End of training

README.md

@@ -2,9 +2,15 @@
 license: apache-2.0
 base_model: alignment-handbook/zephyr-7b-sft-full
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: zephyr-7b-dpo-full-ultrabin-reward-scale-05
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-dpo-full-ultrabin-reward-scale-05
 
-This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on an unknown dataset.
+This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.5419
-- Rewards/chosen: -2.0657
-- Rewards/rejected: -3.5528
+- Rewards/chosen: -2.0652
+- Rewards/rejected: -3.5533
 - Rewards/accuracies: 0.7812
-- Rewards/margins: 1.4871
-- Logps/rejected: -617.9430
-- Logps/chosen: -469.2008
-- Logits/rejected: 3.0322
-- Logits/chosen: 2.1926
+- Rewards/margins: 1.4880
+- Logps/rejected: -617.9895
+- Logps/chosen: -469.1545
+- Logits/rejected: 3.0313
+- Logits/chosen: 2.1931
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": 2.193073272705078,
+    "eval_logits/rejected": 3.0313031673431396,
+    "eval_logps/chosen": -469.1545104980469,
+    "eval_logps/rejected": -617.989501953125,
+    "eval_loss": 0.5419281721115112,
+    "eval_rewards/accuracies": 0.78125,
+    "eval_rewards/chosen": -2.0652458667755127,
+    "eval_rewards/margins": 1.4880279302597046,
+    "eval_rewards/rejected": -3.5532736778259277,
+    "eval_runtime": 103.7243,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 19.282,
+    "eval_steps_per_second": 0.309,
     "total_flos": 0.0,
     "train_loss": 0.5571172007955767,
     "train_runtime": 12724.9946,

config.json

@@ -22,6 +22,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": 2.193073272705078,
+    "eval_logits/rejected": 3.0313031673431396,
+    "eval_logps/chosen": -469.1545104980469,
+    "eval_logps/rejected": -617.989501953125,
+    "eval_loss": 0.5419281721115112,
+    "eval_rewards/accuracies": 0.78125,
+    "eval_rewards/chosen": -2.0652458667755127,
+    "eval_rewards/margins": 1.4880279302597046,
+    "eval_rewards/rejected": -3.5532736778259277,
+    "eval_runtime": 103.7243,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 19.282,
+    "eval_steps_per_second": 0.309
+}