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: 0.4980
-- Rewards/chosen: 0.0533
-- Rewards/rejected: 0.0449
-- Rewards/accuracies: 0.5158
-- Rewards/margins: 0.0084
-- Logps/rejected: -37.2920
-- Logps/chosen: -33.7681
-- Logits/rejected: -2.2264
-- Logits/chosen: -2.2312
+- Loss: 0.4975
+- Rewards/chosen: 0.0550
+- Rewards/rejected: 0.0448
+- Rewards/accuracies: 0.5395
+- Rewards/margins: 0.0102
+- Logps/rejected: -37.2926
+- Logps/chosen: -33.7597
+- Logits/rejected: -2.2271
+- Logits/chosen: -2.2319
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.2319114208221436,
+    "eval_logits/rejected": -2.2271130084991455,
+    "eval_logps/chosen": -33.75971984863281,
+    "eval_logps/rejected": -37.29259490966797,
+    "eval_loss": 0.49754369258880615,
+    "eval_rewards/accuracies": 0.5394518375396729,
+    "eval_rewards/chosen": 0.05496572330594063,
+    "eval_rewards/margins": 0.010160263627767563,
+    "eval_rewards/rejected": 0.044805459678173065,
+    "eval_runtime": 145.6941,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.4876757522682091,
     "train_runtime": 3250.1859,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": -2.2261412143707275,
-    "eval_logits/rejected": -2.2213146686553955,
-    "eval_logps/chosen": -34.100101470947266,
-    "eval_logps/rejected": -37.61717224121094,
-    "eval_loss": 0.6907125115394592,
-    "eval_rewards/accuracies": 0.5132890343666077,
-    "eval_rewards/chosen": -0.013110256753861904,
-    "eval_rewards/margins": 0.007000547368079424,
-    "eval_rewards/rejected": -0.020110804587602615,
-    "eval_runtime": 145.4751,
+    "eval_logits/chosen": -2.2319114208221436,
+    "eval_logits/rejected": -2.2271130084991455,
+    "eval_logps/chosen": -33.75971984863281,
+    "eval_logps/rejected": -37.29259490966797,
+    "eval_loss": 0.49754369258880615,
+    "eval_rewards/accuracies": 0.5394518375396729,
+    "eval_rewards/chosen": 0.05496572330594063,
+    "eval_rewards/margins": 0.010160263627767563,
+    "eval_rewards/rejected": 0.044805459678173065,
+    "eval_runtime": 145.6941,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.358,
-    "eval_steps_per_second": 0.296
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295
 }