update model

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:49f0515aa524f9f8466e9b913bbc06cdebbdcdfd404ca0057876687496978c03
+oid sha256:e8e503d2a6e9e148d20417c2f5a1801a3e1cc30ad9003ea9f29a7e63ead0f6f7
 size 5303326905

tokenizer_config.json

@@ -19,7 +19,7 @@
     "single_word": false
   },
   "errors": "replace",
-  "model_max_length": 1053,
+  "model_max_length": 1536,
   "pad_token": {
     "__type": "AddedToken",
     "content": "<pad>",

trainer_state.json

@@ -1,22 +1,67 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.9389671361502347,
-  "global_step": 600,
+  "epoch": 6.0,
+  "global_step": 3834,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.78,
-      "learning_rate": 1.525551371705218e-05,
-      "loss": 0.0234,
+      "learning_rate": 1.793975255513717e-05,
+      "loss": 0.0242,
       "step": 500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.5250134480903713e-05,
+      "loss": 0.002,
+      "step": 1000
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.2560516406670254e-05,
+      "loss": 0.0013,
+      "step": 1500
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.870898332436795e-06,
+      "loss": 0.0011,
+      "step": 2000
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 7.181280258203336e-06,
+      "loss": 0.0009,
+      "step": 2500
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 4.491662183969877e-06,
+      "loss": 0.0006,
+      "step": 3000
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 1.8020441097364175e-06,
+      "loss": 0.0005,
+      "step": 3500
+    },
+    {
+      "epoch": 6.0,
+      "step": 3834,
+      "total_flos": 2401267745816576.0,
+      "train_loss": 0.0040198439031637265,
+      "train_runtime": 24850.8432,
+      "train_samples_per_second": 19.743,
+      "train_steps_per_second": 0.154
     }
   ],
-  "max_steps": 1917,
-  "num_train_epochs": 3,
-  "total_flos": 362746747027456.0,
+  "max_steps": 3834,
+  "num_train_epochs": 6,
+  "total_flos": 2401267745816576.0,
   "trial_name": null,
   "trial_params": null
 }