apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-05 08:37:06 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

support qwen2 weight only (#3571)
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
Deploy GitHub Pages / deploy (push) Has been cancelled
Details
Publish Job / publish_pre_check (push) Has been cancelled
Details
Publish Job / print_publish_pre_check_outputs (push) Has been cancelled
Details
Publish Job / FD-Clone-Linux (push) Has been cancelled
Details
Publish Job / Show Code Archive Output (push) Has been cancelled
Details
Publish Job / BUILD_SM8090 (push) Has been cancelled
Details
Publish Job / BUILD_SM8689 (push) Has been cancelled
Details
Publish Job / PADDLE_PYPI_UPLOAD_8090 (push) Has been cancelled
Details
Publish Job / PADDLE_PYPI_UPLOAD_8689 (push) Has been cancelled
Details
Publish Job / Run FastDeploy Unit Tests and Coverage (push) Has been cancelled
Details
Publish Job / Run FastDeploy LogProb Tests (push) Has been cancelled
Details
Publish Job / Extracted partial CE model tasks to run in CI. (push) Has been cancelled
Details
Publish Job / Run Base Tests (push) Has been cancelled
Details
Publish Job / Run Accuracy Tests (push) Has been cancelled
Details

Browse Source
This commit is contained in:
bukejiyu
2025-08-24 11:14:34 +08:00
committed by GitHub
parent 77514e3e1e
commit bdbac0aa3d
3 changed files with 25 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
fastdeploy/model_executor/layers/linear.py
View File

@@ -348,8 +348,7 @@ class ColumnParallelLinear(LinearBase):
if self.with_bias:
# col parallel
_set_var_distributed(self.bias, split_axis=1)
if self.nranks > 1:
set_weight_attrs(self.bias, {"output_dim": True})
set_weight_attrs(self.bias, {"output_dim": True})
class MergedColumnParallelLinear(ColumnParallelLinear):
@@ -404,6 +403,7 @@ class MergedColumnParallelLinear(ColumnParallelLinear):
def weight_loader(self, param, loaded_weight, loaded_shard_id: Optional[str] = None):
output_dim = getattr(param, "output_dim", None)
assert output_dim is not None
shard_dim = -1 if output_dim else 0
output_size = param.shape[shard_dim]
if loaded_shard_id is None:
@@ -517,11 +517,12 @@ class QKVParallelLinear(ColumnParallelLinear):
with_bias=with_bias,
add_bias=add_bias,
)
setattr(self.weight, "output_dim", True)
def weight_loader(self, param, loaded_weight, loaded_shard_id: Optional[str] = None):
output_dim = getattr(param, "output_dim", None)
head_dim = param.shape[output_dim] // (self.num_heads_per_rank + 2 * self.kv_num_heads_per_rank)
assert output_dim is not None
dim = -1 if output_dim else 0
head_dim = param.shape[dim] // (self.num_heads_per_rank + 2 * self.kv_num_heads_per_rank)
if loaded_shard_id is None:
# Loaded weight is already fused on disk
shard_offsets = [
@@ -540,7 +541,6 @@ class QKVParallelLinear(ColumnParallelLinear):
assert loaded_shard_id in ["q", "k", "v"]
# Tensor parallelism splits the weight along the output_dim
if self.nranks != 1:
dim = -1 if output_dim else 0
if isinstance(loaded_weight, np.ndarray):
size = loaded_weight.shape[dim]
else:
@@ -717,13 +717,12 @@ class RowParallelLinear(LinearBase):
if self.with_bias:
# col parallel
_set_var_distributed(self.bias, split_axis=0)
if self.nranks > 1:
set_weight_attrs(
self.bias,
{
"output_dim": False,
},
)
set_weight_attrs(
self.bias,
{
"output_dim": False,
},
)
self.reduce_results = reduce_results