Skip to content

[Bugfix] Fix broken deepseek fp8 TP weights loading #24367

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Sep 6, 2025
Merged

[Bugfix] Fix broken deepseek fp8 TP weights loading #24367

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion vllm/model_executor/layers/linear.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -262,7 +262,7 @@ def __init__(
self.tp_size = (get_tensor_model_parallel_world_size()
if not disable_tp else 1)

def __post_init__(self):
def update_param_tp_status(self):
for param in self.parameters():
if isinstance(param, BasevLLMParameter):
param.tp_rank = self.tp_rank
Expand Down Expand Up @@ -459,6 +459,7 @@ def __init__(
})
else:
self.register_parameter("bias", None)
self.update_param_tp_status()

def weight_loader(self, param: Parameter, loaded_weight: torch.Tensor):

Expand Down Expand Up @@ -1250,6 +1251,7 @@ def __init__(
})
else:
self.register_parameter("bias", None)
self.update_param_tp_status()

def weight_loader(self, param: Parameter, loaded_weight: torch.Tensor):
input_dim = getattr(param, "input_dim", None)
Expand Down
3 changes: 2 additions & 1 deletion vllm/model_executor/layers/quantization/fp8.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -270,7 +270,8 @@ def create_weights(
layer.weight_block_size = None

if self.block_quant:
tp_size = get_tensor_model_parallel_world_size()
tp_size = getattr(layer, "tp_size",
get_tensor_model_parallel_world_size())
assert self.quant_config.weight_block_size is not None
layer.weight_block_size = self.quant_config.weight_block_size
block_n, block_k = (
Expand Down