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-live_qa-ep3
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # meditron-7b-dpo-full-wo-live_qa-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:
 - Loss: 0.5356
-- Rewards/chosen: -0.2875
-- Rewards/rejected: -0.7751
-- Rewards/accuracies: 0.7019
-- Rewards/margins: 0.4876
-- Logps/rejected: -1205.2603
-- Logps/chosen: -986.1399
-- Logits/rejected: -0.8883
-- Logits/chosen: -0.8904
+- Rewards/chosen: -0.2871
+- Rewards/rejected: -0.7760
+- Rewards/accuracies: 0.6923
+- Rewards/margins: 0.4889
+- Logps/rejected: -1205.3544
+- Logps/chosen: -986.1032
+- Logits/rejected: -0.8878
+- Logits/chosen: -0.8900
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -0.8900327086448669,
+    "eval_logits/rejected": -0.8877933025360107,
+    "eval_logps/chosen": -986.1032104492188,
+    "eval_logps/rejected": -1205.3543701171875,
+    "eval_loss": 0.5355749130249023,
+    "eval_rewards/accuracies": 0.692307710647583,
+    "eval_rewards/chosen": -0.28710055351257324,
+    "eval_rewards/margins": 0.4889349639415741,
+    "eval_rewards/rejected": -0.7760356068611145,
+    "eval_runtime": 47.8688,
+    "eval_samples": 416,
+    "eval_samples_per_second": 8.69,
+    "eval_steps_per_second": 0.272,
     "train_loss": 0.539328022330415,
     "train_runtime": 3313.1548,
     "train_samples": 13051,

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": -0.8900327086448669,
+    "eval_logits/rejected": -0.8877933025360107,
+    "eval_logps/chosen": -986.1032104492188,
+    "eval_logps/rejected": -1205.3543701171875,
+    "eval_loss": 0.5355749130249023,
+    "eval_rewards/accuracies": 0.692307710647583,
+    "eval_rewards/chosen": -0.28710055351257324,
+    "eval_rewards/margins": 0.4889349639415741,
+    "eval_rewards/rejected": -0.7760356068611145,
+    "eval_runtime": 47.8688,
+    "eval_samples": 416,
+    "eval_samples_per_second": 8.69,
+    "eval_steps_per_second": 0.272
+}