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: 0.6755
-- Rewards/chosen: -0.2256
-- Rewards/rejected: -0.3710
-- Rewards/accuracies: 0.5831
-- Rewards/margins: 0.1454
-- Logps/rejected: -38.1350
-- Logps/chosen: -34.4106
-- Logits/rejected: -2.1010
-- Logits/chosen: -2.1057
+- Loss: 0.6784
+- Rewards/chosen: -0.2292
+- Rewards/rejected: -0.3691
+- Rewards/accuracies: 0.5714
+- Rewards/margins: 0.1398
+- Logps/rejected: -38.1317
+- Logps/chosen: -34.4166
+- Logits/rejected: -2.1013
+- Logits/chosen: -2.1060
 
 ## Model description
 

all_results.json

@@ -1,17 +1,17 @@
 {
     "epoch": 4.0,
-    "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_logits/chosen": -2.106013536453247,
+    "eval_logits/rejected": -2.10129714012146,
+    "eval_logps/chosen": -34.416629791259766,
+    "eval_logps/rejected": -38.131744384765625,
+    "eval_loss": 0.67835533618927,
+    "eval_rewards/accuracies": 0.5714285373687744,
+    "eval_rewards/chosen": -0.22924686968326569,
+    "eval_rewards/margins": 0.13982589542865753,
+    "eval_rewards/rejected": -0.36907273530960083,
+    "eval_runtime": 145.7883,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.354,
+    "eval_samples_per_second": 2.353,
     "eval_steps_per_second": 0.295,
     "train_loss": 0.24285796547864938,
     "train_runtime": 10807.9192,

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "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,
+    "epoch": 4.0,
+    "eval_logits/chosen": -2.106013536453247,
+    "eval_logits/rejected": -2.10129714012146,
+    "eval_logps/chosen": -34.416629791259766,
+    "eval_logps/rejected": -38.131744384765625,
+    "eval_loss": 0.67835533618927,
+    "eval_rewards/accuracies": 0.5714285373687744,
+    "eval_rewards/chosen": -0.22924686968326569,
+    "eval_rewards/margins": 0.13982589542865753,
+    "eval_rewards/rejected": -0.36907273530960083,
+    "eval_runtime": 145.7883,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.354,
+    "eval_samples_per_second": 2.353,
     "eval_steps_per_second": 0.295
 }