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: 0.3276
-- Rewards/chosen: 0.0152
-- Rewards/rejected: 0.0022
-- Rewards/accuracies: 0.5365
-- Rewards/margins: 0.0131
-- Logps/rejected: -37.5142
-- Logps/chosen: -34.0176
-- Logits/rejected: -2.2390
-- Logits/chosen: -2.2439
+- Loss: 0.3447
+- Rewards/chosen: 0.0141
+- Rewards/rejected: 0.0084
+- Rewards/accuracies: 0.5162
+- Rewards/margins: 0.0057
+- Logps/rejected: -37.5073
+- Logps/chosen: -34.0189
+- Logits/rejected: -2.2388
+- Logits/chosen: -2.2437
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.2436861991882324,
+    "eval_logits/rejected": -2.2388129234313965,
+    "eval_logps/chosen": -34.01886749267578,
+    "eval_logps/rejected": -37.50727081298828,
+    "eval_loss": 0.34474730491638184,
+    "eval_rewards/accuracies": 0.5161960124969482,
+    "eval_rewards/chosen": 0.01411630492657423,
+    "eval_rewards/margins": 0.005696756765246391,
+    "eval_rewards/rejected": 0.008419547230005264,
+    "eval_runtime": 145.8183,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.352,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.3306902726749321,
     "train_runtime": 3256.1932,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.24367356300354,
-    "eval_logits/rejected": -2.2387921810150146,
-    "eval_logps/chosen": -34.03134536743164,
-    "eval_logps/rejected": -37.504520416259766,
-    "eval_loss": 0.4422106444835663,
-    "eval_rewards/accuracies": 0.4568106234073639,
-    "eval_rewards/chosen": 0.002567538060247898,
-    "eval_rewards/margins": -0.0071130492724478245,
-    "eval_rewards/rejected": 0.009680588729679585,
-    "eval_runtime": 145.7928,
+    "eval_logits/chosen": -2.2436861991882324,
+    "eval_logits/rejected": -2.2388129234313965,
+    "eval_logps/chosen": -34.01886749267578,
+    "eval_logps/rejected": -37.50727081298828,
+    "eval_loss": 0.34474730491638184,
+    "eval_rewards/accuracies": 0.5161960124969482,
+    "eval_rewards/chosen": 0.01411630492657423,
+    "eval_rewards/margins": 0.005696756765246391,
+    "eval_rewards/rejected": 0.008419547230005264,
+    "eval_runtime": 145.8183,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.353,
+    "eval_samples_per_second": 2.352,
     "eval_steps_per_second": 0.295
 }