Skip to content

Fix w8a8_int8 model shared experts fusion load weights error #5120

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
zhyncs merged 7 commits into from
Apr 12, 2025
Merged

Fix w8a8_int8 model shared experts fusion load weights error #5120

Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
cd6dc5e
fix shared experts fusion load weights with w8a8_int8 model
lambert0312 Apr 7, 2025
e7a6395
Merge branch 'main' into fix_shared_experts_fusion_w8a8_int8
zhyncs Apr 8, 2025
8ed70f2
Merge branch 'sgl-project:main' into fix_shared_experts_fusion_w8a8_int8
lambert0312 Apr 10, 2025
d25a6f0
Merge branch 'main' into fix_shared_experts_fusion_w8a8_int8
lambert0312 Apr 10, 2025
533e10d
Merge branch 'main' into fix_shared_experts_fusion_w8a8_int8
lambert0312 Apr 11, 2025
ebb8fce
Merge branch 'main' into fix_shared_experts_fusion_w8a8_int8
lambert0312 Apr 12, 2025
345b055
Merge branch 'main' into fix_shared_experts_fusion_w8a8_int8
lambert0312 Apr 12, 2025
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
26 changes: 18 additions & 8 deletions python/sglang/srt/models/deepseek_v2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1480,14 +1480,24 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
if self.n_share_experts_fusion is not None and self.n_share_experts_fusion > 0:
weights_list = list(weights)
weights_dict = dict(weights_list)
suffix_list = [
"down_proj.weight",
"down_proj.weight_scale_inv",
"gate_proj.weight",
"gate_proj.weight_scale_inv",
"up_proj.weight",
"up_proj.weight_scale_inv",
]
if self.quant_config.get_name() == "w8a8_int8":
suffix_list = [
"down_proj.weight",
"down_proj.weight_scale",
"gate_proj.weight",
"gate_proj.weight_scale",
"up_proj.weight",
"up_proj.weight_scale",
]
else:
suffix_list = [
"down_proj.weight",
"down_proj.weight_scale_inv",
"gate_proj.weight",
"gate_proj.weight_scale_inv",
"up_proj.weight",
"up_proj.weight_scale_inv",
]
names_to_remove = []
for moe_layer in tqdm(
range(
Expand Down
Loading