End of training

README.md

@@ -2,6 +2,8 @@
 license: apache-2.0
 base_model: google/vit-base-patch16-224-in21k
 tags:
+- image-classification
+- vision
 - generated_from_trainer
 datasets:
 - imagefolder

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 100.0,
-    "eval_accuracy": 0.8431372549019608,
-    "eval_loss": 1.2935162782669067,
-    "eval_runtime": 0.797,
-    "eval_samples_per_second": 127.985,
-    "eval_steps_per_second": 1.255,
-    "train_loss": 1.6557193438212077,
-    "train_runtime": 2084.5935,
-    "train_samples_per_second": 27.535,
-    "train_steps_per_second": 0.288
+    "epoch": 120.0,
+    "eval_accuracy": 0.8627450980392157,
+    "eval_loss": 1.1135263442993164,
+    "eval_runtime": 0.7629,
+    "eval_samples_per_second": 133.697,
+    "eval_steps_per_second": 1.311,
+    "train_loss": 1.4664161258273654,
+    "train_runtime": 3110.7108,
+    "train_samples_per_second": 22.143,
+    "train_steps_per_second": 0.231
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 100.0,
-    "eval_accuracy": 0.8431372549019608,
-    "eval_loss": 1.2935162782669067,
-    "eval_runtime": 0.797,
-    "eval_samples_per_second": 127.985,
-    "eval_steps_per_second": 1.255
+    "epoch": 120.0,
+    "eval_accuracy": 0.8627450980392157,
+    "eval_loss": 1.1135263442993164,
+    "eval_runtime": 0.7629,
+    "eval_samples_per_second": 133.697,
+    "eval_steps_per_second": 1.311
 }

runs/Dec18_02-31-04_da87b97e44af/events.out.tfevents.1702869814.da87b97e44af.3584.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaef516daefcea7cd3682e5c7620dd7abad4914915521609fdfc46a6b6b30a9d
+size 411

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 100.0,
-    "train_loss": 1.6557193438212077,
-    "train_runtime": 2084.5935,
-    "train_samples_per_second": 27.535,
-    "train_steps_per_second": 0.288
+    "epoch": 120.0,
+    "train_loss": 1.4664161258273654,
+    "train_runtime": 3110.7108,
+    "train_samples_per_second": 22.143,
+    "train_steps_per_second": 0.231
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 1.2935162782669067,
-  "best_model_checkpoint": "./drive/MyDrive/repositories/torch_example_image-classification/outputs2/checkpoint-600",
-  "epoch": 100.0,
+  "best_metric": 1.1135263442993164,
+  "best_model_checkpoint": "./drive/MyDrive/repositories/torch_example_image-classification/outputs2/checkpoint-720",
+  "epoch": 120.0,
   "eval_steps": 500,
-  "global_step": 600,
+  "global_step": 720,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -11,1279 +11,1531 @@
     {
       "epoch": 1.0,
       "eval_accuracy": 0.08823529411764706,
-      "eval_loss": 3.2330737113952637,
-      "eval_runtime": 0.7234,
-      "eval_samples_per_second": 141.008,
-      "eval_steps_per_second": 1.382,
+      "eval_loss": 3.2330594062805176,
+      "eval_runtime": 47.8321,
+      "eval_samples_per_second": 2.132,
+      "eval_steps_per_second": 0.021,
       "step": 6
     },
     {
       "epoch": 1.67,
-      "learning_rate": 1.9666666666666666e-05,
+      "learning_rate": 1.9722222222222224e-05,
       "loss": 3.2363,
       "step": 10
     },
     {
       "epoch": 2.0,
       "eval_accuracy": 0.13725490196078433,
-      "eval_loss": 3.202636957168579,
-      "eval_runtime": 0.6909,
-      "eval_samples_per_second": 147.63,
-      "eval_steps_per_second": 1.447,
+      "eval_loss": 3.2025418281555176,
+      "eval_runtime": 0.6721,
+      "eval_samples_per_second": 151.767,
+      "eval_steps_per_second": 1.488,
       "step": 12
     },
     {
       "epoch": 3.0,
       "eval_accuracy": 0.18627450980392157,
-      "eval_loss": 3.1764001846313477,
-      "eval_runtime": 0.8751,
-      "eval_samples_per_second": 116.553,
-      "eval_steps_per_second": 1.143,
+      "eval_loss": 3.176102876663208,
+      "eval_runtime": 0.7424,
+      "eval_samples_per_second": 137.386,
+      "eval_steps_per_second": 1.347,
       "step": 18
     },
     {
       "epoch": 3.33,
-      "learning_rate": 1.9333333333333333e-05,
-      "loss": 3.1624,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 3.1622,
       "step": 20
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.22549019607843138,
-      "eval_loss": 3.124112844467163,
-      "eval_runtime": 1.9559,
-      "eval_samples_per_second": 52.149,
-      "eval_steps_per_second": 0.511,
+      "eval_loss": 3.1237781047821045,
+      "eval_runtime": 0.911,
+      "eval_samples_per_second": 111.961,
+      "eval_steps_per_second": 1.098,
       "step": 24
     },
     {
       "epoch": 5.0,
-      "learning_rate": 1.9e-05,
-      "loss": 3.0924,
+      "learning_rate": 1.916666666666667e-05,
+      "loss": 3.0918,
       "step": 30
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.30392156862745096,
-      "eval_loss": 3.0796170234680176,
-      "eval_runtime": 0.7875,
-      "eval_samples_per_second": 129.524,
-      "eval_steps_per_second": 1.27,
+      "eval_accuracy": 0.3137254901960784,
+      "eval_loss": 3.078857183456421,
+      "eval_runtime": 0.8136,
+      "eval_samples_per_second": 125.367,
+      "eval_steps_per_second": 1.229,
       "step": 30
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.3137254901960784,
-      "eval_loss": 3.0296630859375,
-      "eval_runtime": 0.6746,
-      "eval_samples_per_second": 151.208,
-      "eval_steps_per_second": 1.482,
+      "eval_accuracy": 0.3235294117647059,
+      "eval_loss": 3.028043508529663,
+      "eval_runtime": 0.7696,
+      "eval_samples_per_second": 132.528,
+      "eval_steps_per_second": 1.299,
       "step": 36
     },
     {
       "epoch": 6.67,
-      "learning_rate": 1.866666666666667e-05,
-      "loss": 3.0094,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 3.0081,
       "step": 40
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.3333333333333333,
-      "eval_loss": 2.9901208877563477,
-      "eval_runtime": 0.8568,
-      "eval_samples_per_second": 119.047,
-      "eval_steps_per_second": 1.167,
+      "eval_accuracy": 0.3431372549019608,
+      "eval_loss": 2.9877936840057373,
+      "eval_runtime": 0.695,
+      "eval_samples_per_second": 146.754,
+      "eval_steps_per_second": 1.439,
       "step": 42
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.4019607843137255,
-      "eval_loss": 2.933021306991577,
-      "eval_runtime": 0.865,
-      "eval_samples_per_second": 117.919,
-      "eval_steps_per_second": 1.156,
+      "eval_accuracy": 0.38235294117647056,
+      "eval_loss": 2.931645393371582,
+      "eval_runtime": 0.8566,
+      "eval_samples_per_second": 119.07,
+      "eval_steps_per_second": 1.167,
       "step": 48
     },
     {
       "epoch": 8.33,
-      "learning_rate": 1.8333333333333333e-05,
-      "loss": 2.9143,
+      "learning_rate": 1.8611111111111114e-05,
+      "loss": 2.9118,
       "step": 50
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.4117647058823529,
-      "eval_loss": 2.888922691345215,
-      "eval_runtime": 0.9798,
-      "eval_samples_per_second": 104.098,
-      "eval_steps_per_second": 1.021,
+      "eval_accuracy": 0.43137254901960786,
+      "eval_loss": 2.8864240646362305,
+      "eval_runtime": 0.7935,
+      "eval_samples_per_second": 128.548,
+      "eval_steps_per_second": 1.26,
       "step": 54
     },
     {
       "epoch": 10.0,
-      "learning_rate": 1.8e-05,
-      "loss": 2.8269,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 2.8231,
       "step": 60
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.4411764705882353,
-      "eval_loss": 2.8364291191101074,
-      "eval_runtime": 0.6724,
-      "eval_samples_per_second": 151.704,
-      "eval_steps_per_second": 1.487,
+      "eval_accuracy": 0.45098039215686275,
+      "eval_loss": 2.831418752670288,
+      "eval_runtime": 0.7311,
+      "eval_samples_per_second": 139.516,
+      "eval_steps_per_second": 1.368,
       "step": 60
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.5294117647058824,
-      "eval_loss": 2.785728693008423,
-      "eval_runtime": 0.8028,
-      "eval_samples_per_second": 127.055,
-      "eval_steps_per_second": 1.246,
+      "eval_accuracy": 0.5196078431372549,
+      "eval_loss": 2.781670331954956,
+      "eval_runtime": 1.0842,
+      "eval_samples_per_second": 94.082,
+      "eval_steps_per_second": 0.922,
       "step": 66
     },
     {
       "epoch": 11.67,
-      "learning_rate": 1.7666666666666668e-05,
-      "loss": 2.7201,
+      "learning_rate": 1.8055555555555558e-05,
+      "loss": 2.7149,
       "step": 70
     },
     {
       "epoch": 12.0,
       "eval_accuracy": 0.5196078431372549,
-      "eval_loss": 2.7339251041412354,
-      "eval_runtime": 0.6836,
-      "eval_samples_per_second": 149.204,
-      "eval_steps_per_second": 1.463,
+      "eval_loss": 2.7278239727020264,
+      "eval_runtime": 0.6684,
+      "eval_samples_per_second": 152.601,
+      "eval_steps_per_second": 1.496,
       "step": 72
     },
     {
       "epoch": 13.0,
       "eval_accuracy": 0.5588235294117647,
-      "eval_loss": 2.68558669090271,
-      "eval_runtime": 0.7097,
-      "eval_samples_per_second": 143.73,
-      "eval_steps_per_second": 1.409,
+      "eval_loss": 2.679572343826294,
+      "eval_runtime": 0.6624,
+      "eval_samples_per_second": 153.989,
+      "eval_steps_per_second": 1.51,
       "step": 78
     },
     {
       "epoch": 13.33,
-      "learning_rate": 1.7333333333333336e-05,
-      "loss": 2.6271,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 2.6202,
       "step": 80
     },
     {
       "epoch": 14.0,
       "eval_accuracy": 0.5882352941176471,
-      "eval_loss": 2.628051519393921,
-      "eval_runtime": 0.7248,
-      "eval_samples_per_second": 140.727,
-      "eval_steps_per_second": 1.38,
+      "eval_loss": 2.6203176975250244,
+      "eval_runtime": 0.7886,
+      "eval_samples_per_second": 129.335,
+      "eval_steps_per_second": 1.268,
       "step": 84
     },
     {
       "epoch": 15.0,
-      "learning_rate": 1.7e-05,
-      "loss": 2.5333,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 2.5243,
       "step": 90
     },
     {
       "epoch": 15.0,
       "eval_accuracy": 0.5882352941176471,
-      "eval_loss": 2.576406717300415,
-      "eval_runtime": 0.8135,
-      "eval_samples_per_second": 125.392,
-      "eval_steps_per_second": 1.229,
+      "eval_loss": 2.5673654079437256,
+      "eval_runtime": 0.8258,
+      "eval_samples_per_second": 123.523,
+      "eval_steps_per_second": 1.211,
       "step": 90
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.5980392156862745,
-      "eval_loss": 2.527658462524414,
-      "eval_runtime": 0.7155,
-      "eval_samples_per_second": 142.56,
-      "eval_steps_per_second": 1.398,
+      "eval_accuracy": 0.6078431372549019,
+      "eval_loss": 2.5169873237609863,
+      "eval_runtime": 0.8815,
+      "eval_samples_per_second": 115.707,
+      "eval_steps_per_second": 1.134,
       "step": 96
     },
     {
       "epoch": 16.67,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 2.4387,
+      "learning_rate": 1.7222222222222224e-05,
+      "loss": 2.4279,
       "step": 100
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.5980392156862745,
-      "eval_loss": 2.479215621948242,
-      "eval_runtime": 0.83,
-      "eval_samples_per_second": 122.895,
-      "eval_steps_per_second": 1.205,
+      "eval_accuracy": 0.6176470588235294,
+      "eval_loss": 2.4672491550445557,
+      "eval_runtime": 0.728,
+      "eval_samples_per_second": 140.105,
+      "eval_steps_per_second": 1.374,
       "step": 102
     },
     {
       "epoch": 18.0,
       "eval_accuracy": 0.5980392156862745,
-      "eval_loss": 2.441930055618286,
-      "eval_runtime": 0.9535,
-      "eval_samples_per_second": 106.969,
-      "eval_steps_per_second": 1.049,
+      "eval_loss": 2.428467035293579,
+      "eval_runtime": 1.0034,
+      "eval_samples_per_second": 101.658,
+      "eval_steps_per_second": 0.997,
       "step": 108
     },
     {
       "epoch": 18.33,
-      "learning_rate": 1.6333333333333335e-05,
-      "loss": 2.3536,
+      "learning_rate": 1.6944444444444446e-05,
+      "loss": 2.3404,
       "step": 110
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.6470588235294118,
-      "eval_loss": 2.3931069374084473,
-      "eval_runtime": 0.8198,
-      "eval_samples_per_second": 124.415,
-      "eval_steps_per_second": 1.22,
+      "eval_accuracy": 0.6568627450980392,
+      "eval_loss": 2.3784313201904297,
+      "eval_runtime": 0.9761,
+      "eval_samples_per_second": 104.5,
+      "eval_steps_per_second": 1.025,
       "step": 114
     },
     {
       "epoch": 20.0,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 2.279,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 2.2633,
       "step": 120
     },
     {
       "epoch": 20.0,
       "eval_accuracy": 0.6470588235294118,
-      "eval_loss": 2.3502094745635986,
-      "eval_runtime": 0.8889,
-      "eval_samples_per_second": 114.754,
-      "eval_steps_per_second": 1.125,
+      "eval_loss": 2.3347628116607666,
+      "eval_runtime": 1.0063,
+      "eval_samples_per_second": 101.36,
+      "eval_steps_per_second": 0.994,
       "step": 120
     },
     {
       "epoch": 21.0,
       "eval_accuracy": 0.6666666666666666,
-      "eval_loss": 2.3048012256622314,
-      "eval_runtime": 0.881,
-      "eval_samples_per_second": 115.784,
-      "eval_steps_per_second": 1.135,
+      "eval_loss": 2.2872164249420166,
+      "eval_runtime": 0.9826,
+      "eval_samples_per_second": 103.811,
+      "eval_steps_per_second": 1.018,
       "step": 126
     },
     {
       "epoch": 21.67,
-      "learning_rate": 1.5666666666666667e-05,
-      "loss": 2.2026,
+      "learning_rate": 1.638888888888889e-05,
+      "loss": 2.1838,
       "step": 130
     },
     {
       "epoch": 22.0,
       "eval_accuracy": 0.6568627450980392,
-      "eval_loss": 2.272130250930786,
-      "eval_runtime": 0.8015,
-      "eval_samples_per_second": 127.258,
-      "eval_steps_per_second": 1.248,
+      "eval_loss": 2.2538723945617676,
+      "eval_runtime": 0.7468,
+      "eval_samples_per_second": 136.583,
+      "eval_steps_per_second": 1.339,
       "step": 132
     },
     {
       "epoch": 23.0,
       "eval_accuracy": 0.6764705882352942,
-      "eval_loss": 2.242011308670044,
-      "eval_runtime": 1.0538,
-      "eval_samples_per_second": 96.792,
-      "eval_steps_per_second": 0.949,
+      "eval_loss": 2.2231638431549072,
+      "eval_runtime": 0.6591,
+      "eval_samples_per_second": 154.75,
+      "eval_steps_per_second": 1.517,
       "step": 138
     },
     {
       "epoch": 23.33,
-      "learning_rate": 1.5333333333333334e-05,
-      "loss": 2.1233,
+      "learning_rate": 1.6111111111111115e-05,
+      "loss": 2.1022,
       "step": 140
     },
     {
       "epoch": 24.0,
       "eval_accuracy": 0.6470588235294118,
-      "eval_loss": 2.20729923248291,
-      "eval_runtime": 1.6437,
-      "eval_samples_per_second": 62.057,
-      "eval_steps_per_second": 0.608,
+      "eval_loss": 2.186713218688965,
+      "eval_runtime": 0.668,
+      "eval_samples_per_second": 152.702,
+      "eval_steps_per_second": 1.497,
       "step": 144
     },
     {
       "epoch": 25.0,
-      "learning_rate": 1.5000000000000002e-05,
-      "loss": 2.0591,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 2.0364,
       "step": 150
     },
     {
       "epoch": 25.0,
-      "eval_accuracy": 0.696078431372549,
-      "eval_loss": 2.1703522205352783,
-      "eval_runtime": 1.0255,
-      "eval_samples_per_second": 99.464,
-      "eval_steps_per_second": 0.975,
+      "eval_accuracy": 0.6862745098039216,
+      "eval_loss": 2.1489007472991943,
+      "eval_runtime": 0.6659,
+      "eval_samples_per_second": 153.174,
+      "eval_steps_per_second": 1.502,
       "step": 150
     },
     {
       "epoch": 26.0,
-      "eval_accuracy": 0.7156862745098039,
-      "eval_loss": 2.132262706756592,
-      "eval_runtime": 0.8947,
-      "eval_samples_per_second": 114.002,
-      "eval_steps_per_second": 1.118,
+      "eval_accuracy": 0.7254901960784313,
+      "eval_loss": 2.109919309616089,
+      "eval_runtime": 0.9945,
+      "eval_samples_per_second": 102.565,
+      "eval_steps_per_second": 1.006,
       "step": 156
     },
     {
       "epoch": 26.67,
-      "learning_rate": 1.4666666666666666e-05,
-      "loss": 1.9858,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 1.96,
       "step": 160
     },
     {
       "epoch": 27.0,
-      "eval_accuracy": 0.6862745098039216,
-      "eval_loss": 2.101270914077759,
-      "eval_runtime": 1.1014,
-      "eval_samples_per_second": 92.612,
-      "eval_steps_per_second": 0.908,
+      "eval_accuracy": 0.7156862745098039,
+      "eval_loss": 2.0766732692718506,
+      "eval_runtime": 1.474,
+      "eval_samples_per_second": 69.198,
+      "eval_steps_per_second": 0.678,
       "step": 162
     },
     {
       "epoch": 28.0,
       "eval_accuracy": 0.7156862745098039,
-      "eval_loss": 2.0665531158447266,
-      "eval_runtime": 0.7467,
-      "eval_samples_per_second": 136.598,
-      "eval_steps_per_second": 1.339,
+      "eval_loss": 2.0417261123657227,
+      "eval_runtime": 0.9315,
+      "eval_samples_per_second": 109.504,
+      "eval_steps_per_second": 1.074,
       "step": 168
     },
     {
       "epoch": 28.33,
-      "learning_rate": 1.4333333333333334e-05,
-      "loss": 1.9518,
+      "learning_rate": 1.5277777777777777e-05,
+      "loss": 1.9235,
       "step": 170
     },
     {
       "epoch": 29.0,
       "eval_accuracy": 0.7352941176470589,
-      "eval_loss": 2.0423755645751953,
-      "eval_runtime": 0.8412,
-      "eval_samples_per_second": 121.25,
-      "eval_steps_per_second": 1.189,
+      "eval_loss": 2.0162079334259033,
+      "eval_runtime": 1.6546,
+      "eval_samples_per_second": 61.648,
+      "eval_steps_per_second": 0.604,
       "step": 174
     },
     {
       "epoch": 30.0,
-      "learning_rate": 1.4e-05,
-      "loss": 1.8789,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.8484,
       "step": 180
     },
     {
       "epoch": 30.0,
-      "eval_accuracy": 0.7352941176470589,
-      "eval_loss": 2.0062215328216553,
-      "eval_runtime": 0.7426,
-      "eval_samples_per_second": 137.363,
-      "eval_steps_per_second": 1.347,
+      "eval_accuracy": 0.7450980392156863,
+      "eval_loss": 1.9787482023239136,
+      "eval_runtime": 0.9212,
+      "eval_samples_per_second": 110.726,
+      "eval_steps_per_second": 1.086,
       "step": 180
     },
     {
       "epoch": 31.0,
-      "eval_accuracy": 0.7352941176470589,
-      "eval_loss": 1.9826091527938843,
-      "eval_runtime": 1.5779,
-      "eval_samples_per_second": 64.643,
-      "eval_steps_per_second": 0.634,
+      "eval_accuracy": 0.7450980392156863,
+      "eval_loss": 1.9547613859176636,
+      "eval_runtime": 0.7852,
+      "eval_samples_per_second": 129.907,
+      "eval_steps_per_second": 1.274,
       "step": 186
     },
     {
       "epoch": 31.67,
-      "learning_rate": 1.3666666666666667e-05,
-      "loss": 1.8307,
+      "learning_rate": 1.4722222222222224e-05,
+      "loss": 1.7971,
       "step": 190
     },
     {
       "epoch": 32.0,
-      "eval_accuracy": 0.7647058823529411,
-      "eval_loss": 1.9618277549743652,
-      "eval_runtime": 0.848,
-      "eval_samples_per_second": 120.283,
-      "eval_steps_per_second": 1.179,
+      "eval_accuracy": 0.7549019607843137,
+      "eval_loss": 1.9329289197921753,
+      "eval_runtime": 1.6855,
+      "eval_samples_per_second": 60.517,
+      "eval_steps_per_second": 0.593,
       "step": 192
     },
     {
       "epoch": 33.0,
       "eval_accuracy": 0.7647058823529411,
-      "eval_loss": 1.9387190341949463,
-      "eval_runtime": 0.8972,
-      "eval_samples_per_second": 113.683,
-      "eval_steps_per_second": 1.115,
+      "eval_loss": 1.9051672220230103,
+      "eval_runtime": 0.999,
+      "eval_samples_per_second": 102.103,
+      "eval_steps_per_second": 1.001,
       "step": 198
     },
     {
       "epoch": 33.33,
-      "learning_rate": 1.3333333333333333e-05,
-      "loss": 1.7762,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 1.7409,
       "step": 200
     },
     {
       "epoch": 34.0,
       "eval_accuracy": 0.7549019607843137,
-      "eval_loss": 1.9141510725021362,
-      "eval_runtime": 0.7852,
-      "eval_samples_per_second": 129.898,
-      "eval_steps_per_second": 1.274,
+      "eval_loss": 1.8827149868011475,
+      "eval_runtime": 0.6689,
+      "eval_samples_per_second": 152.484,
+      "eval_steps_per_second": 1.495,
       "step": 204
     },
     {
       "epoch": 35.0,
-      "learning_rate": 1.3000000000000001e-05,
-      "loss": 1.7383,
+      "learning_rate": 1.416666666666667e-05,
+      "loss": 1.7006,
       "step": 210
     },
     {
       "epoch": 35.0,
-      "eval_accuracy": 0.7647058823529411,
-      "eval_loss": 1.8922001123428345,
-      "eval_runtime": 0.6999,
-      "eval_samples_per_second": 145.726,
-      "eval_steps_per_second": 1.429,
+      "eval_accuracy": 0.7745098039215687,
+      "eval_loss": 1.8589290380477905,
+      "eval_runtime": 0.8033,
+      "eval_samples_per_second": 126.972,
+      "eval_steps_per_second": 1.245,
       "step": 210
     },
     {
       "epoch": 36.0,
-      "eval_accuracy": 0.7745098039215687,
-      "eval_loss": 1.864749550819397,
-      "eval_runtime": 0.68,
-      "eval_samples_per_second": 149.999,
-      "eval_steps_per_second": 1.471,
+      "eval_accuracy": 0.7843137254901961,
+      "eval_loss": 1.8293644189834595,
+      "eval_runtime": 1.6485,
+      "eval_samples_per_second": 61.875,
+      "eval_steps_per_second": 0.607,
       "step": 216
     },
     {
       "epoch": 36.67,
-      "learning_rate": 1.2666666666666667e-05,
-      "loss": 1.683,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 1.6426,
       "step": 220
     },
     {
       "epoch": 37.0,
-      "eval_accuracy": 0.7745098039215687,
-      "eval_loss": 1.8461244106292725,
-      "eval_runtime": 0.8708,
-      "eval_samples_per_second": 117.132,
-      "eval_steps_per_second": 1.148,
+      "eval_accuracy": 0.7843137254901961,
+      "eval_loss": 1.8097856044769287,
+      "eval_runtime": 1.5961,
+      "eval_samples_per_second": 63.905,
+      "eval_steps_per_second": 0.627,
       "step": 222
     },
     {
       "epoch": 38.0,
-      "eval_accuracy": 0.7549019607843137,
-      "eval_loss": 1.8185441493988037,
-      "eval_runtime": 0.7505,
-      "eval_samples_per_second": 135.905,
-      "eval_steps_per_second": 1.332,
+      "eval_accuracy": 0.7647058823529411,
+      "eval_loss": 1.7808825969696045,
+      "eval_runtime": 0.7953,
+      "eval_samples_per_second": 128.247,
+      "eval_steps_per_second": 1.257,
       "step": 228
     },
     {
       "epoch": 38.33,
-      "learning_rate": 1.2333333333333334e-05,
-      "loss": 1.6527,
+      "learning_rate": 1.3611111111111113e-05,
+      "loss": 1.6102,
       "step": 230
     },
     {
       "epoch": 39.0,
       "eval_accuracy": 0.7843137254901961,
-      "eval_loss": 1.801927924156189,
-      "eval_runtime": 1.141,
-      "eval_samples_per_second": 89.397,
-      "eval_steps_per_second": 0.876,
+      "eval_loss": 1.764255404472351,
+      "eval_runtime": 0.9811,
+      "eval_samples_per_second": 103.969,
+      "eval_steps_per_second": 1.019,
       "step": 234
     },
     {
       "epoch": 40.0,
-      "learning_rate": 1.2e-05,
-      "loss": 1.615,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.5704,
       "step": 240
     },
     {
       "epoch": 40.0,
-      "eval_accuracy": 0.7941176470588235,
-      "eval_loss": 1.7803996801376343,
-      "eval_runtime": 0.8026,
-      "eval_samples_per_second": 127.081,
-      "eval_steps_per_second": 1.246,
+      "eval_accuracy": 0.803921568627451,
+      "eval_loss": 1.7398592233657837,
+      "eval_runtime": 0.9069,
+      "eval_samples_per_second": 112.465,
+      "eval_steps_per_second": 1.103,
       "step": 240
     },
     {
       "epoch": 41.0,
-      "eval_accuracy": 0.7843137254901961,
-      "eval_loss": 1.7603331804275513,
-      "eval_runtime": 0.7405,
-      "eval_samples_per_second": 137.746,
-      "eval_steps_per_second": 1.35,
+      "eval_accuracy": 0.8137254901960784,
+      "eval_loss": 1.7192758321762085,
+      "eval_runtime": 0.9941,
+      "eval_samples_per_second": 102.601,
+      "eval_steps_per_second": 1.006,
       "step": 246
     },
     {
       "epoch": 41.67,
-      "learning_rate": 1.1666666666666668e-05,
-      "loss": 1.5736,
+      "learning_rate": 1.3055555555555557e-05,
+      "loss": 1.5264,
       "step": 250
     },
     {
       "epoch": 42.0,
-      "eval_accuracy": 0.8137254901960784,
-      "eval_loss": 1.740982174873352,
-      "eval_runtime": 0.6886,
-      "eval_samples_per_second": 148.121,
-      "eval_steps_per_second": 1.452,
+      "eval_accuracy": 0.8333333333333334,
+      "eval_loss": 1.6979612112045288,
+      "eval_runtime": 0.8321,
+      "eval_samples_per_second": 122.58,
+      "eval_steps_per_second": 1.202,
       "step": 252
     },
     {
       "epoch": 43.0,
-      "eval_accuracy": 0.7941176470588235,
-      "eval_loss": 1.7274985313415527,
-      "eval_runtime": 0.6877,
-      "eval_samples_per_second": 148.329,
-      "eval_steps_per_second": 1.454,
+      "eval_accuracy": 0.803921568627451,
+      "eval_loss": 1.6840269565582275,
+      "eval_runtime": 0.8256,
+      "eval_samples_per_second": 123.549,
+      "eval_steps_per_second": 1.211,
       "step": 258
     },
     {
       "epoch": 43.33,
-      "learning_rate": 1.1333333333333334e-05,
-      "loss": 1.5316,
+      "learning_rate": 1.2777777777777777e-05,
+      "loss": 1.4821,
       "step": 260
     },
     {
       "epoch": 44.0,
       "eval_accuracy": 0.8235294117647058,
-      "eval_loss": 1.7081972360610962,
-      "eval_runtime": 0.7845,
-      "eval_samples_per_second": 130.014,
-      "eval_steps_per_second": 1.275,
+      "eval_loss": 1.6644328832626343,
+      "eval_runtime": 0.8752,
+      "eval_samples_per_second": 116.547,
+      "eval_steps_per_second": 1.143,
       "step": 264
     },
     {
       "epoch": 45.0,
-      "learning_rate": 1.1000000000000001e-05,
-      "loss": 1.503,
+      "learning_rate": 1.25e-05,
+      "loss": 1.4506,
       "step": 270
     },
     {
       "epoch": 45.0,
-      "eval_accuracy": 0.8137254901960784,
-      "eval_loss": 1.6913036108016968,
-      "eval_runtime": 0.8968,
-      "eval_samples_per_second": 113.735,
-      "eval_steps_per_second": 1.115,
+      "eval_accuracy": 0.8235294117647058,
+      "eval_loss": 1.646732211112976,
+      "eval_runtime": 0.8249,
+      "eval_samples_per_second": 123.658,
+      "eval_steps_per_second": 1.212,
       "step": 270
     },
     {
       "epoch": 46.0,
-      "eval_accuracy": 0.803921568627451,
-      "eval_loss": 1.680880069732666,
-      "eval_runtime": 0.776,
-      "eval_samples_per_second": 131.44,
-      "eval_steps_per_second": 1.289,
+      "eval_accuracy": 0.8235294117647058,
+      "eval_loss": 1.6333067417144775,
+      "eval_runtime": 0.7966,
+      "eval_samples_per_second": 128.049,
+      "eval_steps_per_second": 1.255,
       "step": 276
     },
     {
       "epoch": 46.67,
-      "learning_rate": 1.0666666666666667e-05,
-      "loss": 1.4905,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 1.4358,
       "step": 280
     },
     {
       "epoch": 47.0,
-      "eval_accuracy": 0.803921568627451,
-      "eval_loss": 1.6573076248168945,
-      "eval_runtime": 0.6724,
-      "eval_samples_per_second": 151.693,
-      "eval_steps_per_second": 1.487,
+      "eval_accuracy": 0.8235294117647058,
+      "eval_loss": 1.6094907522201538,
+      "eval_runtime": 1.3349,
+      "eval_samples_per_second": 76.412,
+      "eval_steps_per_second": 0.749,
       "step": 282
     },
     {
       "epoch": 48.0,
       "eval_accuracy": 0.8235294117647058,
-      "eval_loss": 1.6394025087356567,
-      "eval_runtime": 0.69,
-      "eval_samples_per_second": 147.83,
-      "eval_steps_per_second": 1.449,
+      "eval_loss": 1.5905718803405762,
+      "eval_runtime": 0.8441,
+      "eval_samples_per_second": 120.832,
+      "eval_steps_per_second": 1.185,
       "step": 288
     },
     {
       "epoch": 48.33,
-      "learning_rate": 1.0333333333333335e-05,
-      "loss": 1.4262,
+      "learning_rate": 1.1944444444444444e-05,
+      "loss": 1.3695,
       "step": 290
     },
     {
       "epoch": 49.0,
-      "eval_accuracy": 0.8137254901960784,
-      "eval_loss": 1.6250027418136597,
-      "eval_runtime": 0.7655,
-      "eval_samples_per_second": 133.247,
-      "eval_steps_per_second": 1.306,
+      "eval_accuracy": 0.8431372549019608,
+      "eval_loss": 1.571975827217102,
+      "eval_runtime": 0.8382,
+      "eval_samples_per_second": 121.692,
+      "eval_steps_per_second": 1.193,
       "step": 294
     },
     {
       "epoch": 50.0,
-      "learning_rate": 1e-05,
-      "loss": 1.426,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 1.367,
       "step": 300
     },
     {
       "epoch": 50.0,
       "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 1.6149309873580933,
-      "eval_runtime": 0.7535,
-      "eval_samples_per_second": 135.375,
-      "eval_steps_per_second": 1.327,
+      "eval_loss": 1.5610417127609253,
+      "eval_runtime": 0.7505,
+      "eval_samples_per_second": 135.913,
+      "eval_steps_per_second": 1.332,
       "step": 300
     },
     {
       "epoch": 51.0,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 1.6013031005859375,
-      "eval_runtime": 0.7915,
-      "eval_samples_per_second": 128.866,
-      "eval_steps_per_second": 1.263,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.5439658164978027,
+      "eval_runtime": 1.0176,
+      "eval_samples_per_second": 100.232,
+      "eval_steps_per_second": 0.983,
       "step": 306
     },
     {
       "epoch": 51.67,
-      "learning_rate": 9.666666666666667e-06,
-      "loss": 1.3915,
+      "learning_rate": 1.138888888888889e-05,
+      "loss": 1.3299,
       "step": 310
     },
     {
       "epoch": 52.0,
-      "eval_accuracy": 0.8235294117647058,
-      "eval_loss": 1.5923330783843994,
-      "eval_runtime": 0.7278,
-      "eval_samples_per_second": 140.152,
-      "eval_steps_per_second": 1.374,
+      "eval_accuracy": 0.8333333333333334,
+      "eval_loss": 1.5359292030334473,
+      "eval_runtime": 0.7662,
+      "eval_samples_per_second": 133.117,
+      "eval_steps_per_second": 1.305,
       "step": 312
     },
     {
       "epoch": 53.0,
       "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 1.5717215538024902,
-      "eval_runtime": 0.6608,
-      "eval_samples_per_second": 154.364,
-      "eval_steps_per_second": 1.513,
+      "eval_loss": 1.5128930807113647,
+      "eval_runtime": 0.7573,
+      "eval_samples_per_second": 134.684,
+      "eval_steps_per_second": 1.32,
       "step": 318
     },
     {
       "epoch": 53.33,
-      "learning_rate": 9.333333333333334e-06,
-      "loss": 1.3406,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 1.2765,
       "step": 320
     },
     {
       "epoch": 54.0,
       "eval_accuracy": 0.8235294117647058,
-      "eval_loss": 1.5661754608154297,
-      "eval_runtime": 0.8648,
-      "eval_samples_per_second": 117.94,
-      "eval_steps_per_second": 1.156,
+      "eval_loss": 1.5056854486465454,
+      "eval_runtime": 0.8808,
+      "eval_samples_per_second": 115.804,
+      "eval_steps_per_second": 1.135,
       "step": 324
     },
     {
       "epoch": 55.0,
-      "learning_rate": 9e-06,
-      "loss": 1.3444,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 1.2785,
       "step": 330
     },
     {
       "epoch": 55.0,
       "eval_accuracy": 0.8235294117647058,
-      "eval_loss": 1.5482959747314453,
-      "eval_runtime": 0.7933,
-      "eval_samples_per_second": 128.58,
-      "eval_steps_per_second": 1.261,
+      "eval_loss": 1.4866654872894287,
+      "eval_runtime": 0.7751,
+      "eval_samples_per_second": 131.599,
+      "eval_steps_per_second": 1.29,
       "step": 330
     },
     {
       "epoch": 56.0,
-      "eval_accuracy": 0.8235294117647058,
-      "eval_loss": 1.5372467041015625,
-      "eval_runtime": 0.729,
-      "eval_samples_per_second": 139.912,
-      "eval_steps_per_second": 1.372,
+      "eval_accuracy": 0.8333333333333334,
+      "eval_loss": 1.475063443183899,
+      "eval_runtime": 0.7201,
+      "eval_samples_per_second": 141.644,
+      "eval_steps_per_second": 1.389,
       "step": 336
     },
     {
       "epoch": 56.67,
-      "learning_rate": 8.666666666666668e-06,
-      "loss": 1.3053,
+      "learning_rate": 1.0555555555555557e-05,
+      "loss": 1.2355,
       "step": 340
     },
     {
       "epoch": 57.0,
-      "eval_accuracy": 0.8137254901960784,
-      "eval_loss": 1.522821068763733,
-      "eval_runtime": 0.67,
-      "eval_samples_per_second": 152.241,
-      "eval_steps_per_second": 1.493,
+      "eval_accuracy": 0.8235294117647058,
+      "eval_loss": 1.4553169012069702,
+      "eval_runtime": 0.7072,
+      "eval_samples_per_second": 144.23,
+      "eval_steps_per_second": 1.414,
       "step": 342
     },
     {
       "epoch": 58.0,
       "eval_accuracy": 0.8235294117647058,
-      "eval_loss": 1.513081431388855,
-      "eval_runtime": 0.8482,
-      "eval_samples_per_second": 120.26,
-      "eval_steps_per_second": 1.179,
+      "eval_loss": 1.4491091966629028,
+      "eval_runtime": 0.6583,
+      "eval_samples_per_second": 154.94,
+      "eval_steps_per_second": 1.519,
       "step": 348
     },
     {
       "epoch": 58.33,
-      "learning_rate": 8.333333333333334e-06,
-      "loss": 1.3147,
+      "learning_rate": 1.0277777777777777e-05,
+      "loss": 1.2418,
       "step": 350
     },
     {
       "epoch": 59.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.4987856149673462,
-      "eval_runtime": 0.6717,
-      "eval_samples_per_second": 151.851,
-      "eval_steps_per_second": 1.489,
+      "eval_loss": 1.428889513015747,
+      "eval_runtime": 0.6735,
+      "eval_samples_per_second": 151.449,
+      "eval_steps_per_second": 1.485,
       "step": 354
     },
     {
       "epoch": 60.0,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 1.2796,
+      "learning_rate": 1e-05,
+      "loss": 1.2058,
       "step": 360
     },
     {
       "epoch": 60.0,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 1.4865219593048096,
-      "eval_runtime": 0.7339,
-      "eval_samples_per_second": 138.992,
-      "eval_steps_per_second": 1.363,
+      "eval_accuracy": 0.8235294117647058,
+      "eval_loss": 1.4184516668319702,
+      "eval_runtime": 0.8642,
+      "eval_samples_per_second": 118.027,
+      "eval_steps_per_second": 1.157,
       "step": 360
     },
     {
       "epoch": 61.0,
       "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 1.4788631200790405,
-      "eval_runtime": 0.7971,
-      "eval_samples_per_second": 127.964,
-      "eval_steps_per_second": 1.255,
+      "eval_loss": 1.4103975296020508,
+      "eval_runtime": 0.6996,
+      "eval_samples_per_second": 145.797,
+      "eval_steps_per_second": 1.429,
       "step": 366
     },
     {
       "epoch": 61.67,
-      "learning_rate": 7.666666666666667e-06,
-      "loss": 1.2419,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 1.164,
       "step": 370
     },
     {
       "epoch": 62.0,
-      "eval_accuracy": 0.8137254901960784,
-      "eval_loss": 1.4684420824050903,
-      "eval_runtime": 0.6989,
-      "eval_samples_per_second": 145.949,
-      "eval_steps_per_second": 1.431,
+      "eval_accuracy": 0.8333333333333334,
+      "eval_loss": 1.396785855293274,
+      "eval_runtime": 0.7378,
+      "eval_samples_per_second": 138.245,
+      "eval_steps_per_second": 1.355,
       "step": 372
     },
     {
       "epoch": 63.0,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 1.459528923034668,
-      "eval_runtime": 0.6797,
-      "eval_samples_per_second": 150.063,
-      "eval_steps_per_second": 1.471,
+      "eval_accuracy": 0.8431372549019608,
+      "eval_loss": 1.3845832347869873,
+      "eval_runtime": 0.9059,
+      "eval_samples_per_second": 112.599,
+      "eval_steps_per_second": 1.104,
       "step": 378
     },
     {
       "epoch": 63.33,
-      "learning_rate": 7.333333333333333e-06,
-      "loss": 1.2326,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 1.1529,
       "step": 380
     },
     {
       "epoch": 64.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.447430968284607,
-      "eval_runtime": 0.6753,
-      "eval_samples_per_second": 151.054,
-      "eval_steps_per_second": 1.481,
+      "eval_loss": 1.3697338104248047,
+      "eval_runtime": 0.9585,
+      "eval_samples_per_second": 106.42,
+      "eval_steps_per_second": 1.043,
       "step": 384
     },
     {
       "epoch": 65.0,
-      "learning_rate": 7e-06,
-      "loss": 1.2246,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 1.1408,
       "step": 390
     },
     {
       "epoch": 65.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.4404107332229614,
-      "eval_runtime": 0.6713,
-      "eval_samples_per_second": 151.935,
-      "eval_steps_per_second": 1.49,
+      "eval_loss": 1.36329185962677,
+      "eval_runtime": 0.8799,
+      "eval_samples_per_second": 115.918,
+      "eval_steps_per_second": 1.136,
       "step": 390
     },
     {
       "epoch": 66.0,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 1.4313303232192993,
-      "eval_runtime": 0.6779,
-      "eval_samples_per_second": 150.471,
-      "eval_steps_per_second": 1.475,
+      "eval_accuracy": 0.8431372549019608,
+      "eval_loss": 1.3505079746246338,
+      "eval_runtime": 0.8666,
+      "eval_samples_per_second": 117.702,
+      "eval_steps_per_second": 1.154,
       "step": 396
     },
     {
       "epoch": 66.67,
-      "learning_rate": 6.666666666666667e-06,
-      "loss": 1.1975,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.1102,
       "step": 400
     },
     {
       "epoch": 67.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.4185569286346436,
-      "eval_runtime": 0.8632,
-      "eval_samples_per_second": 118.16,
-      "eval_steps_per_second": 1.158,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.3371140956878662,
+      "eval_runtime": 0.6813,
+      "eval_samples_per_second": 149.711,
+      "eval_steps_per_second": 1.468,
       "step": 402
     },
     {
       "epoch": 68.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.4122827053070068,
-      "eval_runtime": 0.6792,
-      "eval_samples_per_second": 150.175,
-      "eval_steps_per_second": 1.472,
+      "eval_loss": 1.328187108039856,
+      "eval_runtime": 0.9696,
+      "eval_samples_per_second": 105.199,
+      "eval_steps_per_second": 1.031,
       "step": 408
     },
     {
       "epoch": 68.33,
-      "learning_rate": 6.333333333333333e-06,
-      "loss": 1.1804,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.0906,
       "step": 410
     },
     {
       "epoch": 69.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.4083011150360107,
-      "eval_runtime": 0.9425,
-      "eval_samples_per_second": 108.219,
-      "eval_steps_per_second": 1.061,
+      "eval_loss": 1.3239920139312744,
+      "eval_runtime": 0.9604,
+      "eval_samples_per_second": 106.207,
+      "eval_steps_per_second": 1.041,
       "step": 414
     },
     {
       "epoch": 70.0,
-      "learning_rate": 6e-06,
-      "loss": 1.1686,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.0759,
       "step": 420
     },
     {
       "epoch": 70.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.4023797512054443,
-      "eval_runtime": 0.6859,
-      "eval_samples_per_second": 148.71,
-      "eval_steps_per_second": 1.458,
+      "eval_loss": 1.3162660598754883,
+      "eval_runtime": 0.7343,
+      "eval_samples_per_second": 138.913,
+      "eval_steps_per_second": 1.362,
       "step": 420
     },
     {
       "epoch": 71.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.395778775215149,
-      "eval_runtime": 0.7142,
-      "eval_samples_per_second": 142.812,
-      "eval_steps_per_second": 1.4,
+      "eval_loss": 1.3044177293777466,
+      "eval_runtime": 0.9811,
+      "eval_samples_per_second": 103.966,
+      "eval_steps_per_second": 1.019,
       "step": 426
     },
     {
       "epoch": 71.67,
-      "learning_rate": 5.666666666666667e-06,
-      "loss": 1.1608,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 1.0651,
       "step": 430
     },
     {
       "epoch": 72.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.384925365447998,
-      "eval_runtime": 0.792,
-      "eval_samples_per_second": 128.785,
-      "eval_steps_per_second": 1.263,
+      "eval_loss": 1.2924034595489502,
+      "eval_runtime": 0.9153,
+      "eval_samples_per_second": 111.445,
+      "eval_steps_per_second": 1.093,
       "step": 432
     },
     {
       "epoch": 73.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.3785865306854248,
-      "eval_runtime": 0.7537,
-      "eval_samples_per_second": 135.336,
-      "eval_steps_per_second": 1.327,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.2867342233657837,
+      "eval_runtime": 0.9693,
+      "eval_samples_per_second": 105.231,
+      "eval_steps_per_second": 1.032,
       "step": 438
     },
     {
       "epoch": 73.33,
-      "learning_rate": 5.333333333333334e-06,
-      "loss": 1.1491,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.0501,
       "step": 440
     },
     {
       "epoch": 74.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.3704206943511963,
-      "eval_runtime": 0.756,
-      "eval_samples_per_second": 134.922,
-      "eval_steps_per_second": 1.323,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.2749128341674805,
+      "eval_runtime": 1.1764,
+      "eval_samples_per_second": 86.703,
+      "eval_steps_per_second": 0.85,
       "step": 444
     },
     {
       "epoch": 75.0,
-      "learning_rate": 5e-06,
-      "loss": 1.1283,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.0238,
       "step": 450
     },
     {
       "epoch": 75.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.3653908967971802,
-      "eval_runtime": 0.7484,
-      "eval_samples_per_second": 136.284,
-      "eval_steps_per_second": 1.336,
+      "eval_loss": 1.2687901258468628,
+      "eval_runtime": 0.6685,
+      "eval_samples_per_second": 152.571,
+      "eval_steps_per_second": 1.496,
       "step": 450
     },
     {
       "epoch": 76.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3577083349227905,
-      "eval_runtime": 0.7726,
-      "eval_samples_per_second": 132.025,
-      "eval_steps_per_second": 1.294,
+      "eval_loss": 1.2567930221557617,
+      "eval_runtime": 0.7274,
+      "eval_samples_per_second": 140.221,
+      "eval_steps_per_second": 1.375,
       "step": 456
     },
     {
       "epoch": 76.67,
-      "learning_rate": 4.666666666666667e-06,
-      "loss": 1.1113,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 1.0046,
       "step": 460
     },
     {
       "epoch": 77.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.351647973060608,
-      "eval_runtime": 0.7308,
-      "eval_samples_per_second": 139.58,
-      "eval_steps_per_second": 1.368,
+      "eval_loss": 1.2502152919769287,
+      "eval_runtime": 0.7589,
+      "eval_samples_per_second": 134.406,
+      "eval_steps_per_second": 1.318,
       "step": 462
     },
     {
       "epoch": 78.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3462965488433838,
-      "eval_runtime": 0.8148,
-      "eval_samples_per_second": 125.184,
-      "eval_steps_per_second": 1.227,
+      "eval_loss": 1.2459862232208252,
+      "eval_runtime": 1.3626,
+      "eval_samples_per_second": 74.857,
+      "eval_steps_per_second": 0.734,
       "step": 468
     },
     {
       "epoch": 78.33,
-      "learning_rate": 4.333333333333334e-06,
-      "loss": 1.1047,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.9946,
       "step": 470
     },
     {
       "epoch": 79.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.3450775146484375,
-      "eval_runtime": 0.6922,
-      "eval_samples_per_second": 147.359,
-      "eval_steps_per_second": 1.445,
+      "eval_loss": 1.2455214262008667,
+      "eval_runtime": 0.7802,
+      "eval_samples_per_second": 130.729,
+      "eval_steps_per_second": 1.282,
       "step": 474
     },
     {
       "epoch": 80.0,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 1.1132,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.9998,
       "step": 480
     },
     {
       "epoch": 80.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3408479690551758,
-      "eval_runtime": 0.6779,
-      "eval_samples_per_second": 150.459,
-      "eval_steps_per_second": 1.475,
+      "eval_loss": 1.2342967987060547,
+      "eval_runtime": 0.7842,
+      "eval_samples_per_second": 130.063,
+      "eval_steps_per_second": 1.275,
       "step": 480
     },
     {
       "epoch": 81.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.336086630821228,
-      "eval_runtime": 0.9551,
-      "eval_samples_per_second": 106.8,
-      "eval_steps_per_second": 1.047,
+      "eval_loss": 1.2286028861999512,
+      "eval_runtime": 0.6729,
+      "eval_samples_per_second": 151.583,
+      "eval_steps_per_second": 1.486,
       "step": 486
     },
     {
       "epoch": 81.67,
-      "learning_rate": 3.6666666666666666e-06,
-      "loss": 1.0869,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 0.9709,
       "step": 490
     },
     {
       "epoch": 82.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.3308581113815308,
-      "eval_runtime": 0.7685,
-      "eval_samples_per_second": 132.729,
-      "eval_steps_per_second": 1.301,
+      "eval_loss": 1.2194892168045044,
+      "eval_runtime": 0.8526,
+      "eval_samples_per_second": 119.634,
+      "eval_steps_per_second": 1.173,
       "step": 492
     },
     {
       "epoch": 83.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.324817419052124,
-      "eval_runtime": 0.6895,
-      "eval_samples_per_second": 147.938,
-      "eval_steps_per_second": 1.45,
+      "eval_loss": 1.2126073837280273,
+      "eval_runtime": 0.9257,
+      "eval_samples_per_second": 110.192,
+      "eval_steps_per_second": 1.08,
       "step": 498
     },
     {
       "epoch": 83.33,
-      "learning_rate": 3.3333333333333333e-06,
-      "loss": 1.0818,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.963,
       "step": 500
     },
     {
       "epoch": 84.0,
-      "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.320926547050476,
-      "eval_runtime": 0.6927,
-      "eval_samples_per_second": 147.247,
-      "eval_steps_per_second": 1.444,
+      "eval_accuracy": 0.8431372549019608,
+      "eval_loss": 1.2101882696151733,
+      "eval_runtime": 1.5366,
+      "eval_samples_per_second": 66.38,
+      "eval_steps_per_second": 0.651,
       "step": 504
     },
     {
       "epoch": 85.0,
-      "learning_rate": 3e-06,
-      "loss": 1.0732,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.9499,
       "step": 510
     },
     {
       "epoch": 85.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.317958950996399,
-      "eval_runtime": 0.7281,
-      "eval_samples_per_second": 140.089,
-      "eval_steps_per_second": 1.373,
+      "eval_loss": 1.2024163007736206,
+      "eval_runtime": 1.2053,
+      "eval_samples_per_second": 84.623,
+      "eval_steps_per_second": 0.83,
       "step": 510
     },
     {
       "epoch": 86.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.3156795501708984,
-      "eval_runtime": 0.8315,
-      "eval_samples_per_second": 122.671,
-      "eval_steps_per_second": 1.203,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.1980304718017578,
+      "eval_runtime": 0.6794,
+      "eval_samples_per_second": 150.138,
+      "eval_steps_per_second": 1.472,
       "step": 516
     },
     {
       "epoch": 86.67,
-      "learning_rate": 2.666666666666667e-06,
-      "loss": 1.0637,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.937,
       "step": 520
     },
     {
       "epoch": 87.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3119415044784546,
-      "eval_runtime": 0.7917,
-      "eval_samples_per_second": 128.838,
-      "eval_steps_per_second": 1.263,
+      "eval_loss": 1.1911908388137817,
+      "eval_runtime": 0.9503,
+      "eval_samples_per_second": 107.339,
+      "eval_steps_per_second": 1.052,
       "step": 522
     },
     {
       "epoch": 88.0,
-      "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3100982904434204,
-      "eval_runtime": 0.8514,
-      "eval_samples_per_second": 119.809,
-      "eval_steps_per_second": 1.175,
+      "eval_accuracy": 0.8431372549019608,
+      "eval_loss": 1.1882524490356445,
+      "eval_runtime": 0.6719,
+      "eval_samples_per_second": 151.805,
+      "eval_steps_per_second": 1.488,
       "step": 528
     },
     {
       "epoch": 88.33,
-      "learning_rate": 2.3333333333333336e-06,
-      "loss": 1.0703,
+      "learning_rate": 5.2777777777777785e-06,
+      "loss": 0.9389,
       "step": 530
     },
     {
       "epoch": 89.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3083486557006836,
-      "eval_runtime": 0.6809,
-      "eval_samples_per_second": 149.804,
-      "eval_steps_per_second": 1.469,
+      "eval_loss": 1.1845465898513794,
+      "eval_runtime": 1.3029,
+      "eval_samples_per_second": 78.288,
+      "eval_steps_per_second": 0.768,
       "step": 534
     },
     {
       "epoch": 90.0,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 1.055,
+      "learning_rate": 5e-06,
+      "loss": 0.9181,
       "step": 540
     },
     {
       "epoch": 90.0,
       "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3066046237945557,
-      "eval_runtime": 0.7259,
-      "eval_samples_per_second": 140.507,
-      "eval_steps_per_second": 1.378,
+      "eval_loss": 1.1810888051986694,
+      "eval_runtime": 1.3666,
+      "eval_samples_per_second": 74.637,
+      "eval_steps_per_second": 0.732,
       "step": 540
     },
     {
       "epoch": 91.0,
       "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.3046588897705078,
-      "eval_runtime": 0.6982,
-      "eval_samples_per_second": 146.097,
-      "eval_steps_per_second": 1.432,
+      "eval_loss": 1.177653193473816,
+      "eval_runtime": 0.7195,
+      "eval_samples_per_second": 141.761,
+      "eval_steps_per_second": 1.39,
       "step": 546
     },
     {
       "epoch": 91.67,
-      "learning_rate": 1.6666666666666667e-06,
-      "loss": 1.0632,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 0.9219,
       "step": 550
     },
     {
       "epoch": 92.0,
-      "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3031229972839355,
-      "eval_runtime": 0.6959,
-      "eval_samples_per_second": 146.574,
-      "eval_steps_per_second": 1.437,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1742655038833618,
+      "eval_runtime": 0.9915,
+      "eval_samples_per_second": 102.872,
+      "eval_steps_per_second": 1.009,
       "step": 552
     },
     {
       "epoch": 93.0,
-      "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.3012945652008057,
-      "eval_runtime": 0.6812,
-      "eval_samples_per_second": 149.74,
-      "eval_steps_per_second": 1.468,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1674777269363403,
+      "eval_runtime": 1.1959,
+      "eval_samples_per_second": 85.292,
+      "eval_steps_per_second": 0.836,
       "step": 558
     },
     {
       "epoch": 93.33,
-      "learning_rate": 1.3333333333333334e-06,
-      "loss": 1.0509,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.9067,
       "step": 560
     },
     {
       "epoch": 94.0,
-      "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.2976847887039185,
-      "eval_runtime": 0.6876,
-      "eval_samples_per_second": 148.34,
-      "eval_steps_per_second": 1.454,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.159796118736267,
+      "eval_runtime": 0.6745,
+      "eval_samples_per_second": 151.226,
+      "eval_steps_per_second": 1.483,
       "step": 564
     },
     {
       "epoch": 95.0,
-      "learning_rate": 1.0000000000000002e-06,
-      "loss": 1.0496,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.9009,
       "step": 570
     },
     {
       "epoch": 95.0,
-      "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 1.2957763671875,
-      "eval_runtime": 0.6884,
-      "eval_samples_per_second": 148.164,
-      "eval_steps_per_second": 1.453,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1600637435913086,
+      "eval_runtime": 0.9647,
+      "eval_samples_per_second": 105.735,
+      "eval_steps_per_second": 1.037,
       "step": 570
     },
     {
       "epoch": 96.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.2950189113616943,
-      "eval_runtime": 1.0112,
-      "eval_samples_per_second": 100.872,
-      "eval_steps_per_second": 0.989,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.156423807144165,
+      "eval_runtime": 0.7689,
+      "eval_samples_per_second": 132.666,
+      "eval_steps_per_second": 1.301,
       "step": 576
     },
     {
       "epoch": 96.67,
-      "learning_rate": 6.666666666666667e-07,
-      "loss": 1.047,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.8914,
       "step": 580
     },
     {
       "epoch": 97.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.2942228317260742,
-      "eval_runtime": 0.8701,
-      "eval_samples_per_second": 117.229,
-      "eval_steps_per_second": 1.149,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.150528073310852,
+      "eval_runtime": 0.6817,
+      "eval_samples_per_second": 149.634,
+      "eval_steps_per_second": 1.467,
       "step": 582
     },
     {
       "epoch": 98.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.2938711643218994,
-      "eval_runtime": 0.7154,
-      "eval_samples_per_second": 142.581,
-      "eval_steps_per_second": 1.398,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.1487188339233398,
+      "eval_runtime": 1.6683,
+      "eval_samples_per_second": 61.142,
+      "eval_steps_per_second": 0.599,
       "step": 588
     },
     {
       "epoch": 98.33,
-      "learning_rate": 3.3333333333333335e-07,
-      "loss": 1.0317,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 0.8739,
       "step": 590
     },
     {
       "epoch": 99.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.2937029600143433,
-      "eval_runtime": 0.6831,
-      "eval_samples_per_second": 149.311,
-      "eval_steps_per_second": 1.464,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.147989273071289,
+      "eval_runtime": 0.9001,
+      "eval_samples_per_second": 113.319,
+      "eval_steps_per_second": 1.111,
       "step": 594
     },
     {
       "epoch": 100.0,
-      "learning_rate": 0.0,
-      "loss": 1.0391,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.8742,
       "step": 600
     },
     {
       "epoch": 100.0,
-      "eval_accuracy": 0.8431372549019608,
-      "eval_loss": 1.2935162782669067,
-      "eval_runtime": 0.7374,
-      "eval_samples_per_second": 138.332,
-      "eval_steps_per_second": 1.356,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.141270637512207,
+      "eval_runtime": 0.9491,
+      "eval_samples_per_second": 107.465,
+      "eval_steps_per_second": 1.054,
       "step": 600
     },
     {
-      "epoch": 100.0,
-      "step": 600,
-      "total_flos": 4.44899699693568e+18,
-      "train_loss": 1.6557193438212077,
-      "train_runtime": 2084.5935,
-      "train_samples_per_second": 27.535,
-      "train_steps_per_second": 0.288
+      "epoch": 101.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1367518901824951,
+      "eval_runtime": 0.6665,
+      "eval_samples_per_second": 153.047,
+      "eval_steps_per_second": 1.5,
+      "step": 606
+    },
+    {
+      "epoch": 101.67,
+      "learning_rate": 3.055555555555556e-06,
+      "loss": 0.8679,
+      "step": 610
+    },
+    {
+      "epoch": 102.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.136061668395996,
+      "eval_runtime": 0.6609,
+      "eval_samples_per_second": 154.346,
+      "eval_steps_per_second": 1.513,
+      "step": 612
+    },
+    {
+      "epoch": 103.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1316593885421753,
+      "eval_runtime": 0.8305,
+      "eval_samples_per_second": 122.817,
+      "eval_steps_per_second": 1.204,
+      "step": 618
+    },
+    {
+      "epoch": 103.33,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.8516,
+      "step": 620
+    },
+    {
+      "epoch": 104.0,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.1296062469482422,
+      "eval_runtime": 0.7328,
+      "eval_samples_per_second": 139.191,
+      "eval_steps_per_second": 1.365,
+      "step": 624
+    },
+    {
+      "epoch": 105.0,
+      "learning_rate": 2.5e-06,
+      "loss": 0.876,
+      "step": 630
+    },
+    {
+      "epoch": 105.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1287850141525269,
+      "eval_runtime": 0.8191,
+      "eval_samples_per_second": 124.532,
+      "eval_steps_per_second": 1.221,
+      "step": 630
+    },
+    {
+      "epoch": 106.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.126371145248413,
+      "eval_runtime": 0.9539,
+      "eval_samples_per_second": 106.934,
+      "eval_steps_per_second": 1.048,
+      "step": 636
+    },
+    {
+      "epoch": 106.67,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.8591,
+      "step": 640
+    },
+    {
+      "epoch": 107.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1238181591033936,
+      "eval_runtime": 1.9507,
+      "eval_samples_per_second": 52.288,
+      "eval_steps_per_second": 0.513,
+      "step": 642
+    },
+    {
+      "epoch": 108.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1227307319641113,
+      "eval_runtime": 0.8133,
+      "eval_samples_per_second": 125.408,
+      "eval_steps_per_second": 1.229,
+      "step": 648
+    },
+    {
+      "epoch": 108.33,
+      "learning_rate": 1.944444444444445e-06,
+      "loss": 0.8586,
+      "step": 650
+    },
+    {
+      "epoch": 109.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1207623481750488,
+      "eval_runtime": 0.9855,
+      "eval_samples_per_second": 103.502,
+      "eval_steps_per_second": 1.015,
+      "step": 654
+    },
+    {
+      "epoch": 110.0,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.8415,
+      "step": 660
+    },
+    {
+      "epoch": 110.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1194456815719604,
+      "eval_runtime": 1.6344,
+      "eval_samples_per_second": 62.409,
+      "eval_steps_per_second": 0.612,
+      "step": 660
+    },
+    {
+      "epoch": 111.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1184879541397095,
+      "eval_runtime": 0.8861,
+      "eval_samples_per_second": 115.105,
+      "eval_steps_per_second": 1.128,
+      "step": 666
+    },
+    {
+      "epoch": 111.67,
+      "learning_rate": 1.3888888888888892e-06,
+      "loss": 0.8465,
+      "step": 670
+    },
+    {
+      "epoch": 112.0,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.1177918910980225,
+      "eval_runtime": 0.9577,
+      "eval_samples_per_second": 106.503,
+      "eval_steps_per_second": 1.044,
+      "step": 672
+    },
+    {
+      "epoch": 113.0,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.1184459924697876,
+      "eval_runtime": 1.7722,
+      "eval_samples_per_second": 57.556,
+      "eval_steps_per_second": 0.564,
+      "step": 678
+    },
+    {
+      "epoch": 113.33,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.8503,
+      "step": 680
+    },
+    {
+      "epoch": 114.0,
+      "eval_accuracy": 0.8431372549019608,
+      "eval_loss": 1.1182842254638672,
+      "eval_runtime": 0.7295,
+      "eval_samples_per_second": 139.825,
+      "eval_steps_per_second": 1.371,
+      "step": 684
+    },
+    {
+      "epoch": 115.0,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.8332,
+      "step": 690
+    },
+    {
+      "epoch": 115.0,
+      "eval_accuracy": 0.8431372549019608,
+      "eval_loss": 1.1173955202102661,
+      "eval_runtime": 0.9783,
+      "eval_samples_per_second": 104.259,
+      "eval_steps_per_second": 1.022,
+      "step": 690
+    },
+    {
+      "epoch": 116.0,
+      "eval_accuracy": 0.8431372549019608,
+      "eval_loss": 1.1165024042129517,
+      "eval_runtime": 1.541,
+      "eval_samples_per_second": 66.193,
+      "eval_steps_per_second": 0.649,
+      "step": 696
+    },
+    {
+      "epoch": 116.67,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.8476,
+      "step": 700
+    },
+    {
+      "epoch": 117.0,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.1152539253234863,
+      "eval_runtime": 0.7315,
+      "eval_samples_per_second": 139.446,
+      "eval_steps_per_second": 1.367,
+      "step": 702
+    },
+    {
+      "epoch": 118.0,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 1.1142131090164185,
+      "eval_runtime": 1.4479,
+      "eval_samples_per_second": 70.448,
+      "eval_steps_per_second": 0.691,
+      "step": 708
+    },
+    {
+      "epoch": 118.33,
+      "learning_rate": 2.7777777777777776e-07,
+      "loss": 0.8382,
+      "step": 710
+    },
+    {
+      "epoch": 119.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1136654615402222,
+      "eval_runtime": 1.4591,
+      "eval_samples_per_second": 69.906,
+      "eval_steps_per_second": 0.685,
+      "step": 714
+    },
+    {
+      "epoch": 120.0,
+      "learning_rate": 0.0,
+      "loss": 0.8527,
+      "step": 720
+    },
+    {
+      "epoch": 120.0,
+      "eval_accuracy": 0.8627450980392157,
+      "eval_loss": 1.1135263442993164,
+      "eval_runtime": 0.7983,
+      "eval_samples_per_second": 127.768,
+      "eval_steps_per_second": 1.253,
+      "step": 720
+    },
+    {
+      "epoch": 120.0,
+      "step": 720,
+      "total_flos": 5.338796396322816e+18,
+      "train_loss": 1.4664161258273654,
+      "train_runtime": 3110.7108,
+      "train_samples_per_second": 22.143,
+      "train_steps_per_second": 0.231
     }
   ],
   "logging_steps": 10,
-  "max_steps": 600,
+  "max_steps": 720,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 100,
+  "num_train_epochs": 120,
   "save_steps": 500,
-  "total_flos": 4.44899699693568e+18,
+  "total_flos": 5.338796396322816e+18,
   "train_batch_size": 100,
   "trial_name": null,
   "trial_params": null