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: 0.6981
-- Rewards/chosen: 0.0062
-- Rewards/rejected: 0.0101
-- Rewards/accuracies: 0.4983
-- Rewards/margins: -0.0039
-- Logps/rejected: -37.5040
-- Logps/chosen: -34.0268
-- Logits/rejected: -2.2390
-- Logits/chosen: -2.2439
+- Loss: 0.7054
+- Rewards/chosen: -0.0010
+- Rewards/rejected: 0.0163
+- Rewards/accuracies: 0.4726
+- Rewards/margins: -0.0174
+- Logps/rejected: -37.4962
+- Logps/chosen: -34.0358
+- Logits/rejected: -2.2386
+- Logits/chosen: -2.2435
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.2434651851654053,
+    "eval_logits/rejected": -2.2385833263397217,
+    "eval_logps/chosen": -34.03584671020508,
+    "eval_logps/rejected": -37.496219635009766,
+    "eval_loss": 0.7054064273834229,
+    "eval_rewards/accuracies": 0.472591370344162,
+    "eval_rewards/chosen": -0.001035186112858355,
+    "eval_rewards/margins": -0.017356980592012405,
+    "eval_rewards/rejected": 0.01632179319858551,
+    "eval_runtime": 145.8229,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.352,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.6928803004227675,
     "train_runtime": 3255.7403,
     "train_samples": 3079,

eval_results.json

@@ -1,15 +1,15 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.2439703941345215,
-    "eval_logits/rejected": -2.2390873432159424,
-    "eval_logps/chosen": -34.037574768066406,
-    "eval_logps/rejected": -37.50224685668945,
-    "eval_loss": 0.7015302777290344,
-    "eval_rewards/accuracies": 0.4925249218940735,
-    "eval_rewards/chosen": -0.002115294337272644,
-    "eval_rewards/margins": -0.012174182571470737,
-    "eval_rewards/rejected": 0.010058889165520668,
-    "eval_runtime": 145.8307,
+    "eval_logits/chosen": -2.2434651851654053,
+    "eval_logits/rejected": -2.2385833263397217,
+    "eval_logps/chosen": -34.03584671020508,
+    "eval_logps/rejected": -37.496219635009766,
+    "eval_loss": 0.7054064273834229,
+    "eval_rewards/accuracies": 0.472591370344162,
+    "eval_rewards/chosen": -0.001035186112858355,
+    "eval_rewards/margins": -0.017356980592012405,
+    "eval_rewards/rejected": 0.01632179319858551,
+    "eval_runtime": 145.8229,
     "eval_samples": 343,
     "eval_samples_per_second": 2.352,
     "eval_steps_per_second": 0.295