End of training

README.md

@@ -1,9 +1,17 @@
 ---
 library_name: transformers
+license: apache-2.0
+base_model: Qwen/Qwen2.5-1.5B-Instruct
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- YuchenLi01/MATH_Qwen2.5-1.5BInstruct_DPO_MoreUniqueResponseNoGTv3pair1
 model-index:
 - name: genv3pair1NoGT_1.5B_cdpo_ebs32_lr5e-07_beta0.1_epoch8.0_42
   results: []
@@ -14,16 +22,16 @@ should probably proofread and complete it, then remove this comment. -->
 
 # genv3pair1NoGT_1.5B_cdpo_ebs32_lr5e-07_beta0.1_epoch8.0_42
 
-This model was trained from scratch on an unknown dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct) on the YuchenLi01/MATH_Qwen2.5-1.5BInstruct_DPO_MoreUniqueResponseNoGTv3pair1 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3653
-- Rewards/chosen: 1.1076
+- Loss: 0.3658
+- Rewards/chosen: 1.1075
 - Rewards/rejected: 0.0
-- Rewards/accuracies: 0.9250
-- Rewards/margins: 1.1076
-- Logps/rejected: -33.0177
-- Logps/chosen: -18.9939
-- Logits/rejected: -3.5369
+- Rewards/accuracies: 0.9000
+- Rewards/margins: 1.1075
+- Logps/rejected: -33.0278
+- Logps/chosen: -18.9948
+- Logits/rejected: -3.5375
 - Logits/chosen: -3.4431
 
 ## Model description

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 8.0,
+    "eval_logits/chosen": -3.443093776702881,
+    "eval_logits/rejected": -3.537508487701416,
+    "eval_logps/chosen": -18.99481201171875,
+    "eval_logps/rejected": -33.02782440185547,
+    "eval_loss": 0.36576059460639954,
+    "eval_rewards/accuracies": 0.8999999761581421,
+    "eval_rewards/chosen": 1.1075230836868286,
+    "eval_rewards/margins": 1.1075230836868286,
+    "eval_rewards/rejected": 0.0,
+    "eval_runtime": 24.7361,
+    "eval_samples": 310,
+    "eval_samples_per_second": 12.532,
+    "eval_steps_per_second": 0.404,
     "total_flos": 0.0,
     "train_loss": 0.2642531191732464,
     "train_runtime": 10430.2154,

config.json

@@ -23,7 +23,7 @@
   "tie_word_embeddings": true,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.45.2",
-  "use_cache": false,
+  "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151936
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 8.0,
+    "eval_logits/chosen": -3.443093776702881,
+    "eval_logits/rejected": -3.537508487701416,
+    "eval_logps/chosen": -18.99481201171875,
+    "eval_logps/rejected": -33.02782440185547,
+    "eval_loss": 0.36576059460639954,
+    "eval_rewards/accuracies": 0.8999999761581421,
+    "eval_rewards/chosen": 1.1075230836868286,
+    "eval_rewards/margins": 1.1075230836868286,
+    "eval_rewards/rejected": 0.0,
+    "eval_runtime": 24.7361,
+    "eval_samples": 310,
+    "eval_samples_per_second": 12.532,
+    "eval_steps_per_second": 0.404
+}