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: []
@@ -16,17 +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: 0.9913
-- Rewards/chosen: -0.0567
-- Rewards/rejected: -0.0655
-- Rewards/accuracies: 0.5424
+- Rewards/chosen: -0.0569
+- Rewards/rejected: -0.0657
+- Rewards/accuracies: 0.5511
 - Rewards/margins: 0.0088
-- Logps/rejected: -44.0648
-- Logps/chosen: -39.7075
-- Logits/rejected: -1.5832
-- Logits/chosen: -1.5871
+- Logps/rejected: -44.0859
+- Logps/chosen: -39.7278
+- Logits/rejected: -1.5833
+- Logits/chosen: -1.5872
 
 ## Model description
 

all_results.json

@@ -1,17 +1,17 @@
 {
     "epoch": 4.0,
-    "eval_logits/chosen": -2.221278667449951,
-    "eval_logits/rejected": -2.2164483070373535,
-    "eval_logps/chosen": -34.091766357421875,
-    "eval_logps/rejected": -37.60596466064453,
-    "eval_loss": 0.9996774792671204,
-    "eval_rewards/accuracies": 0.5477575063705444,
-    "eval_rewards/chosen": -0.000572161574382335,
-    "eval_rewards/margins": 0.00032129042665474117,
-    "eval_rewards/rejected": -0.0008934520301409066,
-    "eval_runtime": 145.6061,
+    "eval_logits/chosen": -1.5872344970703125,
+    "eval_logits/rejected": -1.58331298828125,
+    "eval_logps/chosen": -39.727821350097656,
+    "eval_logps/rejected": -44.08587646484375,
+    "eval_loss": 0.9912604689598083,
+    "eval_rewards/accuracies": 0.5510797500610352,
+    "eval_rewards/chosen": -0.056932687759399414,
+    "eval_rewards/margins": 0.008759919553995132,
+    "eval_rewards/rejected": -0.06569261103868484,
+    "eval_runtime": 145.7755,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.356,
+    "eval_samples_per_second": 2.353,
     "eval_steps_per_second": 0.295,
     "train_loss": 0.7654441864459546,
     "train_runtime": 10797.8934,

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 1.0,
-    "eval_logits/chosen": -2.221278667449951,
-    "eval_logits/rejected": -2.2164483070373535,
-    "eval_logps/chosen": -34.091766357421875,
-    "eval_logps/rejected": -37.60596466064453,
-    "eval_loss": 0.9996774792671204,
-    "eval_rewards/accuracies": 0.5477575063705444,
-    "eval_rewards/chosen": -0.000572161574382335,
-    "eval_rewards/margins": 0.00032129042665474117,
-    "eval_rewards/rejected": -0.0008934520301409066,
-    "eval_runtime": 145.6061,
+    "epoch": 4.0,
+    "eval_logits/chosen": -1.5872344970703125,
+    "eval_logits/rejected": -1.58331298828125,
+    "eval_logps/chosen": -39.727821350097656,
+    "eval_logps/rejected": -44.08587646484375,
+    "eval_loss": 0.9912604689598083,
+    "eval_rewards/accuracies": 0.5510797500610352,
+    "eval_rewards/chosen": -0.056932687759399414,
+    "eval_rewards/margins": 0.008759919553995132,
+    "eval_rewards/rejected": -0.06569261103868484,
+    "eval_runtime": 145.7755,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.356,
+    "eval_samples_per_second": 2.353,
     "eval_steps_per_second": 0.295
 }