Model save

README.md

@@ -27,7 +27,7 @@ print(output["generated_text"])
 
 ## Training procedure
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/bodasadallah2/review_evaluation/runs/80wak8tc) 
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/bodasadallah2/review_evaluation/runs/snmwv43u) 
 
 
 This model was trained with SFT.

all_results.json

@@ -1,13 +1,8 @@
 {
-    "eval_loss": 0.08637680858373642,
-    "eval_runtime": 20.0984,
-    "eval_samples": 100,
-    "eval_samples_per_second": 6.07,
-    "eval_steps_per_second": 0.796,
-    "total_flos": 1.3268723720781824e+17,
-    "train_loss": 0.12686724121050297,
-    "train_runtime": 1695.1033,
+    "total_flos": 1.3300619754508124e+18,
+    "train_loss": 0.05172654809675914,
+    "train_runtime": 17966.4012,
     "train_samples": 898,
-    "train_samples_per_second": 1.966,
-    "train_steps_per_second": 0.246
+    "train_samples_per_second": 1.855,
+    "train_steps_per_second": 0.232
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "total_flos": 1.3268723720781824e+17,
-    "train_loss": 0.12686724121050297,
-    "train_runtime": 1695.1033,
+    "total_flos": 1.3300619754508124e+18,
+    "train_loss": 0.05172654809675914,
+    "train_runtime": 17966.4012,
     "train_samples": 898,
-    "train_samples_per_second": 1.966,
-    "train_steps_per_second": 0.246
+    "train_samples_per_second": 1.855,
+    "train_steps_per_second": 0.232
 }