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: []
@@ -15,17 +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: 0.6949
-- Rewards/chosen: 0.0029
-- Rewards/rejected: 0.0046
-- Rewards/accuracies: 0.5116
-- Rewards/margins: -0.0017
-- Logps/rejected: -37.5051
-- Logps/chosen: -34.0272
-- Logits/rejected: -2.2395
-- Logits/chosen: -2.2444
+- Loss: 0.6950
+- Rewards/chosen: 0.0036
+- Rewards/rejected: 0.0057
+- Rewards/accuracies: 0.4900
+- Rewards/margins: -0.0021
+- Logps/rejected: -37.5024
+- Logps/chosen: -34.0256
+- Logits/rejected: -2.2389
+- Logits/chosen: -2.2438
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.2437589168548584,
+    "eval_logits/rejected": -2.2388815879821777,
+    "eval_logps/chosen": -34.025550842285156,
+    "eval_logps/rejected": -37.502403259277344,
+    "eval_loss": 0.6950117945671082,
+    "eval_rewards/accuracies": 0.49003320932388306,
+    "eval_rewards/chosen": 0.0036009380128234625,
+    "eval_rewards/margins": -0.0020864540711045265,
+    "eval_rewards/rejected": 0.005687391385436058,
+    "eval_runtime": 145.772,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.353,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.6908149576806403,
     "train_runtime": 3255.681,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.2439303398132324,
-    "eval_logits/rejected": -2.2390527725219727,
-    "eval_logps/chosen": -34.02507400512695,
-    "eval_logps/rejected": -37.508995056152344,
-    "eval_loss": 0.6933426260948181,
-    "eval_rewards/accuracies": 0.5132890343666077,
-    "eval_rewards/chosen": 0.0028422565665096045,
-    "eval_rewards/margins": 0.0005549592897295952,
-    "eval_rewards/rejected": 0.002287296811118722,
-    "eval_runtime": 145.8223,
+    "eval_logits/chosen": -2.2437589168548584,
+    "eval_logits/rejected": -2.2388815879821777,
+    "eval_logps/chosen": -34.025550842285156,
+    "eval_logps/rejected": -37.502403259277344,
+    "eval_loss": 0.6950117945671082,
+    "eval_rewards/accuracies": 0.49003320932388306,
+    "eval_rewards/chosen": 0.0036009380128234625,
+    "eval_rewards/margins": -0.0020864540711045265,
+    "eval_rewards/rejected": 0.005687391385436058,
+    "eval_runtime": 145.772,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.352,
+    "eval_samples_per_second": 2.353,
     "eval_steps_per_second": 0.295
 }