End of training

README.md

@@ -1,9 +1,15 @@
 ---
 base_model: dmis-lab/selfbiorag_7b
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: selfbiorag-7b-dpo-full-wo-healthsearch_qa-ep3
   results: []
@@ -14,17 +20,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # selfbiorag-7b-dpo-full-wo-healthsearch_qa-ep3
 
-This model is a fine-tuned version of [dmis-lab/selfbiorag_7b](https://huggingface.co/dmis-lab/selfbiorag_7b) on an unknown dataset.
+This model is a fine-tuned version of [dmis-lab/selfbiorag_7b](https://huggingface.co/dmis-lab/selfbiorag_7b) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Logits/chosen: -1.6968
-- Logits/rejected: -1.6723
-- Logps/chosen: -158.8701
-- Logps/rejected: -170.2428
-- Loss: 0.6691
-- Rewards/accuracies: 0.6941
-- Rewards/chosen: 0.0706
-- Rewards/margins: 0.0503
-- Rewards/rejected: 0.0202
+- Loss: 0.6669
+- Rewards/chosen: 0.0740
+- Rewards/rejected: 0.0190
+- Rewards/accuracies: 0.6977
+- Rewards/margins: 0.0550
+- Logps/rejected: -170.3629
+- Logps/chosen: -158.5215
+- Logits/rejected: -1.6515
+- Logits/chosen: -1.6773
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -1.6772518157958984,
+    "eval_logits/rejected": -1.6514736413955688,
+    "eval_logps/chosen": -158.52154541015625,
+    "eval_logps/rejected": -170.3629150390625,
+    "eval_loss": 0.6668696403503418,
+    "eval_rewards/accuracies": 0.6976549625396729,
+    "eval_rewards/chosen": 0.07404651492834091,
+    "eval_rewards/margins": 0.05502632260322571,
+    "eval_rewards/rejected": 0.019020190462470055,
+    "eval_runtime": 1787.1604,
+    "eval_samples": 19103,
+    "eval_samples_per_second": 10.689,
+    "eval_steps_per_second": 0.334,
     "train_loss": 0.2519006322069866,
     "train_runtime": 787.0698,
     "train_samples": 10477,

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

@@ -1,16 +1,16 @@
 {
-    "epoch": 3.0,
-    "eval_logits/chosen": -1.680073618888855,
-    "eval_logits/rejected": -1.7620524168014526,
-    "eval_logps/chosen": -349.61962890625,
-    "eval_logps/rejected": -92.87176513671875,
-    "eval_loss": 0.5131469964981079,
-    "eval_rewards/accuracies": 1.0,
-    "eval_rewards/chosen": 0.3873787820339203,
-    "eval_rewards/margins": 0.40528222918510437,
-    "eval_rewards/rejected": -0.017903532832860947,
-    "eval_runtime": 67.3461,
-    "eval_samples": 3077,
-    "eval_samples_per_second": 45.689,
-    "eval_steps_per_second": 0.728
+    "epoch": 1.0,
+    "eval_logits/chosen": -1.6772518157958984,
+    "eval_logits/rejected": -1.6514736413955688,
+    "eval_logps/chosen": -158.52154541015625,
+    "eval_logps/rejected": -170.3629150390625,
+    "eval_loss": 0.6668696403503418,
+    "eval_rewards/accuracies": 0.6976549625396729,
+    "eval_rewards/chosen": 0.07404651492834091,
+    "eval_rewards/margins": 0.05502632260322571,
+    "eval_rewards/rejected": 0.019020190462470055,
+    "eval_runtime": 1787.1604,
+    "eval_samples": 19103,
+    "eval_samples_per_second": 10.689,
+    "eval_steps_per_second": 0.334
 }