End of training

README.md

@@ -2,9 +2,15 @@
 license: llama2
 base_model: epfl-llm/meditron-7b
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: meditron-7b-dpo-full-wo-kqa_silver_wogold-ep3
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # meditron-7b-dpo-full-wo-kqa_silver_wogold-ep3
 
-This model is a fine-tuned version of [epfl-llm/meditron-7b](https://huggingface.co/epfl-llm/meditron-7b) on an unknown dataset.
+This model is a fine-tuned version of [epfl-llm/meditron-7b](https://huggingface.co/epfl-llm/meditron-7b) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Logits/chosen: -0.6676
-- Logits/rejected: -0.8939
-- Logps/chosen: -826.0934
-- Logps/rejected: -1433.1564
-- Loss: 0.6219
-- Rewards/accuracies: 0.7459
-- Rewards/chosen: -0.0445
-- Rewards/margins: 0.1998
-- Rewards/rejected: -0.2443
+- Loss: 0.5793
+- Rewards/chosen: -0.1323
+- Rewards/rejected: -0.4764
+- Rewards/accuracies: 0.7717
+- Rewards/margins: 0.3440
+- Logps/rejected: -1456.3621
+- Logps/chosen: -834.8738
+- Logits/rejected: -0.9041
+- Logits/chosen: -0.7062
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -0.7062377333641052,
+    "eval_logits/rejected": -0.9040690064430237,
+    "eval_logps/chosen": -834.8738403320312,
+    "eval_logps/rejected": -1456.362060546875,
+    "eval_loss": 0.5792528390884399,
+    "eval_rewards/accuracies": 0.77173912525177,
+    "eval_rewards/chosen": -0.1323184221982956,
+    "eval_rewards/margins": 0.34403395652770996,
+    "eval_rewards/rejected": -0.47635239362716675,
+    "eval_runtime": 330.6189,
+    "eval_samples": 2927,
+    "eval_samples_per_second": 8.853,
+    "eval_steps_per_second": 0.278,
     "train_loss": 0.20238888480446554,
     "train_runtime": 919.6394,
     "train_samples": 10540,

config.json

@@ -23,6 +23,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.39.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32017
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": -0.7062377333641052,
+    "eval_logits/rejected": -0.9040690064430237,
+    "eval_logps/chosen": -834.8738403320312,
+    "eval_logps/rejected": -1456.362060546875,
+    "eval_loss": 0.5792528390884399,
+    "eval_rewards/accuracies": 0.77173912525177,
+    "eval_rewards/chosen": -0.1323184221982956,
+    "eval_rewards/margins": 0.34403395652770996,
+    "eval_rewards/rejected": -0.47635239362716675,
+    "eval_runtime": 330.6189,
+    "eval_samples": 2927,
+    "eval_samples_per_second": 8.853,
+    "eval_steps_per_second": 0.278
+}