Upload trainer_state.json

trainer_state.json

@@ -1,83 +1,314 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0,
-  "eval_steps": 500,
-  "global_step": 99,
+  "epoch": 6.0,
+  "eval_steps": 10,
+  "global_step": 198,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.30303030303030304,
-      "grad_norm": 8.750067710876465,
-      "learning_rate": 1.0000000000000002e-06,
-      "loss": 3.3586,
+      "grad_norm": 6.751555442810059,
+      "learning_rate": 6.5224903051542e-06,
+      "loss": 0.1213,
       "step": 10
     },
+    {
+      "epoch": 0.30303030303030304,
+      "eval_loss": 0.9380465149879456,
+      "eval_runtime": 12.8667,
+      "eval_samples_per_second": 10.104,
+      "eval_steps_per_second": 0.699,
+      "step": 10
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 1.701265811920166,
+      "learning_rate": 1.30449806103084e-05,
+      "loss": 0.0674,
+      "step": 20
+    },
     {
       "epoch": 0.6060606060606061,
-      "grad_norm": 9.16357421875,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 3.332,
+      "eval_loss": 0.7714130282402039,
+      "eval_runtime": 13.1222,
+      "eval_samples_per_second": 9.907,
+      "eval_steps_per_second": 0.686,
       "step": 20
     },
     {
       "epoch": 0.9090909090909091,
-      "grad_norm": 8.869489669799805,
-      "learning_rate": 3e-06,
-      "loss": 3.3107,
+      "grad_norm": 0.47370675206184387,
+      "learning_rate": 1.9567470915462598e-05,
+      "loss": 0.0152,
+      "step": 30
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "eval_loss": 0.6980556845664978,
+      "eval_runtime": 13.1297,
+      "eval_samples_per_second": 9.901,
+      "eval_steps_per_second": 0.685,
       "step": 30
     },
     {
       "epoch": 1.2121212121212122,
-      "grad_norm": 8.31998348236084,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 3.2271,
+      "grad_norm": 0.05018394812941551,
+      "learning_rate": 2.60899612206168e-05,
+      "loss": 0.0092,
+      "step": 40
+    },
+    {
+      "epoch": 1.2121212121212122,
+      "eval_loss": 0.6736735105514526,
+      "eval_runtime": 13.2489,
+      "eval_samples_per_second": 9.812,
+      "eval_steps_per_second": 0.679,
       "step": 40
     },
     {
       "epoch": 1.5151515151515151,
-      "grad_norm": 8.316051483154297,
-      "learning_rate": 5e-06,
-      "loss": 3.0808,
+      "grad_norm": 0.09279070049524307,
+      "learning_rate": 3.2612451525771e-05,
+      "loss": 0.0061,
+      "step": 50
+    },
+    {
+      "epoch": 1.5151515151515151,
+      "eval_loss": 0.6550866961479187,
+      "eval_runtime": 13.7983,
+      "eval_samples_per_second": 9.421,
+      "eval_steps_per_second": 0.652,
       "step": 50
     },
     {
       "epoch": 1.8181818181818183,
-      "grad_norm": 8.299915313720703,
-      "learning_rate": 6e-06,
-      "loss": 3.0593,
+      "grad_norm": 0.15316803753376007,
+      "learning_rate": 3.9134941830925195e-05,
+      "loss": 0.0054,
       "step": 60
     },
+    {
+      "epoch": 1.8181818181818183,
+      "eval_loss": 0.6054828763008118,
+      "eval_runtime": 13.111,
+      "eval_samples_per_second": 9.915,
+      "eval_steps_per_second": 0.686,
+      "step": 60
+    },
+    {
+      "epoch": 2.121212121212121,
+      "grad_norm": 0.038146719336509705,
+      "learning_rate": 4.56574321360794e-05,
+      "loss": 0.005,
+      "step": 70
+    },
     {
       "epoch": 2.121212121212121,
-      "grad_norm": 7.488762855529785,
-      "learning_rate": 7.000000000000001e-06,
-      "loss": 2.8781,
+      "eval_loss": 0.5666611194610596,
+      "eval_runtime": 13.2994,
+      "eval_samples_per_second": 9.775,
+      "eval_steps_per_second": 0.677,
       "step": 70
     },
     {
       "epoch": 2.4242424242424243,
-      "grad_norm": 8.20336627960205,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 2.731,
+      "grad_norm": 0.03521318733692169,
+      "learning_rate": 5.21799224412336e-05,
+      "loss": 0.0031,
+      "step": 80
+    },
+    {
+      "epoch": 2.4242424242424243,
+      "eval_loss": 0.5637892484664917,
+      "eval_runtime": 13.5395,
+      "eval_samples_per_second": 9.602,
+      "eval_steps_per_second": 0.665,
       "step": 80
     },
     {
       "epoch": 2.7272727272727275,
-      "grad_norm": 9.079769134521484,
-      "learning_rate": 9e-06,
-      "loss": 2.5706,
+      "grad_norm": 0.021656330674886703,
+      "learning_rate": 5.8702412746387796e-05,
+      "loss": 0.0025,
+      "step": 90
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "eval_loss": 0.5629878044128418,
+      "eval_runtime": 12.8928,
+      "eval_samples_per_second": 10.083,
+      "eval_steps_per_second": 0.698,
       "step": 90
+    },
+    {
+      "epoch": 3.0303030303030303,
+      "grad_norm": 0.023006413131952286,
+      "learning_rate": 6.5224903051542e-05,
+      "loss": 0.0023,
+      "step": 100
+    },
+    {
+      "epoch": 3.0303030303030303,
+      "eval_loss": 0.5643568634986877,
+      "eval_runtime": 14.0003,
+      "eval_samples_per_second": 9.285,
+      "eval_steps_per_second": 0.643,
+      "step": 100
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.018860142678022385,
+      "learning_rate": 7.17473933566962e-05,
+      "loss": 0.0017,
+      "step": 110
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "eval_loss": 0.5654418468475342,
+      "eval_runtime": 12.8973,
+      "eval_samples_per_second": 10.08,
+      "eval_steps_per_second": 0.698,
+      "step": 110
+    },
+    {
+      "epoch": 3.6363636363636362,
+      "grad_norm": 0.011563817039132118,
+      "learning_rate": 7.826988366185039e-05,
+      "loss": 0.0016,
+      "step": 120
+    },
+    {
+      "epoch": 3.6363636363636362,
+      "eval_loss": 0.5633814930915833,
+      "eval_runtime": 13.2174,
+      "eval_samples_per_second": 9.836,
+      "eval_steps_per_second": 0.681,
+      "step": 120
+    },
+    {
+      "epoch": 3.9393939393939394,
+      "grad_norm": 0.013678439892828465,
+      "learning_rate": 8.47923739670046e-05,
+      "loss": 0.0015,
+      "step": 130
+    },
+    {
+      "epoch": 3.9393939393939394,
+      "eval_loss": 0.5603265166282654,
+      "eval_runtime": 13.4745,
+      "eval_samples_per_second": 9.648,
+      "eval_steps_per_second": 0.668,
+      "step": 130
+    },
+    {
+      "epoch": 4.242424242424242,
+      "grad_norm": 0.012644356116652489,
+      "learning_rate": 9.13148642721588e-05,
+      "loss": 0.0013,
+      "step": 140
+    },
+    {
+      "epoch": 4.242424242424242,
+      "eval_loss": 0.5610027313232422,
+      "eval_runtime": 13.2941,
+      "eval_samples_per_second": 9.779,
+      "eval_steps_per_second": 0.677,
+      "step": 140
+    },
+    {
+      "epoch": 4.545454545454545,
+      "grad_norm": 0.010118371807038784,
+      "learning_rate": 9.783735457731299e-05,
+      "loss": 0.001,
+      "step": 150
+    },
+    {
+      "epoch": 4.545454545454545,
+      "eval_loss": 0.5560722947120667,
+      "eval_runtime": 13.2972,
+      "eval_samples_per_second": 9.777,
+      "eval_steps_per_second": 0.677,
+      "step": 150
+    },
+    {
+      "epoch": 4.848484848484849,
+      "grad_norm": 0.00965981837362051,
+      "learning_rate": 0.0001043598448824672,
+      "loss": 0.001,
+      "step": 160
+    },
+    {
+      "epoch": 4.848484848484849,
+      "eval_loss": 0.5564282536506653,
+      "eval_runtime": 13.6075,
+      "eval_samples_per_second": 9.554,
+      "eval_steps_per_second": 0.661,
+      "step": 160
+    },
+    {
+      "epoch": 5.151515151515151,
+      "grad_norm": 0.01098481472581625,
+      "learning_rate": 0.0001108823351876214,
+      "loss": 0.0009,
+      "step": 170
+    },
+    {
+      "epoch": 5.151515151515151,
+      "eval_loss": 0.5595658421516418,
+      "eval_runtime": 13.1715,
+      "eval_samples_per_second": 9.87,
+      "eval_steps_per_second": 0.683,
+      "step": 170
+    },
+    {
+      "epoch": 5.454545454545454,
+      "grad_norm": 0.006646531634032726,
+      "learning_rate": 0.00011740482549277559,
+      "loss": 0.0008,
+      "step": 180
+    },
+    {
+      "epoch": 5.454545454545454,
+      "eval_loss": 0.5603924989700317,
+      "eval_runtime": 13.6131,
+      "eval_samples_per_second": 9.55,
+      "eval_steps_per_second": 0.661,
+      "step": 180
+    },
+    {
+      "epoch": 5.757575757575758,
+      "grad_norm": 0.0069460393860936165,
+      "learning_rate": 0.0001239273157979298,
+      "loss": 0.0007,
+      "step": 190
+    },
+    {
+      "epoch": 5.757575757575758,
+      "eval_loss": 0.5604722499847412,
+      "eval_runtime": 13.0651,
+      "eval_samples_per_second": 9.95,
+      "eval_steps_per_second": 0.689,
+      "step": 190
     }
   ],
   "logging_steps": 10,
-  "max_steps": 99,
+  "max_steps": 198,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
+  "num_train_epochs": 6,
   "save_steps": 500,
   "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
     "TrainerControl": {
       "args": {
         "should_epoch_stop": false,
@@ -89,7 +320,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.20913507233792e+17,
+  "total_flos": 2.41827014467584e+17,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null