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: []
@@ -16,15 +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: 6.0348
-- Rewards/chosen: -0.0125
-- Rewards/rejected: -0.0385
-- Rewards/accuracies: 0.5835
-- Rewards/margins: 0.0260
-- Logps/rejected: -37.7091
-- Logps/chosen: -34.0972
+- Loss: 6.0779
+- Rewards/chosen: -0.0133
+- Rewards/rejected: -0.0383
+- Rewards/accuracies: 0.5714
+- Rewards/margins: 0.0250
+- Logps/rejected: -37.7083
+- Logps/chosen: -34.1011
 - Logits/rejected: -2.2004
 - Logits/chosen: -2.2052
 

all_results.json

@@ -1,18 +1,18 @@
 {
     "epoch": 4.0,
-    "eval_logits/chosen": -2.22882342338562,
-    "eval_logits/rejected": -2.223994731903076,
-    "eval_logps/chosen": -34.021240234375,
-    "eval_logps/rejected": -37.54384994506836,
-    "eval_loss": 6.194193363189697,
-    "eval_rewards/accuracies": 0.5664451718330383,
-    "eval_rewards/chosen": 0.0026620819699019194,
-    "eval_rewards/margins": 0.008108925074338913,
-    "eval_rewards/rejected": -0.0054468438029289246,
-    "eval_runtime": 145.4532,
+    "eval_logits/chosen": -2.205192804336548,
+    "eval_logits/rejected": -2.20035719871521,
+    "eval_logps/chosen": -34.1010856628418,
+    "eval_logps/rejected": -37.70829772949219,
+    "eval_loss": 6.077866077423096,
+    "eval_rewards/accuracies": 0.5714285373687744,
+    "eval_rewards/chosen": -0.013306654058396816,
+    "eval_rewards/margins": 0.025028489530086517,
+    "eval_rewards/rejected": -0.03833514451980591,
+    "eval_runtime": 145.6887,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.358,
-    "eval_steps_per_second": 0.296,
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295,
     "train_loss": 1.3887401007986688,
     "train_runtime": 10806.9485,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 1.0,
-    "eval_logits/chosen": -2.22882342338562,
-    "eval_logits/rejected": -2.223994731903076,
-    "eval_logps/chosen": -34.021240234375,
-    "eval_logps/rejected": -37.54384994506836,
-    "eval_loss": 6.194193363189697,
-    "eval_rewards/accuracies": 0.5664451718330383,
-    "eval_rewards/chosen": 0.0026620819699019194,
-    "eval_rewards/margins": 0.008108925074338913,
-    "eval_rewards/rejected": -0.0054468438029289246,
-    "eval_runtime": 145.4532,
+    "epoch": 4.0,
+    "eval_logits/chosen": -2.205192804336548,
+    "eval_logits/rejected": -2.20035719871521,
+    "eval_logps/chosen": -34.1010856628418,
+    "eval_logps/rejected": -37.70829772949219,
+    "eval_loss": 6.077866077423096,
+    "eval_rewards/accuracies": 0.5714285373687744,
+    "eval_rewards/chosen": -0.013306654058396816,
+    "eval_rewards/margins": 0.025028489530086517,
+    "eval_rewards/rejected": -0.03833514451980591,
+    "eval_runtime": 145.6887,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.358,
-    "eval_steps_per_second": 0.296
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295
 }