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-2e-07-0.5-linear-2.0
   results: []
@@ -16,17 +22,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-align-scan-2e-07-0.5-linear-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: 0.6465
-- Rewards/chosen: 0.3072
-- Rewards/rejected: -0.2969
+- Loss: 0.6464
+- Rewards/chosen: 0.4833
+- Rewards/rejected: -0.1353
 - Rewards/accuracies: 0.3710
-- Rewards/margins: 0.6040
-- Logps/rejected: -81.7221
-- Logps/chosen: -73.8769
-- Logits/rejected: -2.5543
-- Logits/chosen: -2.5709
+- Rewards/margins: 0.6186
+- Logps/rejected: -81.3989
+- Logps/chosen: -73.5245
+- Logits/rejected: -2.5552
+- Logits/chosen: -2.5716
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.0,
+    "eval_logits/chosen": -2.5716235637664795,
+    "eval_logits/rejected": -2.555190324783325,
+    "eval_logps/chosen": -73.5245361328125,
+    "eval_logps/rejected": -81.39889526367188,
+    "eval_loss": 0.6463928818702698,
+    "eval_rewards/accuracies": 0.3710317313671112,
+    "eval_rewards/chosen": 0.48334550857543945,
+    "eval_rewards/margins": 0.6186185479164124,
+    "eval_rewards/rejected": -0.13527308404445648,
+    "eval_runtime": 113.7854,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 17.577,
+    "eval_steps_per_second": 0.554,
     "total_flos": 0.0,
     "train_loss": 0.5264799162901237,
     "train_runtime": 6499.6728,

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": -2.5716235637664795,
+    "eval_logits/rejected": -2.555190324783325,
+    "eval_logps/chosen": -73.5245361328125,
+    "eval_logps/rejected": -81.39889526367188,
+    "eval_loss": 0.6463928818702698,
+    "eval_rewards/accuracies": 0.3710317313671112,
+    "eval_rewards/chosen": 0.48334550857543945,
+    "eval_rewards/margins": 0.6186185479164124,
+    "eval_rewards/rejected": -0.13527308404445648,
+    "eval_runtime": 113.7854,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 17.577,
+    "eval_steps_per_second": 0.554
+}

runs/Sep27_17-55-18_qa-a100-005.crc.nd.edu/events.out.tfevents.1727481086.qa-a100-005.crc.nd.edu.4143722.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88e575093fd126e31e019c637bf8f93b7b18c47e8fb17a9311b8dba068f89092
+size 828