End of training

trainer_state.json

@@ -3,30 +3,23 @@
   "best_model_checkpoint": null,
   "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 501,
+  "global_step": 408,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
-    {
-      "epoch": 2.9940119760479043,
-      "grad_norm": 0.8153374195098877,
-      "learning_rate": 9.98003992015968e-08,
-      "loss": 0.9849,
-      "step": 500
-    },
     {
       "epoch": 3.0,
-      "step": 501,
-      "total_flos": 2.100147738850099e+16,
-      "train_loss": 0.9845637217491211,
-      "train_runtime": 1100.6816,
-      "train_samples_per_second": 1.818,
-      "train_steps_per_second": 0.455
+      "step": 408,
+      "total_flos": 1.5770782755979264e+16,
+      "train_loss": 0.6231819227630017,
+      "train_runtime": 792.1227,
+      "train_samples_per_second": 2.053,
+      "train_steps_per_second": 0.515
     }
   ],
   "logging_steps": 500,
-  "max_steps": 501,
+  "max_steps": 408,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,
@@ -42,7 +35,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.100147738850099e+16,
+  "total_flos": 1.5770782755979264e+16,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null