End of training

README.md

@@ -1,9 +1,15 @@
 ---
 base_model: Minbyul/selfbiorag-7b-wo-kqa_silver_wogold-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_silver_wogold
   results: []
@@ -14,17 +20,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # selfbiorag-7b-dpo-full-sft-wo-kqa_silver_wogold
 
-This model is a fine-tuned version of [Minbyul/selfbiorag-7b-wo-kqa_silver_wogold-sft](https://huggingface.co/Minbyul/selfbiorag-7b-wo-kqa_silver_wogold-sft) on an unknown dataset.
+This model is a fine-tuned version of [Minbyul/selfbiorag-7b-wo-kqa_silver_wogold-sft](https://huggingface.co/Minbyul/selfbiorag-7b-wo-kqa_silver_wogold-sft) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3009
-- Rewards/chosen: -0.9868
-- Rewards/rejected: -4.1905
-- Rewards/accuracies: 0.8632
-- Rewards/margins: 3.2037
-- Logps/rejected: -1216.7521
-- Logps/chosen: -631.9589
-- Logits/rejected: -0.3156
-- Logits/chosen: -0.2396
+- Loss: 0.3027
+- Rewards/chosen: -1.2987
+- Rewards/rejected: -5.2535
+- Rewards/accuracies: 0.8697
+- Rewards/margins: 3.9549
+- Logps/rejected: -1323.0533
+- Logps/chosen: -663.1464
+- Logits/rejected: -0.3508
+- Logits/chosen: -0.2498
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -0.2497607320547104,
+    "eval_logits/rejected": -0.3508274257183075,
+    "eval_logps/chosen": -663.1464233398438,
+    "eval_logps/rejected": -1323.0533447265625,
+    "eval_loss": 0.30267053842544556,
+    "eval_rewards/accuracies": 0.8697368502616882,
+    "eval_rewards/chosen": -1.2986547946929932,
+    "eval_rewards/margins": 3.9548773765563965,
+    "eval_rewards/rejected": -5.2535319328308105,
+    "eval_runtime": 316.807,
+    "eval_samples": 3012,
+    "eval_samples_per_second": 9.507,
+    "eval_steps_per_second": 0.3,
     "train_loss": 0.24118649338243706,
     "train_runtime": 4650.475,
     "train_samples": 18141,

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": -0.2497607320547104,
+    "eval_logits/rejected": -0.3508274257183075,
+    "eval_logps/chosen": -663.1464233398438,
+    "eval_logps/rejected": -1323.0533447265625,
+    "eval_loss": 0.30267053842544556,
+    "eval_rewards/accuracies": 0.8697368502616882,
+    "eval_rewards/chosen": -1.2986547946929932,
+    "eval_rewards/margins": 3.9548773765563965,
+    "eval_rewards/rejected": -5.2535319328308105,
+    "eval_runtime": 316.807,
+    "eval_samples": 3012,
+    "eval_samples_per_second": 9.507,
+    "eval_steps_per_second": 0.3
+}