apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-05 16:48:03 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[v1loader]Reduce EB300B model loading time (#3700)

Browse Source 
* speed up eb45

* update
This commit is contained in:
bukejiyu
2025-09-02 19:13:57 +08:00
committed by GitHub
parent 693c7d781c
commit b6a4115369
4 changed files with 45 additions and 36 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
fastdeploy/model_executor/load_weight_utils.py
View File

@@ -29,7 +29,6 @@ from safetensors import safe_open
from tqdm import tqdm
from fastdeploy.config import FDConfig
from fastdeploy.model_executor.layers.utils import get_tensor
from fastdeploy.model_executor.models.tp_utils import (
check_tensor_parallel_prerequisites,
)
@@ -186,8 +185,7 @@ def fast_weights_iterator(safe_tensor_list: list[str]):
with fast_safe_open(st_file, framework="np") as f:
for name in f.keys():
param_slice = f.get_slice(name)
paddle_tensor = get_tensor(param_slice)
yield name, paddle_tensor
yield name, param_slice
def fastsafetensors_weights_iterator(