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: []
@@ -15,16 +18,16 @@ 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: 0.6937
-- Rewards/chosen: 0.0015
-- Rewards/rejected: 0.0022
-- Rewards/accuracies: 0.5050
-- Rewards/margins: -0.0007
-- Logps/rejected: -37.5058
-- Logps/chosen: -34.0273
-- Logits/rejected: -2.2390
+- Loss: 0.6928
+- Rewards/chosen: 0.0020
+- Rewards/rejected: 0.0008
+- Rewards/accuracies: 0.5100
+- Rewards/margins: 0.0011
+- Logps/rejected: -37.5125
+- Logps/chosen: -34.0248
+- Logits/rejected: -2.2389
 - Logits/chosen: -2.2438
 
 ## Model description

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.243804693222046,
+    "eval_logits/rejected": -2.238922119140625,
+    "eval_logps/chosen": -34.02475357055664,
+    "eval_logps/rejected": -37.51246643066406,
+    "eval_loss": 0.6927995681762695,
+    "eval_rewards/accuracies": 0.5099667906761169,
+    "eval_rewards/chosen": 0.0019583876710385084,
+    "eval_rewards/margins": 0.0011271554976701736,
+    "eval_rewards/rejected": 0.0008312325226143003,
+    "eval_runtime": 145.6648,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.355,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.6919509800997647,
     "train_runtime": 3253.2141,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.243914842605591,
-    "eval_logits/rejected": -2.239037275314331,
-    "eval_logps/chosen": -34.036094665527344,
-    "eval_logps/rejected": -37.504669189453125,
-    "eval_loss": 0.693877100944519,
-    "eval_rewards/accuracies": 0.47508305311203003,
-    "eval_rewards/chosen": -0.00015425789752043784,
-    "eval_rewards/margins": -0.0013494607992470264,
-    "eval_rewards/rejected": 0.0011952028144150972,
-    "eval_runtime": 145.559,
+    "eval_logits/chosen": -2.243804693222046,
+    "eval_logits/rejected": -2.238922119140625,
+    "eval_logps/chosen": -34.02475357055664,
+    "eval_logps/rejected": -37.51246643066406,
+    "eval_loss": 0.6927995681762695,
+    "eval_rewards/accuracies": 0.5099667906761169,
+    "eval_rewards/chosen": 0.0019583876710385084,
+    "eval_rewards/margins": 0.0011271554976701736,
+    "eval_rewards/rejected": 0.0008312325226143003,
+    "eval_runtime": 145.6648,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.356,
+    "eval_samples_per_second": 2.355,
     "eval_steps_per_second": 0.295
 }