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

[V1 Loader]support param create and load for wint2 and xpu backend (#3581)

Browse Source 
* support wint2 backend'

* [V1 Loader]support param create and load for wint2 and xpu backend

* update weight shape name

* update

* update

* update baseline.txt

* update model name

* update baseline.txt

* fix codestyle

* remove debug coode
This commit is contained in:
Zero Rains
2025-08-28 09:49:36 +08:00
committed by GitHub
parent b28a0343a6
commit e37e86b3b8
9 changed files with 307 additions and 326 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
fastdeploy/worker/gpu_model_runner.py
View File

@@ -1752,7 +1752,6 @@ class GPUModelRunner(ModelRunnerBase):
token_type_ids_w_video = token_type_ids
input_ids = inputs["input_ids"]
# convert to img patch id
# TODO(lulinjun): may need to check model_config and model_cfg
image_mask = input_ids == self.model_config.im_patch_id
image_type_ids = inputs["image_type_ids"]
with paddle.amp.auto_cast(

1
fastdeploy/worker/metax_model_runner.py
View File

@@ -1664,7 +1664,6 @@ class MetaxModelRunner(ModelRunnerBase):
token_type_ids_w_video = token_type_ids
input_ids = inputs["input_ids"]
# convert to img patch id
# TODO(lulinjun): may need to check model_config and model_cfg
image_mask = input_ids == self.model_config.im_patch_id
image_type_ids = inputs["image_type_ids"]
with paddle.amp.auto_cast(