End of training

README.md

@@ -2,13 +2,13 @@
 license: apache-2.0
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - sft
-- alignment-handbook
 - generated_from_trainer
 base_model: mistralai/Mistral-7B-Instruct-v0.2
 datasets:
-- generator
+- preference-data
 model-index:
 - name: feedback_p0.1_seed42_level2_rare
   results: []
@@ -19,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # feedback_p0.1_seed42_level2_rare
 
-This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the generator dataset.
+This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the preference-data dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.8673
 

all_results.json

@@ -1,10 +1,10 @@
 {
     "epoch": 0.9988193624557261,
     "eval_loss": 0.8673050403594971,
-    "eval_runtime": 1.0099,
+    "eval_runtime": 2.0994,
     "eval_samples": 10,
-    "eval_samples_per_second": 1.98,
-    "eval_steps_per_second": 0.99,
+    "eval_samples_per_second": 0.953,
+    "eval_steps_per_second": 0.476,
     "total_flos": 1701364321222656.0,
     "train_loss": 0.0,
     "train_runtime": 0.0113,

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 0.9988193624557261,
     "eval_loss": 0.8673050403594971,
-    "eval_runtime": 1.0099,
+    "eval_runtime": 2.0994,
     "eval_samples": 10,
-    "eval_samples_per_second": 1.98,
-    "eval_steps_per_second": 0.99
+    "eval_samples_per_second": 0.953,
+    "eval_steps_per_second": 0.476
 }

runs/Sep02_15-55-36_COE-CS-sv004/events.out.tfevents.1725292576.COE-CS-sv004.2871811.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc3de5c77c8d99941fa4ad7f40572272dfd62dc6e274f51c2f2d0643ff9b3310
+size 359