[DP Attention] Refactor: adding some utility functions (#9136)

This commit is contained in:
Cheng Wan
2025-08-13 21:08:06 -07:00
committed by GitHub
parent b3363cc1aa
commit b87aacb5c5
21 changed files with 216 additions and 159 deletions

View File

@@ -84,7 +84,6 @@ GLOBAL_SERVER_ARGS_KEYS = [
"device",
"disable_chunked_prefix_cache",
"disable_radix_cache",
"enable_dp_attention",
"enable_two_batch_overlap",
"tbo_token_distribution_threshold",
"enable_dp_lm_head",