End of training

README.md

@@ -3,9 +3,15 @@ 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
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: zephyr-7b-align-scan-6e-07-0.53-polynomial-2.0
   results: []
@@ -16,17 +22,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-align-scan-6e-07-0.53-polynomial-2.0
 
-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: 0.8639
-- Rewards/chosen: 0.6923
-- Rewards/rejected: -0.5434
+- Loss: 0.8595
+- Rewards/chosen: 1.8502
+- Rewards/rejected: 0.6070
 - Rewards/accuracies: 0.3393
-- Rewards/margins: 1.2357
-- Logps/rejected: -82.1536
-- Logps/chosen: -73.1851
-- Logits/rejected: -2.6910
-- Logits/chosen: -2.7068
+- Rewards/margins: 1.2433
+- Logps/rejected: -79.9832
+- Logps/chosen: -71.0003
+- Logits/rejected: -2.6929
+- Logits/chosen: -2.7082
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.0,
+    "eval_logits/chosen": -2.708155632019043,
+    "eval_logits/rejected": -2.692856550216675,
+    "eval_logps/chosen": -71.00025177001953,
+    "eval_logps/rejected": -79.98316192626953,
+    "eval_loss": 0.8595444560050964,
+    "eval_rewards/accuracies": 0.3392857015132904,
+    "eval_rewards/chosen": 1.850224494934082,
+    "eval_rewards/margins": 1.243270993232727,
+    "eval_rewards/rejected": 0.6069534420967102,
+    "eval_runtime": 113.8695,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 17.564,
+    "eval_steps_per_second": 0.553,
     "total_flos": 0.0,
     "train_loss": 0.5567006942287139,
     "train_runtime": 6469.9642,

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

@@ -0,0 +1,16 @@
+{
+    "epoch": 2.0,
+    "eval_logits/chosen": -2.708155632019043,
+    "eval_logits/rejected": -2.692856550216675,
+    "eval_logps/chosen": -71.00025177001953,
+    "eval_logps/rejected": -79.98316192626953,
+    "eval_loss": 0.8595444560050964,
+    "eval_rewards/accuracies": 0.3392857015132904,
+    "eval_rewards/chosen": 1.850224494934082,
+    "eval_rewards/margins": 1.243270993232727,
+    "eval_rewards/rejected": 0.6069534420967102,
+    "eval_runtime": 113.8695,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 17.564,
+    "eval_steps_per_second": 0.553
+}

runs/Sep27_22-04-39_qa-a100-005.crc.nd.edu/events.out.tfevents.1727496017.qa-a100-005.crc.nd.edu.4182113.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:230fe578fca9ae53a5147c4602547a80a3c729b38609119a02672af8da789060
+size 828