ハイパーパラメーター調整

yuiseki · Mar 21, 2024 · 11074cd · 11074cd
1 parent 4eb3267
commit 11074cd
Show file tree

Hide file tree

Showing 3 changed files with 8 additions and 8 deletions.
diff --git a/recipes/A5000_24GB_x8/fake-news-detector.yaml b/recipes/A5000_24GB_x8/fake-news-detector.yaml
@@ -11,8 +11,8 @@ dataset_train_split_test_size: 0.2
 lora_r: 8
 lora_alpha: 16
 lora_dropout: 0.05
-train_per_device_train_batch_size: 16
-train_gradient_accumulation_steps: 4
+train_per_device_train_batch_size: 4
+train_gradient_accumulation_steps: 2
 train_num_train_epochs: 4
 train_max_steps: 1000
 train_fp16: True

diff --git a/recipes/A5000_24GB_x8/hate-speech-detector.yaml b/recipes/A5000_24GB_x8/hate-speech-detector.yaml
@@ -11,10 +11,10 @@ dataset_train_split_test_size: 0.2
 lora_r: 8
 lora_alpha: 16
 lora_dropout: 0.05
-train_per_device_train_batch_size: 4
-train_gradient_accumulation_steps: 2
+train_per_device_train_batch_size: 8
+train_gradient_accumulation_steps: 4
 train_num_train_epochs: 4
-train_max_steps: 200
+train_max_steps: 1000
 train_fp16: True
 inference_max_new_tokens: 16
 evaluations:

diff --git a/recipes/A5000_24GB_x8/sentiment-analyzer.yaml b/recipes/A5000_24GB_x8/sentiment-analyzer.yaml
@@ -11,10 +11,10 @@ dataset_train_split_test_size: 0.2
 lora_r: 8
 lora_alpha: 16
 lora_dropout: 0.05
-train_per_device_train_batch_size: 4
-train_gradient_accumulation_steps: 2
+train_per_device_train_batch_size: 8
+train_gradient_accumulation_steps: 4
 train_num_train_epochs: 4
-train_max_steps: 200
+train_max_steps: 1000
 train_fp16: True
 inference_max_new_tokens: 16
 evaluations: