End of training

all_results.json

@@ -1,11 +1,11 @@
 {
     "epoch": 20.0,
-    "eval_exact_match": 73.6802270577105,
-    "eval_f1": 83.81913172157984,
+    "eval_exact_match": 69.46073793755913,
+    "eval_f1": 81.37817461204958,
     "eval_samples": 10790,
-    "train_loss": 0.12154395081276118,
-    "train_runtime": 262.3624,
+    "train_loss": 2.5731575012207033,
+    "train_runtime": 296.5753,
     "train_samples": 1024,
-    "train_samples_per_second": 78.06,
-    "train_steps_per_second": 6.556
+    "train_samples_per_second": 69.055,
+    "train_steps_per_second": 1.079
 }

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:41c62aef08bf8f69e0d83cd5da4f847b47690e7c4ca60d1337590ab7ae1644be
-size 49075762
+oid sha256:d028f460cdc9e6f7c4f5fbd13ca760928d008a2d85be87a3a4f0e850cf79eed5
+size 48796622

eval_results.json

@@ -1,6 +1,6 @@
 {
     "epoch": 20.0,
-    "eval_exact_match": 73.6802270577105,
-    "eval_f1": 83.81913172157984,
+    "eval_exact_match": 69.46073793755913,
+    "eval_f1": 81.37817461204958,
     "eval_samples": 10790
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.12154395081276118,
-    "train_runtime": 262.3624,
+    "train_loss": 2.5731575012207033,
+    "train_runtime": 296.5753,
     "train_samples": 1024,
-    "train_samples_per_second": 78.06,
-    "train_steps_per_second": 6.556
+    "train_samples_per_second": 69.055,
+    "train_steps_per_second": 1.079
 }

trainer_state.json

@@ -2,42 +2,24 @@
   "best_metric": null,
   "best_model_checkpoint": null,
   "epoch": 20.0,
-  "global_step": 1720,
+  "global_step": 320,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
-    {
-      "epoch": 5.81,
-      "learning_rate": 7.881136950904393e-05,
-      "loss": 2.578,
-      "step": 500
-    },
-    {
-      "epoch": 11.63,
-      "learning_rate": 4.651162790697675e-05,
-      "loss": 0.4809,
-      "step": 1000
-    },
-    {
-      "epoch": 17.44,
-      "learning_rate": 1.421188630490956e-05,
-      "loss": 0.3023,
-      "step": 1500
-    },
     {
       "epoch": 20.0,
-      "step": 1720,
-      "total_flos": 1.456427177607168e+16,
-      "train_loss": 0.12154395081276118,
-      "train_runtime": 262.3624,
-      "train_samples_per_second": 78.06,
-      "train_steps_per_second": 6.556
+      "step": 320,
+      "total_flos": 1.456427168169984e+16,
+      "train_loss": 2.5731575012207033,
+      "train_runtime": 296.5753,
+      "train_samples_per_second": 69.055,
+      "train_steps_per_second": 1.079
     }
   ],
-  "max_steps": 1720,
+  "max_steps": 320,
   "num_train_epochs": 20,
-  "total_flos": 1.456427177607168e+16,
+  "total_flos": 1.456427168169984e+16,
   "trial_name": null,
   "trial_params": null
 }