Skip to content
pre-commit

Enable bitsandbytes quantization on AMD GPUs that use warp size 32 #64393

Sign in to view logs

Enable bitsandbytes quantization on AMD GPUs that use warp size 32

Enable bitsandbytes quantization on AMD GPUs that use warp size 32 #64393

Triggered via pull request November 16, 2025 00:19
@sstamenksstamenk
synchronize #27307
sstamenk:enable_bitsandbytes_quant_rocm
Status Failure
Total duration 3m 32s
Artifacts –

pre-commit.yml

on: pull_request
pre-commit
3m 27s
pre-commit
Fit to window
Zoom out
Zoom in

Annotations

2 errors
pre-commit
Process completed with exit code 1.
Ruff (E501): tests/models/quantization/test_bitsandbytes.py#L20
tests/models/quantization/test_bitsandbytes.py:20:89: E501 Line too long (91 > 88)