End of training

README.md

@@ -1,9 +1,15 @@
 ---
 base_model: Minbyul/llama2-7b-wo-live_qa-sft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: llama2-7b-dpo-full-sft-wo-live_qa
   results: []
@@ -14,17 +20,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama2-7b-dpo-full-sft-wo-live_qa
 
-This model is a fine-tuned version of [Minbyul/llama2-7b-wo-live_qa-sft](https://huggingface.co/Minbyul/llama2-7b-wo-live_qa-sft) on an unknown dataset.
+This model is a fine-tuned version of [Minbyul/llama2-7b-wo-live_qa-sft](https://huggingface.co/Minbyul/llama2-7b-wo-live_qa-sft) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5069
-- Rewards/chosen: -0.1530
-- Rewards/rejected: -0.4700
-- Rewards/accuracies: 0.875
-- Rewards/margins: 0.3170
-- Logps/rejected: -668.8459
-- Logps/chosen: -359.0135
-- Logits/rejected: -0.6288
-- Logits/chosen: -0.8302
+- Loss: 0.4649
+- Rewards/chosen: -0.1737
+- Rewards/rejected: -0.6052
+- Rewards/accuracies: 0.9167
+- Rewards/margins: 0.4315
+- Logps/rejected: -682.3730
+- Logps/chosen: -361.0932
+- Logits/rejected: -0.6401
+- Logits/chosen: -0.8344
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -0.8344432711601257,
+    "eval_logits/rejected": -0.6401407718658447,
+    "eval_logps/chosen": -361.0932312011719,
+    "eval_logps/rejected": -682.3729858398438,
+    "eval_loss": 0.46492835879325867,
+    "eval_rewards/accuracies": 0.9166666865348816,
+    "eval_rewards/chosen": -0.1737493872642517,
+    "eval_rewards/margins": 0.431477814912796,
+    "eval_rewards/rejected": -0.6052272319793701,
+    "eval_runtime": 12.8431,
+    "eval_samples": 96,
+    "eval_samples_per_second": 7.475,
+    "eval_steps_per_second": 0.234,
     "train_loss": 0.4338115497871681,
     "train_runtime": 2557.4167,
     "train_samples": 8634,

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": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": -0.8344432711601257,
+    "eval_logits/rejected": -0.6401407718658447,
+    "eval_logps/chosen": -361.0932312011719,
+    "eval_logps/rejected": -682.3729858398438,
+    "eval_loss": 0.46492835879325867,
+    "eval_rewards/accuracies": 0.9166666865348816,
+    "eval_rewards/chosen": -0.1737493872642517,
+    "eval_rewards/margins": 0.431477814912796,
+    "eval_rewards/rejected": -0.6052272319793701,
+    "eval_runtime": 12.8431,
+    "eval_samples": 96,
+    "eval_samples_per_second": 7.475,
+    "eval_steps_per_second": 0.234
+}