End of training

README.md

@@ -1,11 +1,13 @@
 ---
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
-- alignment-handbook
 - generated_from_trainer
 base_model: NbAiLab/nb-gpt-j-6B-v2
+datasets:
+- hugodk-sch/aftonposten_title_prefs
 model-index:
 - name: aftonposten-6b-align-scan
   results: []
@@ -16,17 +18,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # aftonposten-6b-align-scan
 
-This model is a fine-tuned version of [NbAiLab/nb-gpt-j-6B-v2](https://huggingface.co/NbAiLab/nb-gpt-j-6B-v2) on an unknown dataset.
+This model is a fine-tuned version of [data/ap-gpt-j-6b-sft-qlora-04-08](https://huggingface.co/data/ap-gpt-j-6b-sft-qlora-04-08) on the hugodk-sch/aftonposten_title_prefs dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2463.1057
+- Loss: 2461.9446
 - Rewards/chosen: -0.0198
-- Rewards/rejected: -0.0244
-- Rewards/accuracies: 0.5656
-- Rewards/margins: 0.0046
-- Logps/rejected: -39.9582
-- Logps/chosen: -36.0192
-- Logits/rejected: -1.8605
-- Logits/chosen: -1.8648
+- Rewards/rejected: -0.0245
+- Rewards/accuracies: 0.5569
+- Rewards/margins: 0.0047
+- Logps/rejected: -39.9694
+- Logps/chosen: -36.0193
+- Logits/rejected: -1.8604
+- Logits/chosen: -1.8647
 
 ## Model description
 

all_results.json

@@ -1,18 +1,18 @@
 {
     "epoch": 4.0,
-    "eval_logits/chosen": -2.226093053817749,
-    "eval_logits/rejected": -2.221263885498047,
-    "eval_logps/chosen": -34.05622100830078,
-    "eval_logps/rejected": -37.58139419555664,
-    "eval_loss": 2495.922119140625,
-    "eval_rewards/accuracies": 0.529900312423706,
-    "eval_rewards/chosen": -0.00021668746194336563,
-    "eval_rewards/margins": 0.00043102249037474394,
-    "eval_rewards/rejected": -0.0006477099377661943,
-    "eval_runtime": 145.39,
+    "eval_logits/chosen": -1.864721417427063,
+    "eval_logits/rejected": -1.8603547811508179,
+    "eval_logps/chosen": -36.019264221191406,
+    "eval_logps/rejected": -39.96937561035156,
+    "eval_loss": 2461.944580078125,
+    "eval_rewards/accuracies": 0.5568937063217163,
+    "eval_rewards/chosen": -0.019847100600600243,
+    "eval_rewards/margins": 0.0046804738231003284,
+    "eval_rewards/rejected": -0.024527573958039284,
+    "eval_runtime": 145.7077,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.359,
-    "eval_steps_per_second": 0.296,
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295,
     "train_loss": 1756.4536297686689,
     "train_runtime": 10797.265,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 1.0,
-    "eval_logits/chosen": -2.226093053817749,
-    "eval_logits/rejected": -2.221263885498047,
-    "eval_logps/chosen": -34.05622100830078,
-    "eval_logps/rejected": -37.58139419555664,
-    "eval_loss": 2495.922119140625,
-    "eval_rewards/accuracies": 0.529900312423706,
-    "eval_rewards/chosen": -0.00021668746194336563,
-    "eval_rewards/margins": 0.00043102249037474394,
-    "eval_rewards/rejected": -0.0006477099377661943,
-    "eval_runtime": 145.39,
+    "epoch": 4.0,
+    "eval_logits/chosen": -1.864721417427063,
+    "eval_logits/rejected": -1.8603547811508179,
+    "eval_logps/chosen": -36.019264221191406,
+    "eval_logps/rejected": -39.96937561035156,
+    "eval_loss": 2461.944580078125,
+    "eval_rewards/accuracies": 0.5568937063217163,
+    "eval_rewards/chosen": -0.019847100600600243,
+    "eval_rewards/margins": 0.0046804738231003284,
+    "eval_rewards/rejected": -0.024527573958039284,
+    "eval_runtime": 145.7077,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.359,
-    "eval_steps_per_second": 0.296
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295
 }