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: 1.0123
-- Rewards/chosen: 0.0060
-- Rewards/rejected: 0.0180
-- Rewards/accuracies: 0.4551
-- Rewards/margins: -0.0120
-- Logps/rejected: -37.4909
-- Logps/chosen: -34.0260
-- Logits/rejected: -2.2389
-- Logits/chosen: -2.2438
+- Loss: 1.0139
+- Rewards/chosen: -0.0014
+- Rewards/rejected: 0.0126
+- Rewards/accuracies: 0.4664
+- Rewards/margins: -0.0140
+- Logps/rejected: -37.4987
+- Logps/chosen: -34.0366
+- Logits/rejected: -2.2391
+- Logits/chosen: -2.2439
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.243932008743286,
+    "eval_logits/rejected": -2.239053249359131,
+    "eval_logps/chosen": -34.03658676147461,
+    "eval_logps/rejected": -37.49867630004883,
+    "eval_loss": 1.0139278173446655,
+    "eval_rewards/accuracies": 0.4663621187210083,
+    "eval_rewards/chosen": -0.0014253630070015788,
+    "eval_rewards/margins": -0.013985722325742245,
+    "eval_rewards/rejected": 0.01256035827100277,
+    "eval_runtime": 145.6648,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.355,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.9909477320584383,
     "train_runtime": 3254.3492,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.243980884552002,
-    "eval_logits/rejected": -2.2390992641448975,
-    "eval_logps/chosen": -34.03550720214844,
-    "eval_logps/rejected": -37.50373458862305,
-    "eval_loss": 1.0085219144821167,
-    "eval_rewards/accuracies": 0.48172760009765625,
-    "eval_rewards/chosen": -0.0005728387623094022,
-    "eval_rewards/margins": -0.008305290713906288,
-    "eval_rewards/rejected": 0.007732452359050512,
-    "eval_runtime": 145.8784,
+    "eval_logits/chosen": -2.243932008743286,
+    "eval_logits/rejected": -2.239053249359131,
+    "eval_logps/chosen": -34.03658676147461,
+    "eval_logps/rejected": -37.49867630004883,
+    "eval_loss": 1.0139278173446655,
+    "eval_rewards/accuracies": 0.4663621187210083,
+    "eval_rewards/chosen": -0.0014253630070015788,
+    "eval_rewards/margins": -0.013985722325742245,
+    "eval_rewards/rejected": 0.01256035827100277,
+    "eval_runtime": 145.6648,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.351,
+    "eval_samples_per_second": 2.355,
     "eval_steps_per_second": 0.295
 }