diff --git a/examples/qwen/lora.yml b/examples/qwen/lora.yml index 0deff6745..ca2061a5b 100644 --- a/examples/qwen/lora.yml +++ b/examples/qwen/lora.yml @@ -53,7 +53,7 @@ resume_from_checkpoint: local_rank: logging_steps: 1 xformers_attention: -flash_attention: true +flash_attention: warmup_steps: 10 eval_steps: 0.05 diff --git a/examples/qwen/qlora.yml b/examples/qwen/qlora.yml index d0a40461a..224020b7f 100644 --- a/examples/qwen/qlora.yml +++ b/examples/qwen/qlora.yml @@ -53,7 +53,7 @@ resume_from_checkpoint: local_rank: logging_steps: 1 xformers_attention: -flash_attention: true +flash_attention: warmup_steps: 10 eval_steps: 0.05