End of training

README.md

@@ -1,11 +1,13 @@
 ---
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
-- alignment-handbook
 - 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: []
@@ -16,17 +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 the None 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: 0.9280
-- Rewards/chosen: -0.1384
-- Rewards/rejected: -0.2143
-- Rewards/accuracies: 0.5806
-- Rewards/margins: 0.0758
-- Logps/rejected: -38.0523
-- Logps/chosen: -34.3807
+- Loss: 0.9279
+- Rewards/chosen: -0.1421
+- Rewards/rejected: -0.2171
+- Rewards/accuracies: 0.5748
+- Rewards/margins: 0.0750
+- Logps/rejected: -38.0594
+- Logps/chosen: -34.3898
 - Logits/rejected: -2.1358
-- Logits/chosen: -2.1405
+- Logits/chosen: -2.1406
 
 ## Model description
 

all_results.json

@@ -1,17 +1,17 @@
 {
     "epoch": 4.0,
-    "eval_logits/chosen": -2.223268747329712,
-    "eval_logits/rejected": -2.218452215194702,
-    "eval_logps/chosen": -34.1085205078125,
-    "eval_logps/rejected": -37.61680603027344,
-    "eval_loss": 0.9893841743469238,
-    "eval_rewards/accuracies": 0.5124584436416626,
-    "eval_rewards/chosen": -0.02958693355321884,
-    "eval_rewards/margins": 0.010487398132681847,
-    "eval_rewards/rejected": -0.04007433354854584,
-    "eval_runtime": 145.4187,
+    "eval_logits/chosen": -2.1405961513519287,
+    "eval_logits/rejected": -2.13584041595459,
+    "eval_logps/chosen": -34.389774322509766,
+    "eval_logps/rejected": -38.05939865112305,
+    "eval_loss": 0.9278836846351624,
+    "eval_rewards/accuracies": 0.5747508406639099,
+    "eval_rewards/chosen": -0.14208722114562988,
+    "eval_rewards/margins": 0.0750228762626648,
+    "eval_rewards/rejected": -0.21711009740829468,
+    "eval_runtime": 145.1138,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.359,
+    "eval_samples_per_second": 2.364,
     "eval_steps_per_second": 0.296,
     "train_loss": 0.2625940967689861,
     "train_runtime": 10748.6274,

config.json

@@ -18,6 +18,8 @@
   "n_layer": 28,
   "n_positions": 2048,
   "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
     "bnb_4bit_compute_dtype": "float16",
     "bnb_4bit_quant_type": "nf4",
     "bnb_4bit_use_double_quant": false,
@@ -45,7 +47,7 @@
   },
   "tie_word_embeddings": false,
   "tokenizer_class": "GPT2Tokenizer",
-  "transformers_version": "4.37.2",
+  "transformers_version": "4.39.0.dev0",
   "use_cache": true,
   "vocab_size": 50400
 }

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 1.0,
-    "eval_logits/chosen": -2.223268747329712,
-    "eval_logits/rejected": -2.218452215194702,
-    "eval_logps/chosen": -34.1085205078125,
-    "eval_logps/rejected": -37.61680603027344,
-    "eval_loss": 0.9893841743469238,
-    "eval_rewards/accuracies": 0.5124584436416626,
-    "eval_rewards/chosen": -0.02958693355321884,
-    "eval_rewards/margins": 0.010487398132681847,
-    "eval_rewards/rejected": -0.04007433354854584,
-    "eval_runtime": 145.4187,
+    "epoch": 4.0,
+    "eval_logits/chosen": -2.1405961513519287,
+    "eval_logits/rejected": -2.13584041595459,
+    "eval_logps/chosen": -34.389774322509766,
+    "eval_logps/rejected": -38.05939865112305,
+    "eval_loss": 0.9278836846351624,
+    "eval_rewards/accuracies": 0.5747508406639099,
+    "eval_rewards/chosen": -0.14208722114562988,
+    "eval_rewards/margins": 0.0750228762626648,
+    "eval_rewards/rejected": -0.21711009740829468,
+    "eval_runtime": 145.1138,
     "eval_samples": 343,
-    "eval_samples_per_second": 2.359,
+    "eval_samples_per_second": 2.364,
     "eval_steps_per_second": 0.296
 }