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: []
@@ -15,16 +18,16 @@ 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: 1.5962
-- Rewards/chosen: -0.0200
-- Rewards/rejected: -0.0335
-- Rewards/accuracies: 0.5166
-- Rewards/margins: 0.0135
-- Logps/rejected: -37.6004
-- Logps/chosen: -34.0845
-- Logits/rejected: -2.2263
+- Loss: 1.6251
+- Rewards/chosen: -0.0217
+- Rewards/rejected: -0.0319
+- Rewards/accuracies: 0.5079
+- Rewards/margins: 0.0102
+- Logps/rejected: -37.5964
+- Logps/chosen: -34.0889
+- Logits/rejected: -2.2264
 - Logits/chosen: -2.2312
 
 ## Model description

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.2312498092651367,
+    "eval_logits/rejected": -2.2264044284820557,
+    "eval_logps/chosen": -34.088890075683594,
+    "eval_logps/rejected": -37.596412658691406,
+    "eval_loss": 1.625149130821228,
+    "eval_rewards/accuracies": 0.5078904032707214,
+    "eval_rewards/chosen": -0.02173524722456932,
+    "eval_rewards/margins": 0.010181299410760403,
+    "eval_rewards/rejected": -0.0319165475666523,
+    "eval_runtime": 145.6511,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.355,
+    "eval_steps_per_second": 0.295,
     "train_loss": 1.3100324989913346,
     "train_runtime": 3250.9757,
     "train_samples": 3079,

eval_results.json

@@ -1,15 +1,15 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.2320711612701416,
-    "eval_logits/rejected": -2.2272377014160156,
-    "eval_logps/chosen": -33.7447624206543,
-    "eval_logps/rejected": -37.26742935180664,
-    "eval_loss": 0.49609801173210144,
-    "eval_rewards/accuracies": 0.5278239250183105,
-    "eval_rewards/chosen": 0.11591599881649017,
-    "eval_rewards/margins": 0.016239874064922333,
-    "eval_rewards/rejected": 0.09967613220214844,
-    "eval_runtime": 145.6745,
+    "eval_logits/chosen": -2.2312498092651367,
+    "eval_logits/rejected": -2.2264044284820557,
+    "eval_logps/chosen": -34.088890075683594,
+    "eval_logps/rejected": -37.596412658691406,
+    "eval_loss": 1.625149130821228,
+    "eval_rewards/accuracies": 0.5078904032707214,
+    "eval_rewards/chosen": -0.02173524722456932,
+    "eval_rewards/margins": 0.010181299410760403,
+    "eval_rewards/rejected": -0.0319165475666523,
+    "eval_runtime": 145.6511,
     "eval_samples": 343,
     "eval_samples_per_second": 2.355,
     "eval_steps_per_second": 0.295