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: 24.9557
-- Rewards/chosen: -0.0066
-- Rewards/rejected: -0.0089
-- Rewards/accuracies: 0.5216
-- Rewards/margins: 0.0023
-- Logps/rejected: -37.6060
-- Logps/chosen: -34.1010
-- Logits/rejected: -2.2229
-- Logits/chosen: -2.2278
+- Loss: 24.8075
+- Rewards/chosen: -0.0054
+- Rewards/rejected: -0.0092
+- Rewards/accuracies: 0.5133
+- Rewards/margins: 0.0038
+- Logps/rejected: -37.6085
+- Logps/chosen: -34.0885
+- Logits/rejected: -2.2234
+- Logits/chosen: -2.2282
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.228182792663574,
+    "eval_logits/rejected": -2.2233598232269287,
+    "eval_logps/chosen": -34.08854293823242,
+    "eval_logps/rejected": -37.60850143432617,
+    "eval_loss": 24.80753517150879,
+    "eval_rewards/accuracies": 0.5132890343666077,
+    "eval_rewards/chosen": -0.005399257410317659,
+    "eval_rewards/margins": 0.0037888663355261087,
+    "eval_rewards/rejected": -0.009188123047351837,
+    "eval_runtime": 145.6001,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.356,
+    "eval_steps_per_second": 0.295,
     "train_loss": 22.846377404943688,
     "train_runtime": 3252.1936,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.2322750091552734,
-    "eval_logits/rejected": -2.2274460792541504,
-    "eval_logps/chosen": -33.73501205444336,
-    "eval_logps/rejected": -37.2658805847168,
-    "eval_loss": 0.4987604320049286,
-    "eval_rewards/accuracies": 0.5186877250671387,
-    "eval_rewards/chosen": 0.029954075813293457,
-    "eval_rewards/margins": 0.004880108870565891,
-    "eval_rewards/rejected": 0.02507396601140499,
-    "eval_runtime": 145.7154,
+    "eval_logits/chosen": -2.228182792663574,
+    "eval_logits/rejected": -2.2233598232269287,
+    "eval_logps/chosen": -34.08854293823242,
+    "eval_logps/rejected": -37.60850143432617,
+    "eval_loss": 24.80753517150879,
+    "eval_rewards/accuracies": 0.5132890343666077,
+    "eval_rewards/chosen": -0.005399257410317659,
+    "eval_rewards/margins": 0.0037888663355261087,
+    "eval_rewards/rejected": -0.009188123047351837,
+    "eval_runtime": 145.6001,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.354,
+    "eval_samples_per_second": 2.356,
     "eval_steps_per_second": 0.295
 }