Closed kevXZY closed 10 months ago
@kevXZY
I would recommend you to install directly on your mac (this is safe given that the installation will be done in a python venv, and that there are no system library required for macs if you are using metal
)
Performances of VMs are not great, especially when the VM system emulate generic CPUs.
We are currently working on refactoring the documentation, but in a nutshell, to install this on macos:
# One liner to install
git clone https://github.com/imartinez/privateGPT && cd privateGPT && \
python3.11 -m venv .venv && source .venv/bin/activate && \
pip install --upgrade pip poetry && poetry install --with ui,local && ./scripts/setup
# Patch the version of `llama-cpp-python` installed to add metal support
# This have to run with inside your python venv !
CMAKE_ARGS="-DLLAMA_METAL=on" pip install --force-reinstall --no-cache-dir llama-cpp-python
c.f. https://privategpt-preview-8380d3f2-962a-400e-87d4-fd1c4636f174.docs.buildwithfern.com/overview/welcome/quickstart (preview of documentation)
If you still want and persist in VM and Ubuntu world - I guess you simply have to install the missing package, as the error states. (But again, it will not take full advantage of your M2 chips with the chips that apple especially designed for neural network computation)
OS: Ubuntu 22.04.3 LTS ARM 64bit using VMware fusion on Mac M2.
I installed LlamaCPP and still getting this error:
~/privateGPT$ PGPT_PROFILES=local make run poetry run python -m private_gpt 02:13:22.418 [INFO ] private_gpt.settings.settings_loader - Starting application with profiles=['default', 'local'] Traceback (most recent call last): File "/home/kev/.cache/pypoetry/virtualenvs/private-gpt-f8Ius0_a-py3.11/lib/python3.11/site-packages/injector/init.py", line 798, in get return self._context[key]