Skip to content

[bitsandbytes] improve dtype mismatch handling for bnb + lora. #11270

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 11 commits into from
Apr 17, 2025
Merged

[bitsandbytes] improve dtype mismatch handling for bnb + lora. #11270

Changes from all commits
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
5801679
improve dtype mismatch handling for bnb + lora.
sayakpaul Apr 10, 2025
8bf12e8
add a test
sayakpaul Apr 10, 2025
40958f7
Merge branch 'main' into improve-dtype-mismatch-bnb-lora
sayakpaul Apr 10, 2025
58c8183
Merge branch 'main' into improve-dtype-mismatch-bnb-lora
sayakpaul Apr 11, 2025
44e4ded
Merge branch 'main' into improve-dtype-mismatch-bnb-lora
sayakpaul Apr 11, 2025
fd694b0
Merge branch 'main' into improve-dtype-mismatch-bnb-lora
sayakpaul Apr 11, 2025
2107d4c
Merge branch 'main' into improve-dtype-mismatch-bnb-lora
sayakpaul Apr 14, 2025
8b7ef9c
fix and updates
sayakpaul Apr 14, 2025
40c0d89
Merge branch 'main' into improve-dtype-mismatch-bnb-lora
sayakpaul Apr 17, 2025
fccff45
update
sayakpaul Apr 17, 2025
98d84f9
Merge branch 'main' into improve-dtype-mismatch-bnb-lora
sayakpaul Apr 17, 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
8 changes: 5 additions & 3 deletions src/diffusers/quantizers/bitsandbytes/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -171,9 +171,11 @@ def dequantize_bnb_weight(weight: "torch.nn.Parameter", state=None, dtype: "torc

if cls_name == "Params4bit":
output_tensor = bnb.functional.dequantize_4bit(weight.data, weight.quant_state)
logger.warning_once(
f"The model is going to be dequantized in {output_tensor.dtype} - if you want to upcast it to another dtype, make sure to pass the desired dtype when quantizing the model through `bnb_4bit_quant_type` argument of `BitsAndBytesConfig`"
)
msg = f"The model is going to be dequantized in {output_tensor.dtype} - if you want to upcast it to another dtype, make sure to pass the desired dtype when quantizing the model through `bnb_4bit_quant_type` argument of `BitsAndBytesConfig`"
if dtype:
msg = f"The model is going to be first dequantized in {output_tensor.dtype} and type-casted to {dtype}"
output_tensor = output_tensor.to(dtype)
logger.warning_once(msg)
return output_tensor

if state.SCB is None:
Expand Down