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: 0.4944
-- Rewards/chosen: 0.1419
-- Rewards/rejected: 0.1189
-- Rewards/accuracies: 0.5307
-- Rewards/margins: 0.0230
-- Logps/rejected: -37.2789
-- Logps/chosen: -33.7508
-- Logits/rejected: -2.2285
-- Logits/chosen: -2.2333
+- Loss: 0.4958
+- Rewards/chosen: 0.1387
+- Rewards/rejected: 0.1213
+- Rewards/accuracies: 0.5453
+- Rewards/margins: 0.0174
+- Logps/rejected: -37.2740
+- Logps/chosen: -33.7572
+- Logits/rejected: -2.2278
+- Logits/chosen: -2.2326
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.232633352279663,
+    "eval_logits/rejected": -2.2278287410736084,
+    "eval_logps/chosen": -33.75718307495117,
+    "eval_logps/rejected": -37.27397155761719,
+    "eval_loss": 0.49580448865890503,
+    "eval_rewards/accuracies": 0.545265793800354,
+    "eval_rewards/chosen": 0.13868312537670135,
+    "eval_rewards/margins": 0.017359821125864983,
+    "eval_rewards/rejected": 0.12132329493761063,
+    "eval_runtime": 145.5703,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.356,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.4687722819192069,
     "train_runtime": 3250.758,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.22763729095459,
-    "eval_logits/rejected": -2.222792387008667,
-    "eval_logps/chosen": -34.090328216552734,
-    "eval_logps/rejected": -37.59687423706055,
-    "eval_loss": 0.6925650835037231,
-    "eval_rewards/accuracies": 0.5211793780326843,
-    "eval_rewards/chosen": -0.02788795717060566,
-    "eval_rewards/margins": 0.01224011555314064,
-    "eval_rewards/rejected": -0.04012807086110115,
-    "eval_runtime": 145.6508,
+    "eval_logits/chosen": -2.232633352279663,
+    "eval_logits/rejected": -2.2278287410736084,
+    "eval_logps/chosen": -33.75718307495117,
+    "eval_logps/rejected": -37.27397155761719,
+    "eval_loss": 0.49580448865890503,
+    "eval_rewards/accuracies": 0.545265793800354,
+    "eval_rewards/chosen": 0.13868312537670135,
+    "eval_rewards/margins": 0.017359821125864983,
+    "eval_rewards/rejected": 0.12132329493761063,
+    "eval_runtime": 145.5703,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.355,
+    "eval_samples_per_second": 2.356,
     "eval_steps_per_second": 0.295
 }