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: []
@@ -15,16 +18,16 @@ 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: 0.9216
-- Rewards/chosen: -0.0864
-- Rewards/rejected: -0.1686
-- Rewards/accuracies: 0.5573
-- Rewards/margins: 0.0822
-- Logps/rejected: -35.0354
-- Logps/chosen: -31.4553
-- Logits/rejected: -2.8133
+- Loss: 0.9071
+- Rewards/chosen: -0.0857
+- Rewards/rejected: -0.1842
+- Rewards/accuracies: 0.5714
+- Rewards/margins: 0.0984
+- Logps/rejected: -35.0665
+- Logps/chosen: -31.4539
+- Logits/rejected: -2.8132
 - Logits/chosen: -2.8155
 
 ## Model description

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.8154520988464355,
+    "eval_logits/rejected": -2.8132472038269043,
+    "eval_logps/chosen": -31.453943252563477,
+    "eval_logps/rejected": -35.06647872924805,
+    "eval_loss": 0.907057523727417,
+    "eval_rewards/accuracies": 0.5714285373687744,
+    "eval_rewards/chosen": -0.08574554324150085,
+    "eval_rewards/margins": 0.0984167754650116,
+    "eval_rewards/rejected": -0.18416231870651245,
+    "eval_runtime": 112.8077,
+    "eval_samples": 343,
+    "eval_samples_per_second": 3.041,
+    "eval_steps_per_second": 0.381,
     "train_loss": 0.694122297732861,
     "train_runtime": 2717.9333,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.8098936080932617,
-    "eval_logits/rejected": -2.807415723800659,
-    "eval_logps/chosen": -31.408082962036133,
-    "eval_logps/rejected": -34.975181579589844,
-    "eval_loss": 0.9007851481437683,
-    "eval_rewards/accuracies": 0.5859634280204773,
-    "eval_rewards/chosen": -0.1005057543516159,
-    "eval_rewards/margins": 0.12111683934926987,
-    "eval_rewards/rejected": -0.22162257134914398,
-    "eval_runtime": 112.7349,
+    "eval_logits/chosen": -2.8154520988464355,
+    "eval_logits/rejected": -2.8132472038269043,
+    "eval_logps/chosen": -31.453943252563477,
+    "eval_logps/rejected": -35.06647872924805,
+    "eval_loss": 0.907057523727417,
+    "eval_rewards/accuracies": 0.5714285373687744,
+    "eval_rewards/chosen": -0.08574554324150085,
+    "eval_rewards/margins": 0.0984167754650116,
+    "eval_rewards/rejected": -0.18416231870651245,
+    "eval_runtime": 112.8077,
     "eval_samples": 343,
-    "eval_samples_per_second": 3.043,
+    "eval_samples_per_second": 3.041,
     "eval_steps_per_second": 0.381
 }