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 and refine vl (#2866)

Browse Source 
* refine vl config

* delete attn_sep

* fix vl accuracy
This commit is contained in:
Yuanle Liu
2025-07-16 20:59:28 +08:00
committed by GitHub
parent aa76085d1f
commit 63d6e7ce06
11 changed files with 63 additions and 117 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
fastdeploy/worker/worker_process.py
View File

@@ -567,6 +567,9 @@ def initialize_fd_config(args, ranks: int = 1, local_rank: int = 0) -> FDConfig:
decoding_config = DecodingConfig(vars(args))
speculative_config = SpeculativeConfig(vars(args))
parallel_config = ParallelConfig(vars(args))
parallel_config.tensor_parallel_rank = local_rank
parallel_config.tensor_parallel_size = ranks
parallel_config.expert_parallel_rank = int(local_rank / ranks)
load_config = LoadConfig(vars(args))
graph_opt_config = GraphOptimizationConfig(