End of training

README.md

@@ -2,9 +2,15 @@
 license: llama2
 base_model: epfl-llm/meditron-7b
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: meditron-7b-dpo-full-wo-kqa_golden-ep3
   results: []
@@ -15,7 +21,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # meditron-7b-dpo-full-wo-kqa_golden-ep3
 
-This model is a fine-tuned version of [epfl-llm/meditron-7b](https://huggingface.co/epfl-llm/meditron-7b) on an unknown dataset.
+This model is a fine-tuned version of [epfl-llm/meditron-7b](https://huggingface.co/epfl-llm/meditron-7b) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.4459
 - Rewards/chosen: -0.4566

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -0.8309964537620544,
+    "eval_logits/rejected": -0.9202789068222046,
+    "eval_logps/chosen": -859.0581665039062,
+    "eval_logps/rejected": -1444.6895751953125,
+    "eval_loss": 0.4459438920021057,
+    "eval_rewards/accuracies": 0.8068181872367859,
+    "eval_rewards/chosen": -0.4565673768520355,
+    "eval_rewards/margins": 0.9446592926979065,
+    "eval_rewards/rejected": -1.4012266397476196,
+    "eval_runtime": 80.5227,
+    "eval_samples": 676,
+    "eval_samples_per_second": 8.395,
+    "eval_steps_per_second": 0.273,
     "train_loss": 0.5449820566177368,
     "train_runtime": 3295.9797,
     "train_samples": 12791,

config.json

@@ -23,6 +23,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.39.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32017
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": -0.8309964537620544,
+    "eval_logits/rejected": -0.9202789068222046,
+    "eval_logps/chosen": -859.0581665039062,
+    "eval_logps/rejected": -1444.6895751953125,
+    "eval_loss": 0.4459438920021057,
+    "eval_rewards/accuracies": 0.8068181872367859,
+    "eval_rewards/chosen": -0.4565673768520355,
+    "eval_rewards/margins": 0.9446592926979065,
+    "eval_rewards/rejected": -1.4012266397476196,
+    "eval_runtime": 80.5227,
+    "eval_samples": 676,
+    "eval_samples_per_second": 8.395,
+    "eval_steps_per_second": 0.273
+}