",
     "60": "plain",
     "61": "plate",
     "62": "poppy",
@@ -84,7 +87,6 @@
     "67": "ray",
     "68": "road",
     "69": "rocket",
-    "7": "beetle",
     "70": "rose",
     "71": "sea",
     "72": "seal",
@@ -95,7 +97,6 @@
     "77": "snail",
     "78": "snake",
     "79": "spider",
-    "8": "bicycle",
     "80": "squirrel",
     "81": "streetcar",
     "82": "sunflower",
@@ -106,7 +107,6 @@
     "87": "television",
     "88": "tiger",
     "89": "tractor",
-    "9": "bottle",
     "90": "train",
     "91": "trout",
     "92": "tulip",

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8735,
+    "eval_loss": 0.42226287722587585,
+    "eval_runtime": 61.173,
+    "eval_samples_per_second": 163.471,
+    "eval_steps_per_second": 10.217
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:baa87f4c1ffe36c11a1fa3f67e3e6bceb22f0891fe2b34a83d3b03bf4a9de91e
-size 110694319
+oid sha256:1cd68ce90d0ac44e1dcc3ffc94a3ea936d53d1d140d491ba7c308a255951e938
+size 110384947

runs/Dec23_19-28-07_9f2aa0a785d5/events.out.tfevents.1671827711.9f2aa0a785d5.23.9

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1765181576f43b57309a5c5bb9cfb8e4f6bd9235317484a2968084480d8c509d
+size 686

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 5.0,
-    "total_flos": 6.230614598311477e+18,
-    "train_loss": 0.0,
-    "train_runtime": 0.1997,
-    "train_samples_per_second": 1252007.126,
-    "train_steps_per_second": 19556.351
+    "epoch": 8.0,
+    "total_flos": 9.976400755629687e+18,
+    "train_loss": 0.05874636431585964,
+    "train_runtime": 2017.8469,
+    "train_samples_per_second": 495.578,
+    "train_steps_per_second": 7.741
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": 0.8735,
   "best_model_checkpoint": "swin-tiny-finetuned-cifar100/checkpoint-3905",
-  "epoch": 4.99968,
-  "global_step": 3905,
+  "epoch": 7.99968,
+  "global_step": 6248,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -4739,18 +4739,2853 @@
       "step": 3905
     },
     {
-      "epoch": 5.0,
-      "step": 3905,
-      "total_flos": 6.230614598311477e+18,
-      "train_loss": 0.0,
-      "train_runtime": 0.1997,
-      "train_samples_per_second": 1252007.126,
-      "train_steps_per_second": 19556.351
+      "epoch": 5.01,
+      "learning_rate": 3.331910655854318e-05,
+      "loss": 0.1285,
+      "step": 3910
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 3.330487978375302e-05,
+      "loss": 0.1142,
+      "step": 3915
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 3.329065300896287e-05,
+      "loss": 0.1851,
+      "step": 3920
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 3.327642623417272e-05,
+      "loss": 0.1502,
+      "step": 3925
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 3.326219945938256e-05,
+      "loss": 0.1461,
+      "step": 3930
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 3.3247972684592405e-05,
+      "loss": 0.1968,
+      "step": 3935
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 3.3233745909802246e-05,
+      "loss": 0.1589,
+      "step": 3940
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 3.321951913501209e-05,
+      "loss": 0.1806,
+      "step": 3945
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 3.320529236022194e-05,
+      "loss": 0.1707,
+      "step": 3950
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 3.319106558543178e-05,
+      "loss": 0.1472,
+      "step": 3955
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 3.317683881064163e-05,
+      "loss": 0.1815,
+      "step": 3960
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 3.3162612035851475e-05,
+      "loss": 0.2247,
+      "step": 3965
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 3.314838526106132e-05,
+      "loss": 0.1818,
+      "step": 3970
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 3.313415848627116e-05,
+      "loss": 0.13,
+      "step": 3975
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 3.311993171148101e-05,
+      "loss": 0.2268,
+      "step": 3980
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 3.310570493669086e-05,
+      "loss": 0.1953,
+      "step": 3985
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 3.30914781619007e-05,
+      "loss": 0.1226,
+      "step": 3990
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 3.3077251387110545e-05,
+      "loss": 0.1584,
+      "step": 3995
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 3.306302461232039e-05,
+      "loss": 0.1786,
+      "step": 4000
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 3.304879783753023e-05,
+      "loss": 0.1962,
+      "step": 4005
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 3.303457106274008e-05,
+      "loss": 0.1649,
+      "step": 4010
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 3.302034428794993e-05,
+      "loss": 0.2497,
+      "step": 4015
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 3.300611751315977e-05,
+      "loss": 0.1831,
+      "step": 4020
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 3.2991890738369616e-05,
+      "loss": 0.1659,
+      "step": 4025
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 3.297766396357946e-05,
+      "loss": 0.1938,
+      "step": 4030
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 3.2963437188789303e-05,
+      "loss": 0.2537,
+      "step": 4035
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 3.294921041399915e-05,
+      "loss": 0.2067,
+      "step": 4040
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 3.2934983639209e-05,
+      "loss": 0.2082,
+      "step": 4045
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 3.292075686441884e-05,
+      "loss": 0.2009,
+      "step": 4050
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 3.2906530089628686e-05,
+      "loss": 0.2421,
+      "step": 4055
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 3.2892303314838526e-05,
+      "loss": 0.1824,
+      "step": 4060
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 3.2878076540048374e-05,
+      "loss": 0.1365,
+      "step": 4065
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 3.286384976525822e-05,
+      "loss": 0.1509,
+      "step": 4070
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 3.284962299046806e-05,
+      "loss": 0.2064,
+      "step": 4075
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 3.283539621567791e-05,
+      "loss": 0.1502,
+      "step": 4080
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 3.282116944088775e-05,
+      "loss": 0.1109,
+      "step": 4085
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 3.2806942666097596e-05,
+      "loss": 0.1713,
+      "step": 4090
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 3.2792715891307444e-05,
+      "loss": 0.1816,
+      "step": 4095
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 3.2778489116517284e-05,
+      "loss": 0.1827,
+      "step": 4100
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 3.276426234172713e-05,
+      "loss": 0.3017,
+      "step": 4105
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 3.275003556693698e-05,
+      "loss": 0.1865,
+      "step": 4110
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 3.273580879214682e-05,
+      "loss": 0.21,
+      "step": 4115
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 3.272158201735667e-05,
+      "loss": 0.1569,
+      "step": 4120
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 3.2707355242566514e-05,
+      "loss": 0.2136,
+      "step": 4125
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 3.2693128467776354e-05,
+      "loss": 0.2465,
+      "step": 4130
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 3.26789016929862e-05,
+      "loss": 0.2049,
+      "step": 4135
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 3.266467491819605e-05,
+      "loss": 0.2559,
+      "step": 4140
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 3.265044814340589e-05,
+      "loss": 0.183,
+      "step": 4145
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 3.263622136861574e-05,
+      "loss": 0.2204,
+      "step": 4150
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 3.2621994593825584e-05,
+      "loss": 0.1661,
+      "step": 4155
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 3.2607767819035425e-05,
+      "loss": 0.2355,
+      "step": 4160
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 3.259354104424527e-05,
+      "loss": 0.2613,
+      "step": 4165
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 3.257931426945512e-05,
+      "loss": 0.1896,
+      "step": 4170
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 3.2565087494664966e-05,
+      "loss": 0.2326,
+      "step": 4175
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 3.255086071987481e-05,
+      "loss": 0.1743,
+      "step": 4180
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 3.2536633945084654e-05,
+      "loss": 0.12,
+      "step": 4185
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 3.25224071702945e-05,
+      "loss": 0.1525,
+      "step": 4190
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 3.250818039550434e-05,
+      "loss": 0.223,
+      "step": 4195
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 3.249395362071419e-05,
+      "loss": 0.1799,
+      "step": 4200
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 3.247972684592404e-05,
+      "loss": 0.2391,
+      "step": 4205
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 3.246550007113388e-05,
+      "loss": 0.222,
+      "step": 4210
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 3.2451273296343724e-05,
+      "loss": 0.1821,
+      "step": 4215
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 3.2437046521553565e-05,
+      "loss": 0.2977,
+      "step": 4220
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 3.242281974676341e-05,
+      "loss": 0.1666,
+      "step": 4225
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 3.240859297197326e-05,
+      "loss": 0.225,
+      "step": 4230
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 3.23943661971831e-05,
+      "loss": 0.1507,
+      "step": 4235
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 3.238013942239295e-05,
+      "loss": 0.1905,
+      "step": 4240
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 3.236591264760279e-05,
+      "loss": 0.2153,
+      "step": 4245
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 3.2351685872812635e-05,
+      "loss": 0.1666,
+      "step": 4250
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 3.2337459098022476e-05,
+      "loss": 0.2062,
+      "step": 4255
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 3.232323232323232e-05,
+      "loss": 0.178,
+      "step": 4260
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 3.230900554844217e-05,
+      "loss": 0.2068,
+      "step": 4265
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 3.229477877365201e-05,
+      "loss": 0.2439,
+      "step": 4270
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 3.228055199886186e-05,
+      "loss": 0.2313,
+      "step": 4275
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 3.2266325224071705e-05,
+      "loss": 0.1938,
+      "step": 4280
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 3.225209844928155e-05,
+      "loss": 0.1819,
+      "step": 4285
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 3.223787167449139e-05,
+      "loss": 0.2532,
+      "step": 4290
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 3.222364489970124e-05,
+      "loss": 0.2471,
+      "step": 4295
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 3.220941812491109e-05,
+      "loss": 0.2413,
+      "step": 4300
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 3.219519135012093e-05,
+      "loss": 0.1755,
+      "step": 4305
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 3.2180964575330775e-05,
+      "loss": 0.2192,
+      "step": 4310
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 3.216673780054062e-05,
+      "loss": 0.1986,
+      "step": 4315
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 3.215251102575046e-05,
+      "loss": 0.2097,
+      "step": 4320
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 3.213828425096031e-05,
+      "loss": 0.2471,
+      "step": 4325
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 3.212405747617016e-05,
+      "loss": 0.3077,
+      "step": 4330
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 3.210983070138e-05,
+      "loss": 0.1834,
+      "step": 4335
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 3.2095603926589846e-05,
+      "loss": 0.2494,
+      "step": 4340
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 3.208137715179969e-05,
+      "loss": 0.1973,
+      "step": 4345
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 3.2067150377009533e-05,
+      "loss": 0.2002,
+      "step": 4350
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 3.205292360221938e-05,
+      "loss": 0.1791,
+      "step": 4355
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 3.203869682742923e-05,
+      "loss": 0.19,
+      "step": 4360
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 3.202447005263907e-05,
+      "loss": 0.199,
+      "step": 4365
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 3.2010243277848916e-05,
+      "loss": 0.1823,
+      "step": 4370
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 3.199601650305876e-05,
+      "loss": 0.2362,
+      "step": 4375
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 3.1981789728268604e-05,
+      "loss": 0.2319,
+      "step": 4380
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 3.196756295347845e-05,
+      "loss": 0.1667,
+      "step": 4385
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 3.195333617868829e-05,
+      "loss": 0.2043,
+      "step": 4390
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 3.193910940389814e-05,
+      "loss": 0.1878,
+      "step": 4395
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 3.1924882629107986e-05,
+      "loss": 0.2179,
+      "step": 4400
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 3.1910655854317827e-05,
+      "loss": 0.2462,
+      "step": 4405
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 3.1896429079527674e-05,
+      "loss": 0.156,
+      "step": 4410
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 3.1882202304737514e-05,
+      "loss": 0.209,
+      "step": 4415
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 3.186797552994736e-05,
+      "loss": 0.2115,
+      "step": 4420
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 3.185374875515721e-05,
+      "loss": 0.1601,
+      "step": 4425
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 3.183952198036705e-05,
+      "loss": 0.1846,
+      "step": 4430
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 3.18252952055769e-05,
+      "loss": 0.1996,
+      "step": 4435
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 3.1811068430786744e-05,
+      "loss": 0.2333,
+      "step": 4440
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 3.1796841655996584e-05,
+      "loss": 0.2045,
+      "step": 4445
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 3.178261488120643e-05,
+      "loss": 0.1738,
+      "step": 4450
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 3.176838810641628e-05,
+      "loss": 0.2008,
+      "step": 4455
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 3.175416133162612e-05,
+      "loss": 0.2208,
+      "step": 4460
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 3.173993455683597e-05,
+      "loss": 0.2452,
+      "step": 4465
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 3.1725707782045814e-05,
+      "loss": 0.2009,
+      "step": 4470
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 3.1711481007255655e-05,
+      "loss": 0.2077,
+      "step": 4475
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 3.16972542324655e-05,
+      "loss": 0.2101,
+      "step": 4480
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 3.168302745767535e-05,
+      "loss": 0.2303,
+      "step": 4485
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 3.1668800682885197e-05,
+      "loss": 0.178,
+      "step": 4490
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 3.165457390809504e-05,
+      "loss": 0.2568,
+      "step": 4495
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 3.1640347133304884e-05,
+      "loss": 0.2905,
+      "step": 4500
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 3.162612035851473e-05,
+      "loss": 0.1791,
+      "step": 4505
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 3.161189358372457e-05,
+      "loss": 0.2422,
+      "step": 4510
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 3.159766680893442e-05,
+      "loss": 0.2769,
+      "step": 4515
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 3.158344003414427e-05,
+      "loss": 0.1691,
+      "step": 4520
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 3.156921325935411e-05,
+      "loss": 0.2357,
+      "step": 4525
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 3.1554986484563954e-05,
+      "loss": 0.2224,
+      "step": 4530
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 3.1540759709773795e-05,
+      "loss": 0.1777,
+      "step": 4535
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 3.152653293498364e-05,
+      "loss": 0.2366,
+      "step": 4540
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 3.151230616019349e-05,
+      "loss": 0.2274,
+      "step": 4545
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 3.149807938540333e-05,
+      "loss": 0.2297,
+      "step": 4550
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 3.148385261061318e-05,
+      "loss": 0.1871,
+      "step": 4555
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 3.146962583582302e-05,
+      "loss": 0.2088,
+      "step": 4560
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 3.1455399061032865e-05,
+      "loss": 0.2148,
+      "step": 4565
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 3.144117228624271e-05,
+      "loss": 0.1979,
+      "step": 4570
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 3.142694551145255e-05,
+      "loss": 0.1561,
+      "step": 4575
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 3.14127187366624e-05,
+      "loss": 0.272,
+      "step": 4580
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 3.139849196187225e-05,
+      "loss": 0.2451,
+      "step": 4585
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 3.138426518708209e-05,
+      "loss": 0.1517,
+      "step": 4590
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 3.1370038412291935e-05,
+      "loss": 0.264,
+      "step": 4595
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 3.135581163750178e-05,
+      "loss": 0.2004,
+      "step": 4600
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 3.134158486271162e-05,
+      "loss": 0.2062,
+      "step": 4605
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 3.132735808792147e-05,
+      "loss": 0.231,
+      "step": 4610
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 3.131313131313132e-05,
+      "loss": 0.2246,
+      "step": 4615
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 3.129890453834116e-05,
+      "loss": 0.2241,
+      "step": 4620
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 3.1284677763551006e-05,
+      "loss": 0.1962,
+      "step": 4625
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 3.127045098876085e-05,
+      "loss": 0.2359,
+      "step": 4630
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 3.125622421397069e-05,
+      "loss": 0.2372,
+      "step": 4635
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 3.124199743918054e-05,
+      "loss": 0.1803,
+      "step": 4640
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 3.122777066439039e-05,
+      "loss": 0.1941,
+      "step": 4645
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 3.121354388960023e-05,
+      "loss": 0.1994,
+      "step": 4650
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 3.1199317114810076e-05,
+      "loss": 0.1674,
+      "step": 4655
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 3.118509034001992e-05,
+      "loss": 0.1825,
+      "step": 4660
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 3.1170863565229763e-05,
+      "loss": 0.183,
+      "step": 4665
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 3.115663679043961e-05,
+      "loss": 0.2467,
+      "step": 4670
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 3.114241001564946e-05,
+      "loss": 0.2009,
+      "step": 4675
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 3.11281832408593e-05,
+      "loss": 0.2184,
+      "step": 4680
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.1113956466069146e-05,
+      "loss": 0.2589,
+      "step": 4685
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8622,
+      "eval_loss": 0.4774980843067169,
+      "eval_runtime": 58.8421,
+      "eval_samples_per_second": 169.946,
+      "eval_steps_per_second": 10.622,
+      "step": 4686
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 3.109972969127899e-05,
+      "loss": 0.1885,
+      "step": 4690
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 3.1085502916488834e-05,
+      "loss": 0.1172,
+      "step": 4695
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 3.107127614169868e-05,
+      "loss": 0.1204,
+      "step": 4700
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 3.105704936690853e-05,
+      "loss": 0.1306,
+      "step": 4705
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 3.104282259211837e-05,
+      "loss": 0.1688,
+      "step": 4710
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 3.1028595817328216e-05,
+      "loss": 0.1379,
+      "step": 4715
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 3.1014369042538057e-05,
+      "loss": 0.0961,
+      "step": 4720
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 3.1000142267747904e-05,
+      "loss": 0.1494,
+      "step": 4725
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 3.0985915492957744e-05,
+      "loss": 0.1411,
+      "step": 4730
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 3.097168871816759e-05,
+      "loss": 0.1237,
+      "step": 4735
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 3.095746194337744e-05,
+      "loss": 0.2125,
+      "step": 4740
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 3.094323516858728e-05,
+      "loss": 0.0896,
+      "step": 4745
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 3.092900839379713e-05,
+      "loss": 0.1399,
+      "step": 4750
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 3.0914781619006974e-05,
+      "loss": 0.1609,
+      "step": 4755
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 3.0900554844216815e-05,
+      "loss": 0.143,
+      "step": 4760
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 3.088632806942666e-05,
+      "loss": 0.1704,
+      "step": 4765
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 3.087210129463651e-05,
+      "loss": 0.1383,
+      "step": 4770
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 3.085787451984635e-05,
+      "loss": 0.1563,
+      "step": 4775
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 3.08436477450562e-05,
+      "loss": 0.1264,
+      "step": 4780
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 3.0829420970266044e-05,
+      "loss": 0.1716,
+      "step": 4785
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 3.0815194195475885e-05,
+      "loss": 0.162,
+      "step": 4790
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 3.080096742068573e-05,
+      "loss": 0.1408,
+      "step": 4795
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 3.078674064589558e-05,
+      "loss": 0.1411,
+      "step": 4800
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 3.0772513871105427e-05,
+      "loss": 0.1521,
+      "step": 4805
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 3.075828709631527e-05,
+      "loss": 0.1162,
+      "step": 4810
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 3.0744060321525114e-05,
+      "loss": 0.209,
+      "step": 4815
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 3.072983354673496e-05,
+      "loss": 0.1504,
+      "step": 4820
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 3.07156067719448e-05,
+      "loss": 0.1631,
+      "step": 4825
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 3.070137999715465e-05,
+      "loss": 0.1975,
+      "step": 4830
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 3.06871532223645e-05,
+      "loss": 0.155,
+      "step": 4835
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 3.067292644757434e-05,
+      "loss": 0.1489,
+      "step": 4840
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 3.0658699672784185e-05,
+      "loss": 0.218,
+      "step": 4845
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 3.064447289799403e-05,
+      "loss": 0.144,
+      "step": 4850
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 3.063024612320387e-05,
+      "loss": 0.1304,
+      "step": 4855
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 3.061601934841372e-05,
+      "loss": 0.1426,
+      "step": 4860
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 3.060179257362356e-05,
+      "loss": 0.1589,
+      "step": 4865
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 3.058756579883341e-05,
+      "loss": 0.1159,
+      "step": 4870
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 3.0573339024043255e-05,
+      "loss": 0.1731,
+      "step": 4875
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 3.0559112249253095e-05,
+      "loss": 0.1831,
+      "step": 4880
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 3.054488547446294e-05,
+      "loss": 0.1269,
+      "step": 4885
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 3.053065869967278e-05,
+      "loss": 0.1489,
+      "step": 4890
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 3.0516431924882634e-05,
+      "loss": 0.1796,
+      "step": 4895
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 3.0502205150092478e-05,
+      "loss": 0.1082,
+      "step": 4900
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 3.048797837530232e-05,
+      "loss": 0.1524,
+      "step": 4905
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 3.0473751600512165e-05,
+      "loss": 0.1445,
+      "step": 4910
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 3.0459524825722013e-05,
+      "loss": 0.1245,
+      "step": 4915
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 3.0445298050931857e-05,
+      "loss": 0.2082,
+      "step": 4920
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 3.04310712761417e-05,
+      "loss": 0.1317,
+      "step": 4925
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 3.0416844501351548e-05,
+      "loss": 0.1398,
+      "step": 4930
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 3.0402617726561388e-05,
+      "loss": 0.1535,
+      "step": 4935
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 3.0388390951771236e-05,
+      "loss": 0.1526,
+      "step": 4940
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 3.0374164176981083e-05,
+      "loss": 0.1455,
+      "step": 4945
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 3.0359937402190923e-05,
+      "loss": 0.1585,
+      "step": 4950
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 3.034571062740077e-05,
+      "loss": 0.0949,
+      "step": 4955
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 3.0331483852610618e-05,
+      "loss": 0.1112,
+      "step": 4960
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 3.031725707782046e-05,
+      "loss": 0.1918,
+      "step": 4965
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.1097,
+      "step": 4970
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 3.0288803528240153e-05,
+      "loss": 0.1625,
+      "step": 4975
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 3.0274576753449994e-05,
+      "loss": 0.216,
+      "step": 4980
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 3.026034997865984e-05,
+      "loss": 0.1963,
+      "step": 4985
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 3.0246123203869685e-05,
+      "loss": 0.0797,
+      "step": 4990
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 3.023189642907953e-05,
+      "loss": 0.1663,
+      "step": 4995
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 3.0217669654289376e-05,
+      "loss": 0.1426,
+      "step": 5000
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 3.020344287949922e-05,
+      "loss": 0.1107,
+      "step": 5005
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 3.0189216104709067e-05,
+      "loss": 0.1893,
+      "step": 5010
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 3.0174989329918908e-05,
+      "loss": 0.165,
+      "step": 5015
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 3.0160762555128755e-05,
+      "loss": 0.1391,
+      "step": 5020
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 3.0146535780338602e-05,
+      "loss": 0.0906,
+      "step": 5025
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 3.0132309005548443e-05,
+      "loss": 0.1649,
+      "step": 5030
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 3.011808223075829e-05,
+      "loss": 0.1355,
+      "step": 5035
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 3.0103855455968137e-05,
+      "loss": 0.1781,
+      "step": 5040
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 3.0089628681177978e-05,
+      "loss": 0.161,
+      "step": 5045
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 3.0075401906387825e-05,
+      "loss": 0.15,
+      "step": 5050
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 3.0061175131597672e-05,
+      "loss": 0.1811,
+      "step": 5055
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 3.0046948356807513e-05,
+      "loss": 0.1895,
+      "step": 5060
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 3.003272158201736e-05,
+      "loss": 0.1377,
+      "step": 5065
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 3.0018494807227204e-05,
+      "loss": 0.1933,
+      "step": 5070
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 3.0004268032437048e-05,
+      "loss": 0.1105,
+      "step": 5075
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 2.9990041257646895e-05,
+      "loss": 0.134,
+      "step": 5080
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 2.997581448285674e-05,
+      "loss": 0.1476,
+      "step": 5085
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 2.9961587708066583e-05,
+      "loss": 0.1469,
+      "step": 5090
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 2.9947360933276427e-05,
+      "loss": 0.1646,
+      "step": 5095
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 2.9933134158486274e-05,
+      "loss": 0.2304,
+      "step": 5100
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 2.9918907383696115e-05,
+      "loss": 0.1486,
+      "step": 5105
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 2.9904680608905962e-05,
+      "loss": 0.1456,
+      "step": 5110
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 2.989045383411581e-05,
+      "loss": 0.138,
+      "step": 5115
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 2.9876227059325657e-05,
+      "loss": 0.126,
+      "step": 5120
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 2.9862000284535497e-05,
+      "loss": 0.1452,
+      "step": 5125
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 2.9847773509745344e-05,
+      "loss": 0.181,
+      "step": 5130
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 2.9833546734955188e-05,
+      "loss": 0.1938,
+      "step": 5135
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 2.9819319960165032e-05,
+      "loss": 0.1802,
+      "step": 5140
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 2.980509318537488e-05,
+      "loss": 0.1488,
+      "step": 5145
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 2.9790866410584723e-05,
+      "loss": 0.1856,
+      "step": 5150
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 2.9776639635794567e-05,
+      "loss": 0.1021,
+      "step": 5155
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 2.976241286100441e-05,
+      "loss": 0.1413,
+      "step": 5160
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 2.974818608621426e-05,
+      "loss": 0.1775,
+      "step": 5165
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 2.9733959311424102e-05,
+      "loss": 0.095,
+      "step": 5170
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 2.9719732536633946e-05,
+      "loss": 0.1737,
+      "step": 5175
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 2.9705505761843794e-05,
+      "loss": 0.1217,
+      "step": 5180
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 2.9691278987053634e-05,
+      "loss": 0.0981,
+      "step": 5185
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 2.967705221226348e-05,
+      "loss": 0.1533,
+      "step": 5190
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 2.966282543747333e-05,
+      "loss": 0.211,
+      "step": 5195
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 2.964859866268317e-05,
+      "loss": 0.1413,
+      "step": 5200
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 2.9634371887893016e-05,
+      "loss": 0.1936,
+      "step": 5205
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 2.9620145113102864e-05,
+      "loss": 0.1873,
+      "step": 5210
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 2.9605918338312708e-05,
+      "loss": 0.1684,
+      "step": 5215
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 2.959169156352255e-05,
+      "loss": 0.142,
+      "step": 5220
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 2.95774647887324e-05,
+      "loss": 0.1253,
+      "step": 5225
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 2.9563238013942243e-05,
+      "loss": 0.1739,
+      "step": 5230
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 2.9549011239152087e-05,
+      "loss": 0.1478,
+      "step": 5235
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 2.953478446436193e-05,
+      "loss": 0.1689,
+      "step": 5240
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 2.9520557689571778e-05,
+      "loss": 0.1662,
+      "step": 5245
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 2.950633091478162e-05,
+      "loss": 0.136,
+      "step": 5250
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 2.9492104139991466e-05,
+      "loss": 0.1305,
+      "step": 5255
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 2.9477877365201313e-05,
+      "loss": 0.1186,
+      "step": 5260
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 2.9463650590411153e-05,
+      "loss": 0.092,
+      "step": 5265
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 2.9449423815621e-05,
+      "loss": 0.1752,
+      "step": 5270
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 2.9435197040830848e-05,
+      "loss": 0.1549,
+      "step": 5275
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 2.942097026604069e-05,
+      "loss": 0.1603,
+      "step": 5280
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 2.9406743491250536e-05,
+      "loss": 0.1791,
+      "step": 5285
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 2.9392516716460383e-05,
+      "loss": 0.1655,
+      "step": 5290
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 2.9378289941670224e-05,
+      "loss": 0.1347,
+      "step": 5295
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 2.936406316688007e-05,
+      "loss": 0.1365,
+      "step": 5300
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 2.9349836392089918e-05,
+      "loss": 0.1827,
+      "step": 5305
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 2.933560961729976e-05,
+      "loss": 0.2096,
+      "step": 5310
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 2.9321382842509606e-05,
+      "loss": 0.1595,
+      "step": 5315
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 2.930715606771945e-05,
+      "loss": 0.1254,
+      "step": 5320
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 2.9292929292929297e-05,
+      "loss": 0.2147,
+      "step": 5325
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 2.927870251813914e-05,
+      "loss": 0.0987,
+      "step": 5330
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 2.9264475743348985e-05,
+      "loss": 0.187,
+      "step": 5335
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 2.9250248968558832e-05,
+      "loss": 0.1033,
+      "step": 5340
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 2.9236022193768673e-05,
+      "loss": 0.1558,
+      "step": 5345
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 2.922179541897852e-05,
+      "loss": 0.1656,
+      "step": 5350
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 2.9207568644188367e-05,
+      "loss": 0.1316,
+      "step": 5355
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 2.9193341869398208e-05,
+      "loss": 0.1157,
+      "step": 5360
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 2.9179115094608055e-05,
+      "loss": 0.1095,
+      "step": 5365
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 2.9164888319817902e-05,
+      "loss": 0.1144,
+      "step": 5370
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 2.9150661545027743e-05,
+      "loss": 0.1884,
+      "step": 5375
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 2.913643477023759e-05,
+      "loss": 0.1534,
+      "step": 5380
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 2.9122207995447434e-05,
+      "loss": 0.1372,
+      "step": 5385
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 2.9107981220657278e-05,
+      "loss": 0.2208,
+      "step": 5390
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 2.9093754445867125e-05,
+      "loss": 0.1677,
+      "step": 5395
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 2.907952767107697e-05,
+      "loss": 0.1714,
+      "step": 5400
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 2.9065300896286813e-05,
+      "loss": 0.1314,
+      "step": 5405
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 2.9051074121496657e-05,
+      "loss": 0.1602,
+      "step": 5410
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 2.9036847346706504e-05,
+      "loss": 0.1298,
+      "step": 5415
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 2.9022620571916348e-05,
+      "loss": 0.1928,
+      "step": 5420
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 2.9008393797126192e-05,
+      "loss": 0.2219,
+      "step": 5425
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 2.899416702233604e-05,
+      "loss": 0.141,
+      "step": 5430
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 2.8979940247545887e-05,
+      "loss": 0.1183,
+      "step": 5435
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 2.8965713472755727e-05,
+      "loss": 0.1559,
+      "step": 5440
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 2.8951486697965574e-05,
+      "loss": 0.1167,
+      "step": 5445
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 2.893725992317542e-05,
+      "loss": 0.129,
+      "step": 5450
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 2.8923033148385262e-05,
+      "loss": 0.1713,
+      "step": 5455
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 2.890880637359511e-05,
+      "loss": 0.1204,
+      "step": 5460
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.8894579598804953e-05,
+      "loss": 0.1419,
+      "step": 5465
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8642,
+      "eval_loss": 0.490004301071167,
+      "eval_runtime": 59.9772,
+      "eval_samples_per_second": 166.73,
+      "eval_steps_per_second": 10.421,
+      "step": 5467
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.8880352824014797e-05,
+      "loss": 0.1436,
+      "step": 5470
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 2.8866126049224645e-05,
+      "loss": 0.0989,
+      "step": 5475
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 2.885189927443449e-05,
+      "loss": 0.1386,
+      "step": 5480
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 2.8837672499644332e-05,
+      "loss": 0.1292,
+      "step": 5485
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 2.8823445724854176e-05,
+      "loss": 0.0895,
+      "step": 5490
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 2.8809218950064024e-05,
+      "loss": 0.1196,
+      "step": 5495
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 2.8794992175273867e-05,
+      "loss": 0.1005,
+      "step": 5500
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 2.878076540048371e-05,
+      "loss": 0.1087,
+      "step": 5505
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 2.876653862569356e-05,
+      "loss": 0.0914,
+      "step": 5510
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 2.87523118509034e-05,
+      "loss": 0.1147,
+      "step": 5515
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 2.8738085076113246e-05,
+      "loss": 0.1071,
+      "step": 5520
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 2.8723858301323094e-05,
+      "loss": 0.1358,
+      "step": 5525
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 2.870963152653294e-05,
+      "loss": 0.09,
+      "step": 5530
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 2.869540475174278e-05,
+      "loss": 0.1079,
+      "step": 5535
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 2.868117797695263e-05,
+      "loss": 0.0831,
+      "step": 5540
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 2.8666951202162473e-05,
+      "loss": 0.097,
+      "step": 5545
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 2.8652724427372317e-05,
+      "loss": 0.1185,
+      "step": 5550
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 2.8638497652582164e-05,
+      "loss": 0.1015,
+      "step": 5555
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 2.8624270877792008e-05,
+      "loss": 0.1327,
+      "step": 5560
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 2.8610044103001852e-05,
+      "loss": 0.1123,
+      "step": 5565
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 2.8595817328211696e-05,
+      "loss": 0.108,
+      "step": 5570
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 2.8581590553421543e-05,
+      "loss": 0.1279,
+      "step": 5575
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 2.8567363778631387e-05,
+      "loss": 0.1033,
+      "step": 5580
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 2.855313700384123e-05,
+      "loss": 0.0746,
+      "step": 5585
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 2.8538910229051078e-05,
+      "loss": 0.0952,
+      "step": 5590
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 2.852468345426092e-05,
+      "loss": 0.0975,
+      "step": 5595
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 2.8510456679470766e-05,
+      "loss": 0.085,
+      "step": 5600
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 2.8496229904680613e-05,
+      "loss": 0.1156,
+      "step": 5605
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 2.8482003129890454e-05,
+      "loss": 0.0986,
+      "step": 5610
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 2.84677763551003e-05,
+      "loss": 0.2043,
+      "step": 5615
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 2.8453549580310148e-05,
+      "loss": 0.1123,
+      "step": 5620
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 2.843932280551999e-05,
+      "loss": 0.1262,
+      "step": 5625
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 2.8425096030729836e-05,
+      "loss": 0.0971,
+      "step": 5630
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 2.841086925593968e-05,
+      "loss": 0.1056,
+      "step": 5635
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 2.8396642481149527e-05,
+      "loss": 0.1241,
+      "step": 5640
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 2.838241570635937e-05,
+      "loss": 0.0894,
+      "step": 5645
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 2.8368188931569215e-05,
+      "loss": 0.0844,
+      "step": 5650
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 2.8353962156779062e-05,
+      "loss": 0.1121,
+      "step": 5655
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 2.8339735381988903e-05,
+      "loss": 0.1291,
+      "step": 5660
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 2.832550860719875e-05,
+      "loss": 0.1287,
+      "step": 5665
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 2.8311281832408597e-05,
+      "loss": 0.0948,
+      "step": 5670
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 2.8297055057618438e-05,
+      "loss": 0.0957,
+      "step": 5675
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 2.8282828282828285e-05,
+      "loss": 0.1279,
+      "step": 5680
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 2.8268601508038132e-05,
+      "loss": 0.0871,
+      "step": 5685
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 2.8254374733247973e-05,
+      "loss": 0.0998,
+      "step": 5690
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 2.824014795845782e-05,
+      "loss": 0.1114,
+      "step": 5695
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 2.8225921183667667e-05,
+      "loss": 0.2043,
+      "step": 5700
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 2.8211694408877508e-05,
+      "loss": 0.1361,
+      "step": 5705
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 2.8197467634087355e-05,
+      "loss": 0.1129,
+      "step": 5710
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 2.81832408592972e-05,
+      "loss": 0.1198,
+      "step": 5715
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 2.8169014084507043e-05,
+      "loss": 0.0771,
+      "step": 5720
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 2.815478730971689e-05,
+      "loss": 0.0767,
+      "step": 5725
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 2.8140560534926734e-05,
+      "loss": 0.1241,
+      "step": 5730
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 2.8126333760136578e-05,
+      "loss": 0.1501,
+      "step": 5735
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 2.8112106985346422e-05,
+      "loss": 0.1063,
+      "step": 5740
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 2.809788021055627e-05,
+      "loss": 0.0859,
+      "step": 5745
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 2.8083653435766117e-05,
+      "loss": 0.0773,
+      "step": 5750
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 2.8069426660975957e-05,
+      "loss": 0.1341,
+      "step": 5755
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 2.8055199886185804e-05,
+      "loss": 0.1036,
+      "step": 5760
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 2.8040973111395652e-05,
+      "loss": 0.0864,
+      "step": 5765
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 2.8026746336605492e-05,
+      "loss": 0.14,
+      "step": 5770
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 2.801251956181534e-05,
+      "loss": 0.069,
+      "step": 5775
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 2.7998292787025187e-05,
+      "loss": 0.1248,
+      "step": 5780
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 2.7984066012235027e-05,
+      "loss": 0.1395,
+      "step": 5785
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 2.7969839237444875e-05,
+      "loss": 0.1216,
+      "step": 5790
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 2.795561246265472e-05,
+      "loss": 0.1188,
+      "step": 5795
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 2.7941385687864562e-05,
+      "loss": 0.1534,
+      "step": 5800
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 2.792715891307441e-05,
+      "loss": 0.1341,
+      "step": 5805
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 2.7912932138284254e-05,
+      "loss": 0.093,
+      "step": 5810
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 2.7898705363494097e-05,
+      "loss": 0.1099,
+      "step": 5815
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 2.788447858870394e-05,
+      "loss": 0.1631,
+      "step": 5820
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 2.787025181391379e-05,
+      "loss": 0.0992,
+      "step": 5825
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 2.785602503912363e-05,
+      "loss": 0.1339,
+      "step": 5830
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 2.7841798264333476e-05,
+      "loss": 0.0857,
+      "step": 5835
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 2.7827571489543324e-05,
+      "loss": 0.1258,
+      "step": 5840
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 2.781334471475317e-05,
+      "loss": 0.1297,
+      "step": 5845
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 2.779911793996301e-05,
+      "loss": 0.0738,
+      "step": 5850
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 2.778489116517286e-05,
+      "loss": 0.1067,
+      "step": 5855
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 2.7770664390382706e-05,
+      "loss": 0.1489,
+      "step": 5860
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 2.7756437615592547e-05,
+      "loss": 0.1238,
+      "step": 5865
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 2.7742210840802394e-05,
+      "loss": 0.1042,
+      "step": 5870
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 2.7727984066012238e-05,
+      "loss": 0.1127,
+      "step": 5875
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 2.7713757291222082e-05,
+      "loss": 0.1076,
+      "step": 5880
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 2.7699530516431926e-05,
+      "loss": 0.1292,
+      "step": 5885
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 2.7685303741641773e-05,
+      "loss": 0.1058,
+      "step": 5890
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 2.7671076966851617e-05,
+      "loss": 0.0919,
+      "step": 5895
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 2.765685019206146e-05,
+      "loss": 0.1429,
+      "step": 5900
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 2.7642623417271308e-05,
+      "loss": 0.106,
+      "step": 5905
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 2.762839664248115e-05,
+      "loss": 0.086,
+      "step": 5910
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 2.7614169867690996e-05,
+      "loss": 0.1095,
+      "step": 5915
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 2.7599943092900843e-05,
+      "loss": 0.1004,
+      "step": 5920
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 2.7585716318110684e-05,
+      "loss": 0.1113,
+      "step": 5925
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 2.757148954332053e-05,
+      "loss": 0.1335,
+      "step": 5930
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 2.7557262768530378e-05,
+      "loss": 0.0664,
+      "step": 5935
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 2.754303599374022e-05,
+      "loss": 0.1322,
+      "step": 5940
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 2.7528809218950066e-05,
+      "loss": 0.0609,
+      "step": 5945
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 2.7514582444159913e-05,
+      "loss": 0.1927,
+      "step": 5950
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 2.7500355669369757e-05,
+      "loss": 0.0889,
+      "step": 5955
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 2.74861288945796e-05,
+      "loss": 0.1464,
+      "step": 5960
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 2.7471902119789445e-05,
+      "loss": 0.1118,
+      "step": 5965
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 2.7457675344999292e-05,
+      "loss": 0.0954,
+      "step": 5970
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 2.7443448570209136e-05,
+      "loss": 0.1134,
+      "step": 5975
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 2.742922179541898e-05,
+      "loss": 0.0965,
+      "step": 5980
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 2.7414995020628827e-05,
+      "loss": 0.1133,
+      "step": 5985
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 2.7400768245838668e-05,
+      "loss": 0.1118,
+      "step": 5990
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 2.7386541471048515e-05,
+      "loss": 0.1065,
+      "step": 5995
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 2.7372314696258362e-05,
+      "loss": 0.1189,
+      "step": 6000
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 2.7358087921468203e-05,
+      "loss": 0.104,
+      "step": 6005
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 2.734386114667805e-05,
+      "loss": 0.1,
+      "step": 6010
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 2.7329634371887898e-05,
+      "loss": 0.1041,
+      "step": 6015
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 2.7315407597097738e-05,
+      "loss": 0.139,
+      "step": 6020
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 2.7301180822307585e-05,
+      "loss": 0.0824,
+      "step": 6025
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 2.7286954047517433e-05,
+      "loss": 0.1583,
+      "step": 6030
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.1049,
+      "step": 6035
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 2.725850049793712e-05,
+      "loss": 0.18,
+      "step": 6040
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 2.7244273723146964e-05,
+      "loss": 0.1121,
+      "step": 6045
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 2.723004694835681e-05,
+      "loss": 0.0981,
+      "step": 6050
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 2.7215820173566655e-05,
+      "loss": 0.1083,
+      "step": 6055
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 2.72015933987765e-05,
+      "loss": 0.1267,
+      "step": 6060
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 2.7187366623986347e-05,
+      "loss": 0.1045,
+      "step": 6065
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 2.7173139849196187e-05,
+      "loss": 0.1824,
+      "step": 6070
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 2.7158913074406034e-05,
+      "loss": 0.1479,
+      "step": 6075
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 2.7144686299615882e-05,
+      "loss": 0.1746,
+      "step": 6080
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 2.7130459524825722e-05,
+      "loss": 0.1604,
+      "step": 6085
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 2.711623275003557e-05,
+      "loss": 0.0781,
+      "step": 6090
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 2.7102005975245417e-05,
+      "loss": 0.0829,
+      "step": 6095
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 2.7087779200455257e-05,
+      "loss": 0.1885,
+      "step": 6100
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 2.7073552425665105e-05,
+      "loss": 0.1002,
+      "step": 6105
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 2.705932565087495e-05,
+      "loss": 0.1507,
+      "step": 6110
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 2.7045098876084792e-05,
+      "loss": 0.1359,
+      "step": 6115
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 2.703087210129464e-05,
+      "loss": 0.119,
+      "step": 6120
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 2.7016645326504484e-05,
+      "loss": 0.1012,
+      "step": 6125
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 2.7002418551714328e-05,
+      "loss": 0.1272,
+      "step": 6130
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 2.698819177692417e-05,
+      "loss": 0.0746,
+      "step": 6135
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 2.697396500213402e-05,
+      "loss": 0.1057,
+      "step": 6140
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 2.6959738227343863e-05,
+      "loss": 0.0748,
+      "step": 6145
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 2.6945511452553707e-05,
+      "loss": 0.0647,
+      "step": 6150
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 2.6931284677763554e-05,
+      "loss": 0.1712,
+      "step": 6155
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 2.69170579029734e-05,
+      "loss": 0.1391,
+      "step": 6160
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 2.690283112818324e-05,
+      "loss": 0.1111,
+      "step": 6165
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 2.688860435339309e-05,
+      "loss": 0.0622,
+      "step": 6170
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 2.6874377578602936e-05,
+      "loss": 0.1563,
+      "step": 6175
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 2.6860150803812777e-05,
+      "loss": 0.1771,
+      "step": 6180
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 2.6845924029022624e-05,
+      "loss": 0.1229,
+      "step": 6185
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 2.6831697254232468e-05,
+      "loss": 0.1404,
+      "step": 6190
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 2.6817470479442312e-05,
+      "loss": 0.1076,
+      "step": 6195
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 2.680324370465216e-05,
+      "loss": 0.1175,
+      "step": 6200
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 2.6789016929862003e-05,
+      "loss": 0.1649,
+      "step": 6205
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 2.6774790155071847e-05,
+      "loss": 0.1314,
+      "step": 6210
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 2.676056338028169e-05,
+      "loss": 0.1571,
+      "step": 6215
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 2.6746336605491538e-05,
+      "loss": 0.1484,
+      "step": 6220
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 2.6732109830701382e-05,
+      "loss": 0.1064,
+      "step": 6225
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 2.6717883055911226e-05,
+      "loss": 0.0954,
+      "step": 6230
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 2.6703656281121073e-05,
+      "loss": 0.0797,
+      "step": 6235
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 2.6689429506330914e-05,
+      "loss": 0.142,
+      "step": 6240
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 2.667520273154076e-05,
+      "loss": 0.1513,
+      "step": 6245
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8667,
+      "eval_loss": 0.49564066529273987,
+      "eval_runtime": 60.3762,
+      "eval_samples_per_second": 165.628,
+      "eval_steps_per_second": 10.352,
+      "step": 6248
+    },
+    {
+      "epoch": 8.0,
+      "step": 6248,
+      "total_flos": 9.976400755629687e+18,
+      "train_loss": 0.05874636431585964,
+      "train_runtime": 2017.8469,
+      "train_samples_per_second": 495.578,
+      "train_steps_per_second": 7.741
     }
   ],
-  "max_steps": 3905,
-  "num_train_epochs": 5,
-  "total_flos": 6.230614598311477e+18,
+  "max_steps": 15620,
+  "num_train_epochs": 20,
+  "total_flos": 9.976400755629687e+18,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fd4fa7a5742bbd6246bbd60ac6ffb547d10f3ccac8955980fa25bd0d41e6549b
+oid sha256:15a4e5d8a601da8dd43edd99b047858d3da32c6b32780a1e59087ddca1fc7092
 size 3311