End of training

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [facebook/convnextv2-base-1k-224](https://huggingface.co/facebook/convnextv2-base-1k-224) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.7475
-- Accuracy: 0.3520
+- Loss: 1.3599
+- Accuracy: 0.4190
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 27.91,
-    "eval_accuracy": 0.4382284382284382,
-    "eval_loss": 1.447946548461914,
-    "eval_runtime": 8.404,
-    "eval_samples_per_second": 51.047,
-    "eval_steps_per_second": 1.666,
-    "total_flos": 3.007291871298355e+18,
-    "train_loss": 1.1542485936482747,
-    "train_runtime": 2419.37,
-    "train_samples_per_second": 16.864,
-    "train_steps_per_second": 0.124
+    "epoch": 30.0,
+    "eval_accuracy": 0.41899441340782123,
+    "eval_loss": 1.359870195388794,
+    "eval_runtime": 7.4764,
+    "eval_samples_per_second": 47.884,
+    "eval_steps_per_second": 1.605,
+    "total_flos": 4.840276186658304e+18,
+    "train_loss": 1.0325976332028708,
+    "train_runtime": 3493.542,
+    "train_samples_per_second": 17.492,
+    "train_steps_per_second": 0.137
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 27.91,
-    "eval_accuracy": 0.4382284382284382,
-    "eval_loss": 1.447946548461914,
-    "eval_runtime": 8.404,
-    "eval_samples_per_second": 51.047,
-    "eval_steps_per_second": 1.666
+    "epoch": 30.0,
+    "eval_accuracy": 0.41899441340782123,
+    "eval_loss": 1.359870195388794,
+    "eval_runtime": 7.4764,
+    "eval_samples_per_second": 47.884,
+    "eval_steps_per_second": 1.605
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 27.91,
-    "total_flos": 3.007291871298355e+18,
-    "train_loss": 1.1542485936482747,
-    "train_runtime": 2419.37,
-    "train_samples_per_second": 16.864,
-    "train_steps_per_second": 0.124
+    "epoch": 30.0,
+    "total_flos": 4.840276186658304e+18,
+    "train_loss": 1.0325976332028708,
+    "train_runtime": 3493.542,
+    "train_samples_per_second": 17.492,
+    "train_steps_per_second": 0.137
 }

trainer_state.json

@@ -1,460 +1,586 @@
 {
-  "best_metric": 0.4382284382284382,
-  "best_model_checkpoint": "convnextv2-base-1k-224-for-pre_evaluation/checkpoint-268",
-  "epoch": 27.906976744186046,
+  "best_metric": 0.41899441340782123,
+  "best_model_checkpoint": "convnextv2-base-1k-224-for-pre_evaluation/checkpoint-128",
+  "epoch": 30.0,
   "eval_steps": 500,
-  "global_step": 300,
+  "global_step": 480,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.93,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 1.5952,
+      "epoch": 0.62,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 1.6,
       "step": 10
     },
     {
-      "epoch": 0.93,
-      "eval_accuracy": 0.29603729603729606,
-      "eval_loss": 1.5510554313659668,
-      "eval_runtime": 8.4324,
-      "eval_samples_per_second": 50.875,
-      "eval_steps_per_second": 1.66,
-      "step": 10
+      "epoch": 1.0,
+      "eval_accuracy": 0.29608938547486036,
+      "eval_loss": 1.5315604209899902,
+      "eval_runtime": 6.8342,
+      "eval_samples_per_second": 52.384,
+      "eval_steps_per_second": 1.756,
+      "step": 16
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 1.5238,
+      "epoch": 1.25,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 1.535,
       "step": 20
     },
     {
-      "epoch": 1.95,
-      "eval_accuracy": 0.34265734265734266,
-      "eval_loss": 1.5091006755828857,
-      "eval_runtime": 8.9811,
-      "eval_samples_per_second": 47.767,
-      "eval_steps_per_second": 1.559,
-      "step": 21
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 5e-05,
-      "loss": 1.4881,
+      "epoch": 1.88,
+      "learning_rate": 3.125e-05,
+      "loss": 1.5084,
       "step": 30
     },
     {
-      "epoch": 2.98,
-      "eval_accuracy": 0.34498834498834496,
-      "eval_loss": 1.4853538274765015,
-      "eval_runtime": 8.6991,
-      "eval_samples_per_second": 49.315,
-      "eval_steps_per_second": 1.609,
+      "epoch": 2.0,
+      "eval_accuracy": 0.2849162011173184,
+      "eval_loss": 1.5060781240463257,
+      "eval_runtime": 7.4696,
+      "eval_samples_per_second": 47.928,
+      "eval_steps_per_second": 1.607,
       "step": 32
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 4.814814814814815e-05,
-      "loss": 1.4708,
+      "epoch": 2.5,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.5134,
       "step": 40
     },
     {
-      "epoch": 4.0,
-      "eval_accuracy": 0.3473193473193473,
-      "eval_loss": 1.4616328477859497,
-      "eval_runtime": 8.3701,
-      "eval_samples_per_second": 51.254,
-      "eval_steps_per_second": 1.673,
-      "step": 43
+      "epoch": 3.0,
+      "eval_accuracy": 0.3240223463687151,
+      "eval_loss": 1.4968407154083252,
+      "eval_runtime": 7.1172,
+      "eval_samples_per_second": 50.301,
+      "eval_steps_per_second": 1.686,
+      "step": 48
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 4.62962962962963e-05,
-      "loss": 1.4361,
+      "epoch": 3.12,
+      "learning_rate": 4.976851851851852e-05,
+      "loss": 1.4694,
       "step": 50
     },
     {
-      "epoch": 4.93,
-      "eval_accuracy": 0.34498834498834496,
-      "eval_loss": 1.4416619539260864,
-      "eval_runtime": 8.7032,
-      "eval_samples_per_second": 49.292,
-      "eval_steps_per_second": 1.609,
-      "step": 53
-    },
-    {
-      "epoch": 5.58,
-      "learning_rate": 4.4444444444444447e-05,
-      "loss": 1.3764,
+      "epoch": 3.75,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 1.4663,
       "step": 60
     },
     {
-      "epoch": 5.95,
-      "eval_accuracy": 0.3752913752913753,
-      "eval_loss": 1.4134629964828491,
-      "eval_runtime": 8.9398,
-      "eval_samples_per_second": 47.988,
-      "eval_steps_per_second": 1.566,
+      "epoch": 4.0,
+      "eval_accuracy": 0.33519553072625696,
+      "eval_loss": 1.4607384204864502,
+      "eval_runtime": 7.4013,
+      "eval_samples_per_second": 48.37,
+      "eval_steps_per_second": 1.621,
       "step": 64
     },
     {
-      "epoch": 6.51,
-      "learning_rate": 4.259259259259259e-05,
-      "loss": 1.3333,
+      "epoch": 4.38,
+      "learning_rate": 4.745370370370371e-05,
+      "loss": 1.4375,
       "step": 70
     },
     {
-      "epoch": 6.98,
-      "eval_accuracy": 0.3986013986013986,
-      "eval_loss": 1.3822472095489502,
-      "eval_runtime": 8.7499,
-      "eval_samples_per_second": 49.029,
-      "eval_steps_per_second": 1.6,
-      "step": 75
-    },
-    {
-      "epoch": 7.44,
-      "learning_rate": 4.074074074074074e-05,
-      "loss": 1.3296,
+      "epoch": 5.0,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 1.4046,
       "step": 80
     },
     {
-      "epoch": 8.0,
-      "eval_accuracy": 0.36363636363636365,
-      "eval_loss": 1.4111592769622803,
-      "eval_runtime": 8.8496,
-      "eval_samples_per_second": 48.477,
-      "eval_steps_per_second": 1.582,
-      "step": 86
+      "epoch": 5.0,
+      "eval_accuracy": 0.3268156424581006,
+      "eval_loss": 1.4509011507034302,
+      "eval_runtime": 6.83,
+      "eval_samples_per_second": 52.416,
+      "eval_steps_per_second": 1.757,
+      "step": 80
     },
     {
-      "epoch": 8.37,
-      "learning_rate": 3.888888888888889e-05,
-      "loss": 1.2798,
+      "epoch": 5.62,
+      "learning_rate": 4.5138888888888894e-05,
+      "loss": 1.4085,
       "step": 90
     },
     {
-      "epoch": 8.93,
-      "eval_accuracy": 0.38927738927738925,
-      "eval_loss": 1.4038037061691284,
-      "eval_runtime": 8.5853,
-      "eval_samples_per_second": 49.969,
-      "eval_steps_per_second": 1.631,
+      "epoch": 6.0,
+      "eval_accuracy": 0.388268156424581,
+      "eval_loss": 1.4423185586929321,
+      "eval_runtime": 7.3683,
+      "eval_samples_per_second": 48.587,
+      "eval_steps_per_second": 1.629,
       "step": 96
     },
     {
-      "epoch": 9.3,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 1.3129,
+      "epoch": 6.25,
+      "learning_rate": 4.3981481481481486e-05,
+      "loss": 1.3765,
       "step": 100
     },
     {
-      "epoch": 9.95,
-      "eval_accuracy": 0.3776223776223776,
-      "eval_loss": 1.424072265625,
-      "eval_runtime": 8.6578,
-      "eval_samples_per_second": 49.551,
-      "eval_steps_per_second": 1.617,
-      "step": 107
-    },
-    {
-      "epoch": 10.23,
-      "learning_rate": 3.518518518518519e-05,
-      "loss": 1.3014,
+      "epoch": 6.88,
+      "learning_rate": 4.282407407407408e-05,
+      "loss": 1.3443,
       "step": 110
     },
     {
-      "epoch": 10.98,
-      "eval_accuracy": 0.38927738927738925,
-      "eval_loss": 1.356952428817749,
-      "eval_runtime": 8.5422,
-      "eval_samples_per_second": 50.221,
-      "eval_steps_per_second": 1.639,
-      "step": 118
+      "epoch": 7.0,
+      "eval_accuracy": 0.4022346368715084,
+      "eval_loss": 1.400512456893921,
+      "eval_runtime": 6.9156,
+      "eval_samples_per_second": 51.767,
+      "eval_steps_per_second": 1.735,
+      "step": 112
     },
     {
-      "epoch": 11.16,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 1.2332,
+      "epoch": 7.5,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.3025,
       "step": 120
     },
     {
-      "epoch": 12.0,
-      "eval_accuracy": 0.38927738927738925,
-      "eval_loss": 1.4072706699371338,
-      "eval_runtime": 8.3607,
-      "eval_samples_per_second": 51.312,
-      "eval_steps_per_second": 1.675,
-      "step": 129
+      "epoch": 8.0,
+      "eval_accuracy": 0.41899441340782123,
+      "eval_loss": 1.359870195388794,
+      "eval_runtime": 7.041,
+      "eval_samples_per_second": 50.845,
+      "eval_steps_per_second": 1.704,
+      "step": 128
     },
     {
-      "epoch": 12.09,
-      "learning_rate": 3.148148148148148e-05,
-      "loss": 1.212,
+      "epoch": 8.12,
+      "learning_rate": 4.0509259259259265e-05,
+      "loss": 1.2668,
       "step": 130
     },
     {
-      "epoch": 12.93,
-      "eval_accuracy": 0.40326340326340326,
-      "eval_loss": 1.376956820487976,
-      "eval_runtime": 8.3675,
-      "eval_samples_per_second": 51.27,
-      "eval_steps_per_second": 1.673,
-      "step": 139
-    },
-    {
-      "epoch": 13.02,
-      "learning_rate": 2.962962962962963e-05,
-      "loss": 1.1844,
+      "epoch": 8.75,
+      "learning_rate": 3.935185185185186e-05,
+      "loss": 1.2627,
       "step": 140
     },
     {
-      "epoch": 13.95,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 1.1763,
-      "step": 150
+      "epoch": 9.0,
+      "eval_accuracy": 0.39106145251396646,
+      "eval_loss": 1.3638169765472412,
+      "eval_runtime": 7.2532,
+      "eval_samples_per_second": 49.357,
+      "eval_steps_per_second": 1.654,
+      "step": 144
     },
     {
-      "epoch": 13.95,
-      "eval_accuracy": 0.3962703962703963,
-      "eval_loss": 1.3891488313674927,
-      "eval_runtime": 8.5717,
-      "eval_samples_per_second": 50.049,
-      "eval_steps_per_second": 1.633,
+      "epoch": 9.38,
+      "learning_rate": 3.8194444444444444e-05,
+      "loss": 1.2427,
       "step": 150
     },
     {
-      "epoch": 14.88,
-      "learning_rate": 2.5925925925925925e-05,
-      "loss": 1.124,
+      "epoch": 10.0,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 1.2099,
       "step": 160
     },
     {
-      "epoch": 14.98,
-      "eval_accuracy": 0.4125874125874126,
-      "eval_loss": 1.3915237188339233,
-      "eval_runtime": 8.5638,
-      "eval_samples_per_second": 50.095,
-      "eval_steps_per_second": 1.635,
-      "step": 161
+      "epoch": 10.0,
+      "eval_accuracy": 0.34916201117318435,
+      "eval_loss": 1.4057648181915283,
+      "eval_runtime": 7.508,
+      "eval_samples_per_second": 47.682,
+      "eval_steps_per_second": 1.598,
+      "step": 160
     },
     {
-      "epoch": 15.81,
-      "learning_rate": 2.4074074074074074e-05,
-      "loss": 1.0963,
+      "epoch": 10.62,
+      "learning_rate": 3.587962962962963e-05,
+      "loss": 1.2086,
       "step": 170
     },
     {
-      "epoch": 16.0,
-      "eval_accuracy": 0.4149184149184149,
-      "eval_loss": 1.4098657369613647,
-      "eval_runtime": 8.8116,
-      "eval_samples_per_second": 48.686,
-      "eval_steps_per_second": 1.589,
-      "step": 172
+      "epoch": 11.0,
+      "eval_accuracy": 0.3407821229050279,
+      "eval_loss": 1.443146824836731,
+      "eval_runtime": 7.4283,
+      "eval_samples_per_second": 48.194,
+      "eval_steps_per_second": 1.615,
+      "step": 176
     },
     {
-      "epoch": 16.74,
-      "learning_rate": 2.2222222222222223e-05,
-      "loss": 1.0547,
+      "epoch": 11.25,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 1.1684,
       "step": 180
     },
     {
-      "epoch": 16.93,
-      "eval_accuracy": 0.40326340326340326,
-      "eval_loss": 1.4206278324127197,
-      "eval_runtime": 8.7717,
-      "eval_samples_per_second": 48.907,
-      "eval_steps_per_second": 1.596,
-      "step": 182
-    },
-    {
-      "epoch": 17.67,
-      "learning_rate": 2.037037037037037e-05,
-      "loss": 1.0631,
+      "epoch": 11.88,
+      "learning_rate": 3.3564814814814815e-05,
+      "loss": 1.1393,
       "step": 190
     },
     {
-      "epoch": 17.95,
-      "eval_accuracy": 0.4195804195804196,
-      "eval_loss": 1.4040827751159668,
-      "eval_runtime": 8.3983,
-      "eval_samples_per_second": 51.082,
-      "eval_steps_per_second": 1.667,
-      "step": 193
+      "epoch": 12.0,
+      "eval_accuracy": 0.34916201117318435,
+      "eval_loss": 1.4143450260162354,
+      "eval_runtime": 7.1396,
+      "eval_samples_per_second": 50.143,
+      "eval_steps_per_second": 1.681,
+      "step": 192
     },
     {
-      "epoch": 18.6,
-      "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.9911,
+      "epoch": 12.5,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 1.1039,
       "step": 200
     },
     {
-      "epoch": 18.98,
-      "eval_accuracy": 0.4149184149184149,
-      "eval_loss": 1.4271957874298096,
-      "eval_runtime": 8.2919,
-      "eval_samples_per_second": 51.737,
-      "eval_steps_per_second": 1.688,
-      "step": 204
+      "epoch": 13.0,
+      "eval_accuracy": 0.388268156424581,
+      "eval_loss": 1.4305065870285034,
+      "eval_runtime": 6.8442,
+      "eval_samples_per_second": 52.307,
+      "eval_steps_per_second": 1.753,
+      "step": 208
     },
     {
-      "epoch": 19.53,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 1.005,
+      "epoch": 13.12,
+      "learning_rate": 3.125e-05,
+      "loss": 1.0641,
       "step": 210
     },
     {
-      "epoch": 20.0,
-      "eval_accuracy": 0.4219114219114219,
-      "eval_loss": 1.42105233669281,
-      "eval_runtime": 8.2769,
-      "eval_samples_per_second": 51.831,
-      "eval_steps_per_second": 1.691,
-      "step": 215
-    },
-    {
-      "epoch": 20.47,
-      "learning_rate": 1.4814814814814815e-05,
-      "loss": 0.9663,
+      "epoch": 13.75,
+      "learning_rate": 3.0092592592592593e-05,
+      "loss": 1.0551,
       "step": 220
     },
     {
-      "epoch": 20.93,
-      "eval_accuracy": 0.40093240093240096,
-      "eval_loss": 1.466171145439148,
-      "eval_runtime": 9.4718,
-      "eval_samples_per_second": 45.292,
-      "eval_steps_per_second": 1.478,
-      "step": 225
+      "epoch": 14.0,
+      "eval_accuracy": 0.35195530726256985,
+      "eval_loss": 1.5202596187591553,
+      "eval_runtime": 7.2925,
+      "eval_samples_per_second": 49.091,
+      "eval_steps_per_second": 1.646,
+      "step": 224
     },
     {
-      "epoch": 21.4,
-      "learning_rate": 1.2962962962962962e-05,
-      "loss": 0.9533,
+      "epoch": 14.38,
+      "learning_rate": 2.8935185185185186e-05,
+      "loss": 1.0686,
       "step": 230
     },
     {
-      "epoch": 21.95,
-      "eval_accuracy": 0.43356643356643354,
-      "eval_loss": 1.428614616394043,
-      "eval_runtime": 8.343,
-      "eval_samples_per_second": 51.42,
-      "eval_steps_per_second": 1.678,
-      "step": 236
-    },
-    {
-      "epoch": 22.33,
-      "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.9506,
+      "epoch": 15.0,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 1.0368,
       "step": 240
     },
     {
-      "epoch": 22.98,
-      "eval_accuracy": 0.43123543123543123,
-      "eval_loss": 1.413465976715088,
-      "eval_runtime": 8.7694,
-      "eval_samples_per_second": 48.92,
-      "eval_steps_per_second": 1.596,
-      "step": 247
+      "epoch": 15.0,
+      "eval_accuracy": 0.3324022346368715,
+      "eval_loss": 1.5117393732070923,
+      "eval_runtime": 6.8377,
+      "eval_samples_per_second": 52.356,
+      "eval_steps_per_second": 1.755,
+      "step": 240
     },
     {
-      "epoch": 23.26,
-      "learning_rate": 9.259259259259259e-06,
-      "loss": 0.8973,
+      "epoch": 15.62,
+      "learning_rate": 2.6620370370370372e-05,
+      "loss": 0.9753,
       "step": 250
     },
     {
-      "epoch": 24.0,
-      "eval_accuracy": 0.42657342657342656,
-      "eval_loss": 1.442847728729248,
-      "eval_runtime": 8.7464,
-      "eval_samples_per_second": 49.049,
-      "eval_steps_per_second": 1.601,
-      "step": 258
+      "epoch": 16.0,
+      "eval_accuracy": 0.3770949720670391,
+      "eval_loss": 1.4545259475708008,
+      "eval_runtime": 7.0796,
+      "eval_samples_per_second": 50.568,
+      "eval_steps_per_second": 1.695,
+      "step": 256
     },
     {
-      "epoch": 24.19,
-      "learning_rate": 7.4074074074074075e-06,
-      "loss": 0.8807,
+      "epoch": 16.25,
+      "learning_rate": 2.5462962962962965e-05,
+      "loss": 0.9677,
       "step": 260
     },
     {
-      "epoch": 24.93,
-      "eval_accuracy": 0.4382284382284382,
-      "eval_loss": 1.447946548461914,
-      "eval_runtime": 8.3339,
-      "eval_samples_per_second": 51.476,
-      "eval_steps_per_second": 1.68,
-      "step": 268
-    },
-    {
-      "epoch": 25.12,
-      "learning_rate": 5.555555555555556e-06,
-      "loss": 0.8731,
+      "epoch": 16.88,
+      "learning_rate": 2.4305555555555558e-05,
+      "loss": 0.938,
       "step": 270
     },
     {
-      "epoch": 25.95,
-      "eval_accuracy": 0.4289044289044289,
-      "eval_loss": 1.4429428577423096,
-      "eval_runtime": 8.7998,
-      "eval_samples_per_second": 48.751,
-      "eval_steps_per_second": 1.591,
-      "step": 279
+      "epoch": 17.0,
+      "eval_accuracy": 0.33519553072625696,
+      "eval_loss": 1.5396308898925781,
+      "eval_runtime": 7.1168,
+      "eval_samples_per_second": 50.304,
+      "eval_steps_per_second": 1.686,
+      "step": 272
     },
     {
-      "epoch": 26.05,
-      "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.8366,
+      "epoch": 17.5,
+      "learning_rate": 2.314814814814815e-05,
+      "loss": 0.899,
       "step": 280
     },
     {
-      "epoch": 26.98,
-      "learning_rate": 1.8518518518518519e-06,
-      "loss": 0.8472,
-      "step": 290
+      "epoch": 18.0,
+      "eval_accuracy": 0.3407821229050279,
+      "eval_loss": 1.5770219564437866,
+      "eval_runtime": 6.8822,
+      "eval_samples_per_second": 52.018,
+      "eval_steps_per_second": 1.744,
+      "step": 288
     },
     {
-      "epoch": 26.98,
-      "eval_accuracy": 0.43123543123543123,
-      "eval_loss": 1.4461231231689453,
-      "eval_runtime": 8.7766,
-      "eval_samples_per_second": 48.88,
-      "eval_steps_per_second": 1.595,
+      "epoch": 18.12,
+      "learning_rate": 2.1990740740740743e-05,
+      "loss": 0.9047,
       "step": 290
     },
     {
-      "epoch": 27.91,
-      "learning_rate": 0.0,
-      "loss": 0.8348,
+      "epoch": 18.75,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.8629,
       "step": 300
     },
     {
-      "epoch": 27.91,
-      "eval_accuracy": 0.43356643356643354,
-      "eval_loss": 1.453087568283081,
-      "eval_runtime": 8.3523,
-      "eval_samples_per_second": 51.363,
-      "eval_steps_per_second": 1.676,
-      "step": 300
+      "epoch": 19.0,
+      "eval_accuracy": 0.3128491620111732,
+      "eval_loss": 1.7105906009674072,
+      "eval_runtime": 7.3502,
+      "eval_samples_per_second": 48.706,
+      "eval_steps_per_second": 1.633,
+      "step": 304
+    },
+    {
+      "epoch": 19.38,
+      "learning_rate": 1.967592592592593e-05,
+      "loss": 0.8624,
+      "step": 310
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.8674,
+      "step": 320
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.33519553072625696,
+      "eval_loss": 1.5864217281341553,
+      "eval_runtime": 7.1963,
+      "eval_samples_per_second": 49.748,
+      "eval_steps_per_second": 1.668,
+      "step": 320
+    },
+    {
+      "epoch": 20.62,
+      "learning_rate": 1.736111111111111e-05,
+      "loss": 0.7789,
+      "step": 330
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.3407821229050279,
+      "eval_loss": 1.6129050254821777,
+      "eval_runtime": 7.1496,
+      "eval_samples_per_second": 50.073,
+      "eval_steps_per_second": 1.678,
+      "step": 336
+    },
+    {
+      "epoch": 21.25,
+      "learning_rate": 1.6203703703703704e-05,
+      "loss": 0.8161,
+      "step": 340
+    },
+    {
+      "epoch": 21.88,
+      "learning_rate": 1.5046296296296297e-05,
+      "loss": 0.7426,
+      "step": 350
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.36033519553072624,
+      "eval_loss": 1.6353477239608765,
+      "eval_runtime": 7.4456,
+      "eval_samples_per_second": 48.082,
+      "eval_steps_per_second": 1.612,
+      "step": 352
+    },
+    {
+      "epoch": 22.5,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.7677,
+      "step": 360
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.3463687150837989,
+      "eval_loss": 1.6793445348739624,
+      "eval_runtime": 6.994,
+      "eval_samples_per_second": 51.187,
+      "eval_steps_per_second": 1.716,
+      "step": 368
+    },
+    {
+      "epoch": 23.12,
+      "learning_rate": 1.2731481481481482e-05,
+      "loss": 0.7327,
+      "step": 370
+    },
+    {
+      "epoch": 23.75,
+      "learning_rate": 1.1574074074074075e-05,
+      "loss": 0.7172,
+      "step": 380
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.3575418994413408,
+      "eval_loss": 1.6759321689605713,
+      "eval_runtime": 7.4394,
+      "eval_samples_per_second": 48.122,
+      "eval_steps_per_second": 1.613,
+      "step": 384
+    },
+    {
+      "epoch": 24.38,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 0.6759,
+      "step": 390
+    },
+    {
+      "epoch": 25.0,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.6809,
+      "step": 400
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.3659217877094972,
+      "eval_loss": 1.701292634010315,
+      "eval_runtime": 7.4138,
+      "eval_samples_per_second": 48.288,
+      "eval_steps_per_second": 1.619,
+      "step": 400
+    },
+    {
+      "epoch": 25.62,
+      "learning_rate": 8.101851851851852e-06,
+      "loss": 0.6619,
+      "step": 410
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.36312849162011174,
+      "eval_loss": 1.7108293771743774,
+      "eval_runtime": 7.238,
+      "eval_samples_per_second": 49.461,
+      "eval_steps_per_second": 1.658,
+      "step": 416
+    },
+    {
+      "epoch": 26.25,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.6773,
+      "step": 420
+    },
+    {
+      "epoch": 26.88,
+      "learning_rate": 5.787037037037038e-06,
+      "loss": 0.6656,
+      "step": 430
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.3715083798882682,
+      "eval_loss": 1.7327028512954712,
+      "eval_runtime": 6.8416,
+      "eval_samples_per_second": 52.327,
+      "eval_steps_per_second": 1.754,
+      "step": 432
     },
     {
-      "epoch": 27.91,
-      "step": 300,
-      "total_flos": 3.007291871298355e+18,
-      "train_loss": 1.1542485936482747,
-      "train_runtime": 2419.37,
-      "train_samples_per_second": 16.864,
-      "train_steps_per_second": 0.124
+      "epoch": 27.5,
+      "learning_rate": 4.6296296296296296e-06,
+      "loss": 0.6258,
+      "step": 440
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.3547486033519553,
+      "eval_loss": 1.7377949953079224,
+      "eval_runtime": 7.2785,
+      "eval_samples_per_second": 49.186,
+      "eval_steps_per_second": 1.649,
+      "step": 448
+    },
+    {
+      "epoch": 28.12,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 0.6646,
+      "step": 450
+    },
+    {
+      "epoch": 28.75,
+      "learning_rate": 2.3148148148148148e-06,
+      "loss": 0.6173,
+      "step": 460
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.36033519553072624,
+      "eval_loss": 1.7461235523223877,
+      "eval_runtime": 6.8622,
+      "eval_samples_per_second": 52.17,
+      "eval_steps_per_second": 1.749,
+      "step": 464
+    },
+    {
+      "epoch": 29.38,
+      "learning_rate": 1.1574074074074074e-06,
+      "loss": 0.6482,
+      "step": 470
+    },
+    {
+      "epoch": 30.0,
+      "learning_rate": 0.0,
+      "loss": 0.6214,
+      "step": 480
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.35195530726256985,
+      "eval_loss": 1.7475444078445435,
+      "eval_runtime": 7.4355,
+      "eval_samples_per_second": 48.148,
+      "eval_steps_per_second": 1.614,
+      "step": 480
+    },
+    {
+      "epoch": 30.0,
+      "step": 480,
+      "total_flos": 4.840276186658304e+18,
+      "train_loss": 1.0325976332028708,
+      "train_runtime": 3493.542,
+      "train_samples_per_second": 17.492,
+      "train_steps_per_second": 0.137
     }
   ],
   "logging_steps": 10,
-  "max_steps": 300,
+  "max_steps": 480,
   "num_train_epochs": 30,
   "save_steps": 500,
-  "total_flos": 3.007291871298355e+18,
+  "total_flos": 4.840276186658304e+18,
   "trial_name": null,
   "trial_params": null
 }