End of training

all_results.json

@@ -5,15 +5,15 @@
     "eval_rouge1": 0.3297614987151056,
     "eval_rouge2": 0.08429294540985294,
     "eval_rougeL": 0.2561476738686219,
-    "eval_runtime": 28.613,
+    "eval_runtime": 26.2557,
     "eval_samples": 847,
-    "eval_samples_per_second": 29.602,
-    "eval_steps_per_second": 3.705,
+    "eval_samples_per_second": 32.26,
+    "eval_steps_per_second": 4.037,
     "perplexity": 22.834978708992665,
     "total_flos": 2000451796992000.0,
-    "train_loss": 4.049833148250759,
-    "train_runtime": 274.3289,
+    "train_loss": 4.049833530916316,
+    "train_runtime": 263.2573,
     "train_samples": 2552,
-    "train_samples_per_second": 18.605,
-    "train_steps_per_second": 2.326
+    "train_samples_per_second": 19.388,
+    "train_steps_per_second": 2.423
 }

egy_training_log.txt

@@ -2,3 +2,5 @@ WARNING:accelerate.utils.other:Detected kernel version 5.4.0, which is below the
 WARNING:root:Epoch 1.0: No losses recorded yet.
 INFO:absl:Using default tokenizer.
 INFO:root:Epoch 2.0: Train Loss = 4.3013, Eval Loss = 3.128293514251709
+INFO:__main__:*** Evaluate ***
+INFO:absl:Using default tokenizer.

eval_results.json

@@ -5,9 +5,9 @@
     "eval_rouge1": 0.3297614987151056,
     "eval_rouge2": 0.08429294540985294,
     "eval_rougeL": 0.2561476738686219,
-    "eval_runtime": 28.613,
+    "eval_runtime": 26.2557,
     "eval_samples": 847,
-    "eval_samples_per_second": 29.602,
-    "eval_steps_per_second": 3.705,
+    "eval_samples_per_second": 32.26,
+    "eval_steps_per_second": 4.037,
     "perplexity": 22.834978708992665
 }

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 2.0,
     "total_flos": 2000451796992000.0,
-    "train_loss": 4.049833148250759,
-    "train_runtime": 274.3289,
+    "train_loss": 4.049833530916316,
+    "train_runtime": 263.2573,
     "train_samples": 2552,
-    "train_samples_per_second": 18.605,
-    "train_steps_per_second": 2.326
+    "train_samples_per_second": 19.388,
+    "train_steps_per_second": 2.423
 }

trainer_state.json

@@ -10,7 +10,7 @@
   "log_history": [
     {
       "epoch": 1.567398119122257,
-      "grad_norm": 1.3536407947540283,
+      "grad_norm": 1.3536410331726074,
       "learning_rate": 5e-05,
       "loss": 4.3013,
       "step": 500
@@ -22,19 +22,19 @@
       "eval_rouge1": 0.3297614987151056,
       "eval_rouge2": 0.08429294540985294,
       "eval_rougeL": 0.2561476738686219,
-      "eval_runtime": 29.502,
-      "eval_samples_per_second": 28.71,
-      "eval_steps_per_second": 3.593,
+      "eval_runtime": 26.806,
+      "eval_samples_per_second": 31.597,
+      "eval_steps_per_second": 3.954,
       "step": 500
     },
     {
       "epoch": 2.0,
       "step": 638,
       "total_flos": 2000451796992000.0,
-      "train_loss": 4.049833148250759,
-      "train_runtime": 274.3289,
-      "train_samples_per_second": 18.605,
-      "train_steps_per_second": 2.326
+      "train_loss": 4.049833530916316,
+      "train_runtime": 263.2573,
+      "train_samples_per_second": 19.388,
+      "train_steps_per_second": 2.423
     }
   ],
   "logging_steps": 500,