End of training

README.md

@@ -1,10 +1,13 @@
 ---
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - 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: []
@@ -15,15 +18,15 @@ 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: 1.0355
-- Rewards/chosen: 0.0020
-- Rewards/rejected: 0.0020
-- Rewards/accuracies: 0.4817
-- Rewards/margins: 0.0000
-- Logps/rejected: -37.5127
-- Logps/chosen: -34.0306
+- Loss: 1.0334
+- Rewards/chosen: 0.0053
+- Rewards/rejected: 0.0035
+- Rewards/accuracies: 0.5012
+- Rewards/margins: 0.0019
+- Logps/rejected: -37.5097
+- Logps/chosen: -34.0239
 - Logits/rejected: -2.2390
 - Logits/chosen: -2.2439
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.243878126144409,
+    "eval_logits/rejected": -2.239006519317627,
+    "eval_logps/chosen": -34.02389144897461,
+    "eval_logps/rejected": -37.50969696044922,
+    "eval_loss": 1.033429503440857,
+    "eval_rewards/accuracies": 0.5012458562850952,
+    "eval_rewards/chosen": 0.005329286679625511,
+    "eval_rewards/margins": 0.0018695228500291705,
+    "eval_rewards/rejected": 0.003459764178842306,
+    "eval_runtime": 145.7674,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.353,
+    "eval_steps_per_second": 0.295,
     "train_loss": 1.0120837347848075,
     "train_runtime": 3256.3987,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.243934392929077,
-    "eval_logits/rejected": -2.2390573024749756,
-    "eval_logps/chosen": -34.03672409057617,
-    "eval_logps/rejected": -37.51164627075195,
-    "eval_loss": 1.6173344850540161,
-    "eval_rewards/accuracies": 0.52491694688797,
-    "eval_rewards/chosen": -0.0008701992337591946,
-    "eval_rewards/margins": -0.0028602415695786476,
-    "eval_rewards/rejected": 0.0019900428596884012,
-    "eval_runtime": 145.7376,
+    "eval_logits/chosen": -2.243878126144409,
+    "eval_logits/rejected": -2.239006519317627,
+    "eval_logps/chosen": -34.02389144897461,
+    "eval_logps/rejected": -37.50969696044922,
+    "eval_loss": 1.033429503440857,
+    "eval_rewards/accuracies": 0.5012458562850952,
+    "eval_rewards/chosen": 0.005329286679625511,
+    "eval_rewards/margins": 0.0018695228500291705,
+    "eval_rewards/rejected": 0.003459764178842306,
+    "eval_runtime": 145.7674,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.354,
+    "eval_samples_per_second": 2.353,
     "eval_steps_per_second": 0.295
 }