End of training

README.md

@@ -1,10 +1,13 @@
 ---
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
 base_model: NorLLM-AI/NorMistral-7B
+datasets:
+- hugodk-sch/aftonposten_title_prefs
 model-index:
 - name: norllm-ai-normistral-7b-align-scan
   results: []
@@ -15,17 +18,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # norllm-ai-normistral-7b-align-scan
 
-This model is a fine-tuned version of [NorLLM-AI/NorMistral-7B](https://huggingface.co/NorLLM-AI/NorMistral-7B) on the None dataset.
+This model is a fine-tuned version of [data/norllm-ai-normistral-7b-sft-qlora](https://huggingface.co/data/norllm-ai-normistral-7b-sft-qlora) on the hugodk-sch/aftonposten_title_prefs dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9202
-- Rewards/chosen: -0.0777
-- Rewards/rejected: -0.1629
-- Rewards/accuracies: 0.5656
-- Rewards/margins: 0.0852
-- Logps/rejected: -35.1054
-- Logps/chosen: -31.4766
-- Logits/rejected: -2.8221
-- Logits/chosen: -2.8242
+- Loss: 0.9105
+- Rewards/chosen: -0.0836
+- Rewards/rejected: -0.1783
+- Rewards/accuracies: 0.5515
+- Rewards/margins: 0.0947
+- Logps/rejected: -35.1438
+- Logps/chosen: -31.4914
+- Logits/rejected: -2.8211
+- Logits/chosen: -2.8232
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.8232436180114746,
+    "eval_logits/rejected": -2.8211076259613037,
+    "eval_logps/chosen": -31.491378784179688,
+    "eval_logps/rejected": -35.14382553100586,
+    "eval_loss": 0.9104546904563904,
+    "eval_rewards/accuracies": 0.5514950156211853,
+    "eval_rewards/chosen": -0.08357102423906326,
+    "eval_rewards/margins": 0.09469697624444962,
+    "eval_rewards/rejected": -0.17826800048351288,
+    "eval_runtime": 112.9595,
+    "eval_samples": 343,
+    "eval_samples_per_second": 3.036,
+    "eval_steps_per_second": 0.381,
     "train_loss": 0.7203964704042906,
     "train_runtime": 2720.5927,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.821291208267212,
-    "eval_logits/rejected": -2.818777084350586,
-    "eval_logps/chosen": -31.391199111938477,
-    "eval_logps/rejected": -34.970741271972656,
-    "eval_loss": 0.8988423347473145,
-    "eval_rewards/accuracies": 0.5772424936294556,
-    "eval_rewards/chosen": -0.07612106204032898,
-    "eval_rewards/margins": 0.11468725651502609,
-    "eval_rewards/rejected": -0.19080831110477448,
-    "eval_runtime": 112.7622,
+    "eval_logits/chosen": -2.8232436180114746,
+    "eval_logits/rejected": -2.8211076259613037,
+    "eval_logps/chosen": -31.491378784179688,
+    "eval_logps/rejected": -35.14382553100586,
+    "eval_loss": 0.9104546904563904,
+    "eval_rewards/accuracies": 0.5514950156211853,
+    "eval_rewards/chosen": -0.08357102423906326,
+    "eval_rewards/margins": 0.09469697624444962,
+    "eval_rewards/rejected": -0.17826800048351288,
+    "eval_runtime": 112.9595,
     "eval_samples": 343,
-    "eval_samples_per_second": 3.042,
+    "eval_samples_per_second": 3.036,
     "eval_steps_per_second": 0.381
 }