apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-23 00:17:25 +08:00
Code Issues Actions 19 Packages Projects Releases Wiki Activity

[Features] support hugging face qwen3 moe (#3649)
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
Deploy GitHub Pages / deploy (push) Has been cancelled
Details

Browse Source 
* split ut

* qwen3-30B-A3B

* fix

* add test

* add test_torch_model.py

* fix test_torch_model.py

* delete print

* fix moe

* delete init.py

* fix

* fix

---------

Co-authored-by: bukejiyu <395822456@qq.com>
Co-authored-by: bukejiyu <52310069+bukejiyu@users.noreply.github.com>
This commit is contained in:
lizexu123
2025-08-30 15:26:05 +08:00
committed by GitHub
parent f206474cc7
commit 455205f991
9 changed files with 437 additions and 258 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fastdeploy/model_executor/layers/linear.py
+1 -2
View File
@@ -294,6 +294,7 @@ class ReplicatedLinear(LinearBase):
weight_loader=(
self.weight_loader if hasattr(self, "weight_loader") else default_weight_loader(self.fd_config)
),
model_format=fd_config.model_config.model_format,
)
@@ -446,7 +447,6 @@ class MergedColumnParallelLinear(ColumnParallelLinear):
shard_size = (self.local_rank + 1) * block_size
loaded_weight = slice_fn(loaded_weight, output_dim, start=shard_offset, end=shard_size)
loaded_weight = get_tensor(loaded_weight)
if not param._is_initialized():
param.initialize()
param_shard_size = output_size // 2
@@ -574,7 +574,6 @@ class QKVParallelLinear(ColumnParallelLinear):
shard_size = (shard_id + 1) * block_size
loaded_weight = slice_fn(loaded_weight, output_dim, start=shard_offset, end=shard_size)
loaded_weight = get_tensor(loaded_weight)
if not param._is_initialized():
param.initialize()