frozenleaves/vllm
1
0
Fork 0
mirror of https://github.com/vllm-project/vllm.git synced 2025-10-20 14:53:52 +08:00
Code Issues Packages Projects Releases Wiki Activity

[[Misc]Upgrade bitsandbytes to the latest version 0.45.0 (#11201)

Browse Source
This commit is contained in:
Jee Jee Li
2024-12-15 11:03:06 +08:00
committed by GitHub
parent 886936837c
commit 15859f2357
5 changed files with 8 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
Dockerfile
View File

@ -218,7 +218,7 @@ FROM vllm-base AS vllm-openai
# install additional dependencies for openai api server
RUN --mount=type=cache,target=/root/.cache/pip \
pip install accelerate hf_transfer 'modelscope!=1.15.0' 'bitsandbytes>=0.44.0' timm==0.9.10
pip install accelerate hf_transfer 'modelscope!=1.15.0' 'bitsandbytes>=0.45.0' timm==0.9.10
ENV VLLM_USAGE_SOURCE production-docker-image

2
docs/source/quantization/bnb.rst
View File

@ -11,7 +11,7 @@ Below are the steps to utilize BitsAndBytes with vLLM.
.. code-block:: console
$ pip install bitsandbytes>=0.44.0
$ pip install bitsandbytes>=0.45.0
vLLM reads the model's config file and supports both in-flight quantization and pre-quantized checkpoint.

2
requirements-test.in
View File

@ -25,7 +25,7 @@ datamodel_code_generator # required for minicpm3 test
lm-eval[api]==0.4.4 # required for model evaluation test
# quantization
bitsandbytes>=0.44.0
bitsandbytes>=0.45.0
buildkite-test-collector==0.1.9
numpy < 2.0.0

2
requirements-test.txt
View File

@ -37,7 +37,7 @@ audioread==3.0.1
# via librosa
awscli==1.35.23
# via -r requirements-test.in
bitsandbytes==0.44.1
bitsandbytes>=0.45.0
# via -r requirements-test.in
black==24.10.0
# via datamodel-code-generator

8
vllm/model_executor/layers/quantization/bitsandbytes.py
View File

@ -145,12 +145,12 @@ class BitsAndBytesLinearMethod(LinearMethodBase):
def __init__(self, quant_config: BitsAndBytesConfig):
try:
import bitsandbytes
if bitsandbytes.__version__ < "0.44.0":
if bitsandbytes.__version__ < "0.45.0":
raise ImportError("bitsandbytes version is wrong. Please "
"install bitsandbytes>=0.44.0.")
"install bitsandbytes>=0.45.0.")
except ImportError as err:
raise ImportError("Please install bitsandbytes>=0.44.0 via "
"`pip install bitsandbytes>=0.44.0` to use "
raise ImportError("Please install bitsandbytes>=0.45.0 via "
"`pip install bitsandbytes>=0.45.0` to use "
"bitsandbytes quantizer.") from err
self.quant_config = quant_config