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: 0.9961
-- Rewards/chosen: -0.0016
-- Rewards/rejected: -0.0055
-- Rewards/accuracies: 0.5594
-- Rewards/margins: 0.0039
-- Logps/rejected: -37.5717
-- Logps/chosen: -34.0508
-- Logits/rejected: -2.2195
-- Logits/chosen: -2.2243
+- Loss: 0.9947
+- Rewards/chosen: -0.0004
+- Rewards/rejected: -0.0057
+- Rewards/accuracies: 0.5710
+- Rewards/margins: 0.0053
+- Logps/rejected: -37.5734
+- Logps/chosen: -34.0383
+- Logits/rejected: -2.2197
+- Logits/chosen: -2.2245
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.22452712059021,
+    "eval_logits/rejected": -2.219710350036621,
+    "eval_logps/chosen": -34.03827667236328,
+    "eval_logps/rejected": -37.573402404785156,
+    "eval_loss": 0.9946814179420471,
+    "eval_rewards/accuracies": 0.5710132718086243,
+    "eval_rewards/chosen": -0.00037224299740046263,
+    "eval_rewards/margins": 0.005306063685566187,
+    "eval_rewards/rejected": -0.005678307265043259,
+    "eval_runtime": 145.3784,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.359,
+    "eval_steps_per_second": 0.296,
     "train_loss": 0.9734619264478808,
     "train_runtime": 3253.307,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "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_logits/chosen": -2.22452712059021,
+    "eval_logits/rejected": -2.219710350036621,
+    "eval_logps/chosen": -34.03827667236328,
+    "eval_logps/rejected": -37.573402404785156,
+    "eval_loss": 0.9946814179420471,
+    "eval_rewards/accuracies": 0.5710132718086243,
+    "eval_rewards/chosen": -0.00037224299740046263,
+    "eval_rewards/margins": 0.005306063685566187,
+    "eval_rewards/rejected": -0.005678307265043259,
+    "eval_runtime": 145.3784,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.356,
-    "eval_steps_per_second": 0.295
+    "eval_samples_per_second": 2.359,
+    "eval_steps_per_second": 0.296
 }