End of training

README.md

@@ -1,9 +1,15 @@
 ---
 base_model: Minbyul/selfbiorag-7b-wo-kqa_golden-sft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: selfbiorag-7b-dpo-full-sft-wo-kqa_golden
   results: []
@@ -14,17 +20,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # selfbiorag-7b-dpo-full-sft-wo-kqa_golden
 
-This model is a fine-tuned version of [Minbyul/selfbiorag-7b-wo-kqa_golden-sft](https://huggingface.co/Minbyul/selfbiorag-7b-wo-kqa_golden-sft) on an unknown dataset.
+This model is a fine-tuned version of [Minbyul/selfbiorag-7b-wo-kqa_golden-sft](https://huggingface.co/Minbyul/selfbiorag-7b-wo-kqa_golden-sft) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2413
-- Rewards/chosen: -1.0961
-- Rewards/rejected: -13.1849
+- Loss: 0.2401
+- Rewards/chosen: -1.0928
+- Rewards/rejected: -13.1704
 - Rewards/accuracies: 0.8942
-- Rewards/margins: 12.0888
-- Logps/rejected: -2033.0142
-- Logps/chosen: -567.6829
-- Logits/rejected: -0.2111
-- Logits/chosen: -0.3569
+- Rewards/margins: 12.0777
+- Logps/rejected: -2031.5652
+- Logps/chosen: -567.3484
+- Logits/rejected: -0.2100
+- Logits/chosen: -0.3532
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -0.3531719446182251,
+    "eval_logits/rejected": -0.21002119779586792,
+    "eval_logps/chosen": -567.3484497070312,
+    "eval_logps/rejected": -2031.565185546875,
+    "eval_loss": 0.240110382437706,
+    "eval_rewards/accuracies": 0.8942307829856873,
+    "eval_rewards/chosen": -1.0927813053131104,
+    "eval_rewards/margins": 12.077652931213379,
+    "eval_rewards/rejected": -13.17043399810791,
+    "eval_runtime": 41.3566,
+    "eval_samples": 396,
+    "eval_samples_per_second": 9.575,
+    "eval_steps_per_second": 0.314,
     "train_loss": 0.23018195102980107,
     "train_runtime": 4792.5755,
     "train_samples": 20757,

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": 32016
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": -0.3531719446182251,
+    "eval_logits/rejected": -0.21002119779586792,
+    "eval_logps/chosen": -567.3484497070312,
+    "eval_logps/rejected": -2031.565185546875,
+    "eval_loss": 0.240110382437706,
+    "eval_rewards/accuracies": 0.8942307829856873,
+    "eval_rewards/chosen": -1.0927813053131104,
+    "eval_rewards/margins": 12.077652931213379,
+    "eval_rewards/rejected": -13.17043399810791,
+    "eval_runtime": 41.3566,
+    "eval_samples": 396,
+    "eval_samples_per_second": 9.575,
+    "eval_steps_per_second": 0.314
+}