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: 0.4906
-- Rewards/chosen: -1.5814
-- Rewards/rejected: -2.6280
+- Loss: 0.4904
+- Rewards/chosen: -1.5809
+- Rewards/rejected: -2.6285
 - Rewards/accuracies: 0.7698
-- Rewards/margins: 1.0467
-- Logps/rejected: -523.0108
-- Logps/chosen: -440.1050
-- Logits/rejected: 0.7152
-- Logits/chosen: -0.3568
+- Rewards/margins: 1.0476
+- Logps/rejected: -523.0543
+- Logps/chosen: -440.0590
+- Logits/rejected: 0.7096
+- Logits/chosen: -0.3600
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.9994767137624281,
+    "eval_logits/chosen": -0.35997089743614197,
+    "eval_logits/rejected": 0.7096247673034668,
+    "eval_logps/chosen": -440.0589599609375,
+    "eval_logps/rejected": -523.0543212890625,
+    "eval_loss": 0.4903615117073059,
+    "eval_rewards/accuracies": 0.7698412537574768,
+    "eval_rewards/chosen": -1.5809098482131958,
+    "eval_rewards/margins": 1.0475648641586304,
+    "eval_rewards/rejected": -2.6284749507904053,
+    "eval_runtime": 176.963,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 11.302,
+    "eval_steps_per_second": 0.356,
     "total_flos": 0.0,
     "train_loss": 0.5247097397349891,
     "train_runtime": 18219.8315,

config.json

@@ -21,6 +21,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.40.1",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 0.9994767137624281,
+    "eval_logits/chosen": -0.35997089743614197,
+    "eval_logits/rejected": 0.7096247673034668,
+    "eval_logps/chosen": -440.0589599609375,
+    "eval_logps/rejected": -523.0543212890625,
+    "eval_loss": 0.4903615117073059,
+    "eval_rewards/accuracies": 0.7698412537574768,
+    "eval_rewards/chosen": -1.5809098482131958,
+    "eval_rewards/margins": 1.0475648641586304,
+    "eval_rewards/rejected": -2.6284749507904053,
+    "eval_runtime": 176.963,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 11.302,
+    "eval_steps_per_second": 0.356
+}

runs/Apr25_13-59-39_dsw-3393-565b8f59b6-l5p7k/events.out.tfevents.1714073194.dsw-3393-565b8f59b6-l5p7k.12328.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d05a07fb2bfe27d97c0c0586ad1c55aac5cef3dfd7ebca853849c98d0b12a4b
+size 780