End of training

README.md

@@ -1,11 +1,13 @@
 ---
-license: apache-2.0
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
 base_model: norallm/normistral-7b-warm
+datasets:
+- hugodk-sch/aftonposten_title_prefs
 model-index:
 - name: ap-normistral-7b-align-scan
   results: []
@@ -16,17 +18,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # ap-normistral-7b-align-scan
 
-This model is a fine-tuned version of [norallm/normistral-7b-warm](https://huggingface.co/norallm/normistral-7b-warm) on the None dataset.
+This model is a fine-tuned version of [data/ap-normistral-7b-sft-qlora](https://huggingface.co/data/ap-normistral-7b-sft-qlora) on the hugodk-sch/aftonposten_title_prefs dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4980
-- Rewards/chosen: 0.1165
-- Rewards/rejected: 0.1232
-- Rewards/accuracies: 0.5017
-- Rewards/margins: -0.0067
-- Logps/rejected: -35.7905
-- Logps/chosen: -32.2767
-- Logits/rejected: 98.3630
-- Logits/chosen: 98.3713
+- Loss: 0.5060
+- Rewards/chosen: 0.0978
+- Rewards/rejected: 0.1310
+- Rewards/accuracies: 0.5220
+- Rewards/margins: -0.0333
+- Logps/rejected: -35.7793
+- Logps/chosen: -32.3035
+- Logits/rejected: 98.3593
+- Logits/chosen: 98.3666
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": 98.3665542602539,
+    "eval_logits/rejected": 98.35928344726562,
+    "eval_logps/chosen": -32.303466796875,
+    "eval_logps/rejected": -35.77931594848633,
+    "eval_loss": 0.5059725642204285,
+    "eval_rewards/accuracies": 0.5220099687576294,
+    "eval_rewards/chosen": 0.09779632091522217,
+    "eval_rewards/margins": -0.033251672983169556,
+    "eval_rewards/rejected": 0.13104799389839172,
+    "eval_runtime": 103.819,
+    "eval_samples": 343,
+    "eval_samples_per_second": 3.304,
+    "eval_steps_per_second": 0.414,
     "train_loss": 0.38857739250381274,
     "train_runtime": 2559.074,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": 98.25546264648438,
-    "eval_logits/rejected": 98.2234878540039,
-    "eval_logps/chosen": -32.8620491027832,
-    "eval_logps/rejected": -36.65834426879883,
-    "eval_loss": 0.6803030967712402,
-    "eval_rewards/accuracies": 0.5186877250671387,
-    "eval_rewards/chosen": -0.0837748572230339,
-    "eval_rewards/margins": 0.05458855256438255,
-    "eval_rewards/rejected": -0.13836342096328735,
-    "eval_runtime": 103.6862,
+    "eval_logits/chosen": 98.3665542602539,
+    "eval_logits/rejected": 98.35928344726562,
+    "eval_logps/chosen": -32.303466796875,
+    "eval_logps/rejected": -35.77931594848633,
+    "eval_loss": 0.5059725642204285,
+    "eval_rewards/accuracies": 0.5220099687576294,
+    "eval_rewards/chosen": 0.09779632091522217,
+    "eval_rewards/margins": -0.033251672983169556,
+    "eval_rewards/rejected": 0.13104799389839172,
+    "eval_runtime": 103.819,
     "eval_samples": 343,
-    "eval_samples_per_second": 3.308,
-    "eval_steps_per_second": 0.415
+    "eval_samples_per_second": 3.304,
+    "eval_steps_per_second": 0.414
 }