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
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-dpo-full
 
-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: 0.5023
-- Rewards/chosen: -1.0875
-- Rewards/rejected: -2.0687
-- Rewards/accuracies: 0.7695
-- Rewards/margins: 0.9812
-- Logps/rejected: -469.5586
-- Logps/chosen: -371.3432
-- Logits/rejected: -0.3666
-- Logits/chosen: -1.1044
+- Loss: 0.5018
+- Rewards/chosen: -1.1027
+- Rewards/rejected: -2.0673
+- Rewards/accuracies: 0.7656
+- Rewards/margins: 0.9646
+- Logps/rejected: -469.4237
+- Logps/chosen: -372.8636
+- Logits/rejected: -0.4162
+- Logits/chosen: -1.1470
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -1.1469733715057373,
+    "eval_logits/rejected": -0.4162125587463379,
+    "eval_logps/chosen": -372.8635559082031,
+    "eval_logps/rejected": -469.4237060546875,
+    "eval_loss": 0.5017749071121216,
+    "eval_rewards/accuracies": 0.765625,
+    "eval_rewards/chosen": -1.1026902198791504,
+    "eval_rewards/margins": 0.9646469354629517,
+    "eval_rewards/rejected": -2.0673370361328125,
+    "eval_runtime": 91.4963,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 21.859,
+    "eval_steps_per_second": 0.35,
     "train_loss": 0.5374924991918907,
     "train_runtime": 7413.7576,
     "train_samples": 61134,

config.json

@@ -21,6 +21,6 @@
   "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": -1.1469733715057373,
+    "eval_logits/rejected": -0.4162125587463379,
+    "eval_logps/chosen": -372.8635559082031,
+    "eval_logps/rejected": -469.4237060546875,
+    "eval_loss": 0.5017749071121216,
+    "eval_rewards/accuracies": 0.765625,
+    "eval_rewards/chosen": -1.1026902198791504,
+    "eval_rewards/margins": 0.9646469354629517,
+    "eval_rewards/rejected": -2.0673370361328125,
+    "eval_runtime": 91.4963,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 21.859,
+    "eval_steps_per_second": 0.35
+}