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 the None 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: 0.2103
-- Rewards/chosen: 4.5125
-- Rewards/rejected: -5.2808
-- Rewards/accuracies: 0.9258
-- Rewards/margins: 9.7933
-- Logps/rejected: -190.5157
-- Logps/chosen: -248.4084
-- Logits/rejected: -2.4965
-- Logits/chosen: -2.5123
+- Loss: 0.2141
+- Rewards/chosen: 4.2427
+- Rewards/rejected: -5.5362
+- Rewards/accuracies: 0.9141
+- Rewards/margins: 9.7789
+- Logps/rejected: -191.0266
+- Logps/chosen: -248.9478
+- Logits/rejected: -2.4882
+- Logits/chosen: -2.5045
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.5045135021209717,
+    "eval_logits/rejected": -2.4882149696350098,
+    "eval_logps/chosen": -248.94781494140625,
+    "eval_logps/rejected": -191.02655029296875,
+    "eval_loss": 0.21413667500019073,
+    "eval_rewards/accuracies": 0.9140625,
+    "eval_rewards/chosen": 4.242725372314453,
+    "eval_rewards/margins": 9.778908729553223,
+    "eval_rewards/rejected": -5.536184310913086,
+    "eval_runtime": 97.1606,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 20.584,
+    "eval_steps_per_second": 0.329,
     "train_loss": 0.4018711235732713,
     "train_runtime": 7633.33,
     "train_samples": 61135,

config.json

@@ -21,6 +21,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.38.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -1,15 +1,15 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -0.30772945284843445,
-    "eval_logits/rejected": 1.072676658630371,
-    "eval_logps/chosen": -474.955810546875,
-    "eval_logps/rejected": -812.6320190429688,
-    "eval_loss": 0.21300281584262848,
-    "eval_rewards/accuracies": 0.8984375,
-    "eval_rewards/chosen": -2.175225019454956,
-    "eval_rewards/margins": 4.151553153991699,
-    "eval_rewards/rejected": -6.326778411865234,
-    "eval_runtime": 97.1627,
+    "eval_logits/chosen": -2.5045135021209717,
+    "eval_logits/rejected": -2.4882149696350098,
+    "eval_logps/chosen": -248.94781494140625,
+    "eval_logps/rejected": -191.02655029296875,
+    "eval_loss": 0.21413667500019073,
+    "eval_rewards/accuracies": 0.9140625,
+    "eval_rewards/chosen": 4.242725372314453,
+    "eval_rewards/margins": 9.778908729553223,
+    "eval_rewards/rejected": -5.536184310913086,
+    "eval_runtime": 97.1606,
     "eval_samples": 2000,
     "eval_samples_per_second": 20.584,
     "eval_steps_per_second": 0.329

runs/May14_20-23-03_n136-100-194/events.out.tfevents.1715697794.n136-100-194.2851926.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86aa99cc611b52022209fc7630990b7dba32a69b2b745eed25d62dd4f9fc6486
+size 828