diff --git a/apps/grpo/qwen3_8b.yaml b/apps/grpo/qwen3_8b.yaml index 9b2f70edd..534e5b92a 100644 --- a/apps/grpo/qwen3_8b.yaml +++ b/apps/grpo/qwen3_8b.yaml @@ -41,7 +41,6 @@ policy: # Trainer configuration trainer: - use_dcp: true model: name: qwen3 flavor: 8B