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: 0.6872
-- Rewards/chosen: -0.0004
-- Rewards/rejected: -0.0276
-- Rewards/accuracies: 0.5598
-- Rewards/margins: 0.0272
-- Logps/rejected: -37.5627
-- Logps/chosen: -34.0353
+- Loss: 0.6903
+- Rewards/chosen: -0.0025
+- Rewards/rejected: -0.0237
+- Rewards/accuracies: 0.5519
+- Rewards/margins: 0.0212
+- Logps/rejected: -37.5562
+- Logps/chosen: -34.0388
 - Logits/rejected: -2.2221
-- Logits/chosen: -2.2269
+- Logits/chosen: -2.2270
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.226956844329834,
+    "eval_logits/rejected": -2.2221333980560303,
+    "eval_logps/chosen": -34.03876876831055,
+    "eval_logps/rejected": -37.55620193481445,
+    "eval_loss": 0.6902801394462585,
+    "eval_rewards/accuracies": 0.5519103407859802,
+    "eval_rewards/chosen": -0.002528026234358549,
+    "eval_rewards/margins": 0.021220751106739044,
+    "eval_rewards/rejected": -0.02374877780675888,
+    "eval_runtime": 145.6815,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.6364745784115482,
     "train_runtime": 3251.6479,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.2225441932678223,
-    "eval_logits/rejected": -2.217723846435547,
-    "eval_logps/chosen": -34.06193161010742,
-    "eval_logps/rejected": -37.57505416870117,
-    "eval_loss": 0.9846408367156982,
-    "eval_rewards/accuracies": 0.5137043595314026,
-    "eval_rewards/chosen": -0.013691178523004055,
-    "eval_rewards/margins": 0.015524487011134624,
-    "eval_rewards/rejected": -0.02921566367149353,
-    "eval_runtime": 145.6349,
+    "eval_logits/chosen": -2.226956844329834,
+    "eval_logits/rejected": -2.2221333980560303,
+    "eval_logps/chosen": -34.03876876831055,
+    "eval_logps/rejected": -37.55620193481445,
+    "eval_loss": 0.6902801394462585,
+    "eval_rewards/accuracies": 0.5519103407859802,
+    "eval_rewards/chosen": -0.002528026234358549,
+    "eval_rewards/margins": 0.021220751106739044,
+    "eval_rewards/rejected": -0.02374877780675888,
+    "eval_runtime": 145.6815,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.355,
+    "eval_samples_per_second": 2.354,
     "eval_steps_per_second": 0.295
 }