End of training

all_results.json

@@ -2,13 +2,13 @@
     "epoch": 2.0,
     "eval_exact_match": 58.04655870445344,
     "eval_f1": 70.00816554397369,
-    "eval_runtime": 37.0671,
+    "eval_runtime": 36.0165,
     "eval_samples": 6056,
-    "eval_samples_per_second": 163.379,
-    "eval_steps_per_second": 20.422,
+    "eval_samples_per_second": 168.145,
+    "eval_steps_per_second": 21.018,
     "train_loss": 0.013409816984013865,
-    "train_runtime": 36.1557,
+    "train_runtime": 37.8508,
     "train_samples": 87739,
-    "train_samples_per_second": 4853.397,
-    "train_steps_per_second": 404.473
+    "train_samples_per_second": 4636.049,
+    "train_steps_per_second": 386.359
 }

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cda9b172f98c9e9376f0df6c872fa698d46bbfd4d45b552f9d3a980ab1aa0a50
-size 28157111
+oid sha256:271e840903ac63bafb21a0f09ae77179513098f3b85c0ae312971ac0114b9c9f
+size 28157318

eval_results.json

@@ -2,8 +2,8 @@
     "epoch": 2.0,
     "eval_exact_match": 58.04655870445344,
     "eval_f1": 70.00816554397369,
-    "eval_runtime": 37.0671,
+    "eval_runtime": 36.0165,
     "eval_samples": 6056,
-    "eval_samples_per_second": 163.379,
-    "eval_steps_per_second": 20.422
+    "eval_samples_per_second": 168.145,
+    "eval_steps_per_second": 21.018
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.0,
     "train_loss": 0.013409816984013865,
-    "train_runtime": 36.1557,
+    "train_runtime": 37.8508,
     "train_samples": 87739,
-    "train_samples_per_second": 4853.397,
-    "train_steps_per_second": 404.473
+    "train_samples_per_second": 4636.049,
+    "train_steps_per_second": 386.359
 }

trainer_state.json

@@ -187,9 +187,9 @@
       "step": 14624,
       "total_flos": 1847651335621632.0,
       "train_loss": 0.013409816984013865,
-      "train_runtime": 36.1557,
-      "train_samples_per_second": 4853.397,
-      "train_steps_per_second": 404.473
+      "train_runtime": 37.8508,
+      "train_samples_per_second": 4636.049,
+      "train_steps_per_second": 386.359
     }
   ],
   "logging_steps": 500,