apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-12-24 13:28:13 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Speculative Decoding][MTP]Support mtp in epdptp mode (#4614)

Browse Source 
* support mtp many features

* support mtp reshard in rl mode

* fix function

* support mtp ep

* support mtp in hybird-dp-tp mode

* default open scheduler_v1 in mtp
This commit is contained in:
freeliuzc
2025-10-28 16:02:47 +08:00
committed by GitHub
parent b4014834a9
commit c63361fd1d
10 changed files with 124 additions and 74 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
fastdeploy/worker/worker_process.py
View File

@@ -778,13 +778,6 @@ def initialize_fd_config(args, ranks: int = 1, local_rank: int = 0) -> FDConfig:
logger.info(f"- Dynamic load weight: {load_config.dynamic_load_weight}")
logger.info(f"- Load strategy: {load_config.load_strategy}")
if (
args.speculative_config is not None
and ("method" in args.speculative_config)
and (args.speculative_config["method"] is not None)
):
logger.info("Set ENABLE_V1_KVCACHE_SCHEDULER to 0 due to not support speculative decoding now.")
envs.ENABLE_V1_KVCACHE_SCHEDULER = 0
if args.splitwise_role != "mixed" and args.cache_transfer_protocol != "rdma":
envs.ENABLE_V1_KVCACHE_SCHEDULER = 0
if not current_platform.is_cuda() and not current_platform.is_xpu():