Skip to content

[Bugfix] Fix bnb 8bit model weights loading #19917

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
jeejeelee merged 1 commit into from
Jun 21, 2025
Merged

[Bugfix] Fix bnb 8bit model weights loading #19917

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: 2 additions & 2 deletions vllm/model_executor/model_loader/bitsandbytes_loader.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -577,10 +577,10 @@ def dequantize_dq(quant_states: dict) -> None:
thereby avoiding this computational overhead during inference. This comes
at the cost of increased memory usage.
"""
from bitsandbytes.functional import dequantize_blockwise
from bitsandbytes.functional import QuantState, dequantize_blockwise
for _, quant_state in quant_states.items():
# Copied from: https://github.com/bitsandbytes-foundation/bitsandbytes/blob/0.45.3/bitsandbytes/functional.py#L1352-#L1356
if quant_state.nested:
if isinstance(quant_state, QuantState) and quant_state.nested:
absmax = dequantize_blockwise(quant_state.absmax,
quant_state.state2)
absmax += quant_state.offset
Expand Down