End of training

README.md

@@ -3,10 +3,16 @@ library_name: transformers
 license: apache-2.0
 base_model: alignment-handbook/zephyr-7b-sft-full
 tags:
+- alignment-handbook
+- trl
+- dpo
+- generated_from_trainer
 - trl
 - dpo
 - alignment-handbook
 - generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: zephyr-7b-align-scan-0.0-0.2-polynomial-3
   results: []
@@ -17,17 +23,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-align-scan-0.0-0.2-polynomial-3
 
-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:
-- Logits/chosen: -2.5251
-- Logits/rejected: -2.5078
-- Logps/chosen: -75.2992
-- Logps/rejected: -83.7654
-- Loss: 0.6467
+- Loss: 0.6601
+- Rewards/chosen: -1.0432
+- Rewards/rejected: -1.4686
 - Rewards/accuracies: 0.3313
-- Rewards/chosen: -0.1914
-- Rewards/margins: 0.4333
-- Rewards/rejected: -0.6247
+- Rewards/margins: 0.4254
+- Logps/rejected: -89.9617
+- Logps/chosen: -80.7659
+- Logits/rejected: -2.4878
+- Logits/chosen: -2.5051
 
 ## Model description
 

all_results.json

@@ -4,15 +4,15 @@
     "eval_logits/rejected": -2.487805128097534,
     "eval_logps/chosen": -80.76593017578125,
     "eval_logps/rejected": -89.9616928100586,
-    "eval_loss": 0.701084554195404,
+    "eval_loss": 0.6600826382637024,
     "eval_rewards/accuracies": 0.3313491940498352,
-    "eval_rewards/chosen": -1.4863682985305786,
-    "eval_rewards/margins": 0.6061023473739624,
-    "eval_rewards/rejected": -2.092470407485962,
-    "eval_runtime": 114.0061,
+    "eval_rewards/chosen": -1.0432164669036865,
+    "eval_rewards/margins": 0.4253966212272644,
+    "eval_rewards/rejected": -1.4686132669448853,
+    "eval_runtime": 114.8599,
     "eval_samples": 2000,
-    "eval_samples_per_second": 17.543,
-    "eval_steps_per_second": 0.553,
+    "eval_samples_per_second": 17.413,
+    "eval_steps_per_second": 0.548,
     "total_flos": 0.0,
     "train_loss": 0.0,
     "train_runtime": 0.0237,

config.json

@@ -22,6 +22,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -4,13 +4,13 @@
     "eval_logits/rejected": -2.487805128097534,
     "eval_logps/chosen": -80.76593017578125,
     "eval_logps/rejected": -89.9616928100586,
-    "eval_loss": 0.701084554195404,
+    "eval_loss": 0.6600826382637024,
     "eval_rewards/accuracies": 0.3313491940498352,
-    "eval_rewards/chosen": -1.4863682985305786,
-    "eval_rewards/margins": 0.6061023473739624,
-    "eval_rewards/rejected": -2.092470407485962,
-    "eval_runtime": 114.0061,
+    "eval_rewards/chosen": -1.0432164669036865,
+    "eval_rewards/margins": 0.4253966212272644,
+    "eval_rewards/rejected": -1.4686132669448853,
+    "eval_runtime": 114.8599,
     "eval_samples": 2000,
-    "eval_samples_per_second": 17.543,
-    "eval_steps_per_second": 0.553
+    "eval_samples_per_second": 17.413,
+    "eval_steps_per_second": 0.548
 }

runs/Sep24_14-18-45_qa-a100-004.crc.nd.edu/events.out.tfevents.1727202288.qa-a100-004.crc.nd.edu.1892028.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f046e2300711a8b87398ba1ea44191421ea3ac7b05615bffa2469951cd48dcc
+size 828