Skip to content

Fix Qwen3.5/3.6 load_weights stacked-mapping name mutation (gate_gate_up_proj / qkqkv_proj) #475

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

Open
bongmiin wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Fix Qwen3.5/3.6 load_weights stacked-mapping name mutation (gate_gate_up_proj / qkqkv_proj) #475

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
d2fda4e
Fix Fix Qwen3.5/3.6 load_weights stacked-mapping name mutation
bongmiin Jun 12, 2026
cd4bd8e
Fix indentation and correct name mapping logic
bongmiin Jun 12, 2026
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
14 changes: 8 additions & 6 deletions vllm/patches/vllm_for_multi_arc.patch
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11851,16 +11851,17 @@ index 000000000..b01b55f70
+ if "mlp.experts" in name:
+ continue
+
+ name = name.replace(weight_name, param_name)
+ name_mapped = name.replace(weight_name, param_name)
+ # Skip loading extra bias for GPTQ models.
+ if name.endswith(".bias") and name not in params_dict:
+ if name_mapped.endswith(".bias") and name_mapped not in params_dict:
+ continue
+ # Skip layers on other devices.
+ if is_pp_missing_parameter(name, self):
+ continue
+ # name = apply_attn_prefix(name, params_dict)
+ if name not in params_dict:
+ if name_mapped not in params_dict:
+ continue
+ name = name_mapped
+ param = params_dict[name]
+
+ # Use weight_loader_v2 for tuple shard_id (GDN projections).
Expand Down Expand Up @@ -12604,15 +12605,16 @@ index 000000000..547c1885a
+ if "mlp.experts" in name:
+ continue
+
+ name = name.replace(weight_name, param_name)
+ name_mapped = name.replace(weight_name, param_name)
+ # Skip loading extra bias for GPTQ models.
+ if name.endswith(".bias") and name not in params_dict:
+ if name_mapped.endswith(".bias") and name_mapped not in params_dict:
+ continue
+ # Skip layers on other devices.
+ if is_pp_missing_parameter(name, self):
+ if is_pp_missing_parameter(name_mapped, self):
+ continue
+ if name not in params_dict:
+ continue
+ name = name_mapped
+ param = params_dict[name]
+ weight_loader = param.weight_loader
+ weight_loader(param, loaded_weight, shard_id)
Expand Down