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: 0.6902
-- Rewards/chosen: 0.0156
-- Rewards/rejected: 0.0052
-- Rewards/accuracies: 0.5133
-- Rewards/margins: 0.0103
-- Logps/rejected: -37.5091
-- Logps/chosen: -34.0123
-- Logits/rejected: -2.2392
-- Logits/chosen: -2.2441
+- Loss: 0.7015
+- Rewards/chosen: -0.0021
+- Rewards/rejected: 0.0101
+- Rewards/accuracies: 0.4925
+- Rewards/margins: -0.0122
+- Logps/rejected: -37.5022
+- Logps/chosen: -34.0376
+- Logits/rejected: -2.2391
+- Logits/chosen: -2.2440
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "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_samples": 343,
+    "eval_samples_per_second": 2.352,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.6898305428492558,
     "train_runtime": 3256.6227,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.244286060333252,
-    "eval_logits/rejected": -2.2393975257873535,
-    "eval_logps/chosen": -34.03020095825195,
-    "eval_logps/rejected": -37.50043869018555,
-    "eval_loss": 0.6983639597892761,
-    "eval_rewards/accuracies": 0.4987541437149048,
-    "eval_rewards/chosen": 0.0026094578206539154,
-    "eval_rewards/margins": -0.007098525296896696,
-    "eval_rewards/rejected": 0.00970798172056675,
-    "eval_runtime": 145.7128,
+    "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_samples": 343,
-    "eval_samples_per_second": 2.354,
+    "eval_samples_per_second": 2.352,
     "eval_steps_per_second": 0.295
 }