End of training

README.md

@@ -2,9 +2,15 @@
 license: apache-2.0
 base_model: Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: biomistral-7b-wo-kqa_golden-iter-dpo-step2
   results: []
@@ -15,7 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # biomistral-7b-wo-kqa_golden-iter-dpo-step2
 
-This model is a fine-tuned version of [Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1](https://huggingface.co/Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1) on an unknown dataset.
+This model is a fine-tuned version of [Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1](https://huggingface.co/Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1) on the HuggingFaceH4/ultrafeedback_binarized dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.6909
+- Rewards/chosen: 0.0063
+- Rewards/rejected: 0.0057
+- Rewards/accuracies: 0.5625
+- Rewards/margins: 0.0006
+- Logps/rejected: -193.8717
+- Logps/chosen: -168.4928
+- Logits/rejected: -2.2060
+- Logits/chosen: -2.9391
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.9391283988952637,
+    "eval_logits/rejected": -2.2059857845306396,
+    "eval_logps/chosen": -168.49281311035156,
+    "eval_logps/rejected": -193.87171936035156,
+    "eval_loss": 0.6908513903617859,
+    "eval_rewards/accuracies": 0.5625,
+    "eval_rewards/chosen": 0.00629099877551198,
+    "eval_rewards/margins": 0.0006325023132376373,
+    "eval_rewards/rejected": 0.0056584966368973255,
+    "eval_runtime": 20.5327,
+    "eval_samples": 231,
+    "eval_samples_per_second": 11.25,
+    "eval_steps_per_second": 0.39,
     "train_loss": 0.680875283606509,
     "train_runtime": 1044.3353,
     "train_samples": 6013,

config.json

@@ -21,6 +21,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.38.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": -2.9391283988952637,
+    "eval_logits/rejected": -2.2059857845306396,
+    "eval_logps/chosen": -168.49281311035156,
+    "eval_logps/rejected": -193.87171936035156,
+    "eval_loss": 0.6908513903617859,
+    "eval_rewards/accuracies": 0.5625,
+    "eval_rewards/chosen": 0.00629099877551198,
+    "eval_rewards/margins": 0.0006325023132376373,
+    "eval_rewards/rejected": 0.0056584966368973255,
+    "eval_runtime": 20.5327,
+    "eval_samples": 231,
+    "eval_samples_per_second": 11.25,
+    "eval_steps_per_second": 0.39
+}