BLD: limit gptqmodel installation to specified version (#4798)

Author: zwt-1234

xinference/deploy/docker/Dockerfile.cpu

@@ -21,7 +21,7 @@ ARG PIP_INDEX=https://pypi.org/simple
 RUN python -m pip install --upgrade -i "$PIP_INDEX" pip "setuptools<82" wheel "packaging>=24.2" && \
     pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu && \
     pip install -i "$PIP_INDEX" --upgrade-strategy only-if-needed -r /opt/inference/xinference/deploy/docker/requirements_cpu/requirements_cpu-base.txt && \
-    pip install --no-cache-dir -i "$PIP_INDEX" --no-build-isolation gptqmodel && \
+    pip install --no-cache-dir -i "$PIP_INDEX" --no-build-isolation "gptqmodel<6.0.0" && \
     pip install -i "$PIP_INDEX" --upgrade-strategy only-if-needed -r /opt/inference/xinference/deploy/docker/requirements_cpu/requirements_cpu-ml.txt && \
     pip install -i "$PIP_INDEX" --upgrade-strategy  only-if-needed -r /opt/inference/xinference/deploy/docker/requirements_cpu/requirements_cpu-models.txt && \
     cd /opt/inference && \