End of training

README.md

@@ -2,9 +2,15 @@
 license: apache-2.0
 base_model: Minbyul/biomistral-7b-wo-live_qa-sft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: biomistral-7b-dpo-full-sft-wo-live_qa
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # biomistral-7b-dpo-full-sft-wo-live_qa
 
-This model is a fine-tuned version of [Minbyul/biomistral-7b-wo-live_qa-sft](https://huggingface.co/Minbyul/biomistral-7b-wo-live_qa-sft) on an unknown dataset.
+This model is a fine-tuned version of [Minbyul/biomistral-7b-wo-live_qa-sft](https://huggingface.co/Minbyul/biomistral-7b-wo-live_qa-sft) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4731
-- Rewards/chosen: -0.1249
-- Rewards/rejected: -0.7085
+- Loss: 0.4714
+- Rewards/chosen: -0.1308
+- Rewards/rejected: -0.7214
 - Rewards/accuracies: 0.75
-- Rewards/margins: 0.5837
-- Logps/rejected: -486.1749
-- Logps/chosen: -89.4136
-- Logits/rejected: -4.0476
-- Logits/chosen: -5.2050
+- Rewards/margins: 0.5906
+- Logps/rejected: -487.4595
+- Logps/chosen: -90.0080
+- Logits/rejected: -4.0457
+- Logits/chosen: -5.2055
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -5.205519676208496,
+    "eval_logits/rejected": -4.045749187469482,
+    "eval_logps/chosen": -90.00802612304688,
+    "eval_logps/rejected": -487.45953369140625,
+    "eval_loss": 0.47143450379371643,
+    "eval_rewards/accuracies": 0.75,
+    "eval_rewards/chosen": -0.13080169260501862,
+    "eval_rewards/margins": 0.5905601382255554,
+    "eval_rewards/rejected": -0.7213618159294128,
+    "eval_runtime": 5.4486,
+    "eval_samples": 4,
+    "eval_samples_per_second": 0.734,
+    "eval_steps_per_second": 0.184,
     "train_loss": 0.24888706256131657,
     "train_runtime": 2705.2939,
     "train_samples": 7808,

config.json

@@ -21,6 +21,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.39.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": -5.205519676208496,
+    "eval_logits/rejected": -4.045749187469482,
+    "eval_logps/chosen": -90.00802612304688,
+    "eval_logps/rejected": -487.45953369140625,
+    "eval_loss": 0.47143450379371643,
+    "eval_rewards/accuracies": 0.75,
+    "eval_rewards/chosen": -0.13080169260501862,
+    "eval_rewards/margins": 0.5905601382255554,
+    "eval_rewards/rejected": -0.7213618159294128,
+    "eval_runtime": 5.4486,
+    "eval_samples": 4,
+    "eval_samples_per_second": 0.734,
+    "eval_steps_per_second": 0.184
+}