apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-05 08:37:06 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Feature] Set v1 scheduler as default in develop (#3807)

Browse Source 
* Set scheduler v1 as default

* Set scheduler v1 as default

* Set scheduler v1 as default

* Set scheduler v1 as default

* Set scheduler v1 as default

* close V1 in guided_decoding

* fix vl ci

* close V1 in guided_decoding
This commit is contained in:
chenjian
2025-09-04 15:16:56 +08:00
committed by GitHub
parent e83251699f
commit 22c165d6dd
8 changed files with 50 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
fastdeploy/engine/args_utils.py
View File

@@ -392,6 +392,14 @@ class EngineArgs:
raise NotImplementedError("Logprob does not support enable_expert_parallel.")
if not current_platform.is_cuda():
raise NotImplementedError("Only CUDA platform supports logprob.")
if self.speculative_config is not None:
envs.ENABLE_V1_KVCACHE_SCHEDULER = 0
if self.splitwise_role != "mixed":
envs.ENABLE_V1_KVCACHE_SCHEDULER = 0
if not current_platform.is_cuda():
envs.ENABLE_V1_KVCACHE_SCHEDULER = 0
if self.guided_decoding_backend != "off":
envs.ENABLE_V1_KVCACHE_SCHEDULER = 0
@staticmethod
def add_cli_args(parser: FlexibleArgumentParser) -> FlexibleArgumentParser: