apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-04 08:16:42 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Fix] disable scheduler v1 in guided decoding (#3877)
Some checks failed
CE Compile Job / ce_job_pre_check (push) Has been cancelled
Details
CE Compile Job / print_ce_job_pre_check_outputs (push) Has been cancelled
Details
CE Compile Job / FD-Clone-Linux (push) Has been cancelled
Details
CE Compile Job / Show Code Archive Output (push) Has been cancelled
Details
CE Compile Job / BUILD_SM8090 (push) Has been cancelled
Details
CE Compile Job / BUILD_SM8689 (push) Has been cancelled
Details
CE Compile Job / CE_UPLOAD (push) Has been cancelled
Details

Browse Source 
* disable scheduler v1 in guided decoding

* disable scheduler v1 in guided decoding
This commit is contained in:
chenjian
2025-09-04 20:54:55 +08:00
committed by GitHub
parent afcde19277
commit 8567ada09e
2 changed files with 5 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
fastdeploy/engine/args_utils.py
View File

@@ -398,6 +398,8 @@ class EngineArgs:
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:

3
fastdeploy/worker/worker_process.py
View File

@@ -755,6 +755,9 @@ def initialize_fd_config(args, ranks: int = 1, local_rank: int = 0) -> FDConfig:
if not current_platform.is_cuda():
logger.info("Set ENABLE_V1_KVCACHE_SCHEDULER to 0 due to not supported.")
envs.ENABLE_V1_KVCACHE_SCHEDULER = 0
if parallel_config.guided_decoding_backend != "off":
logger.info("Set ENABLE_V1_KVCACHE_SCHEDULER to 0 due to not supported guided_decoding.")
envs.ENABLE_V1_KVCACHE_SCHEDULER = 0
fd_config = FDConfig(
model_config=model_config,