BCS Blog

COBOL, z/OS performance measurement, and systems programming updates.

install oLLM

By Arch Brooks on July 31, 2026 in Artificial Intelligence

Back to Blog ← Back

Permalink

If you mean oLLM (the Python inference framework, not Ollama), there is currently no Gentoo ebuild, so the recommended installation is into a Python virtual environment using pip.

1. Install prerequisites

sudo emerge --ask \
    dev-lang/python \
    dev-python/pip \
    dev-python/virtualenv \
    dev-vcs/git

2. Create a virtual environment

mkdir -p ~/venvs
python3 -m venv ~/venvs/ollm
source ~/venvs/ollm/bin/activate

3. Upgrade pip

pip install --upgrade pip setuptools wheel

4. Install oLLM

pip install --no-build-isolation ollm

Or install the latest development version:

git clone https://github.com/Mega4alik/ollm.git
cd ollm
pip install --no-build-isolation -e .

The project also documents optional CUDA acceleration packages for NVIDIA GPUs, but those are unnecessary on your Intel Arc system.

5. Verify

python -c "import ollm; print('oLLM installed successfully')"

6. Activate when needed

source ~/venvs/ollm/bin/activate

Deactivate with:

deactivate

Since you've previously configured Intel Arc, oneAPI/SYCL, and local AI tooling on your Gentoo system, oLLM will run on the CPU unless its supported backends are configured appropriately. The upstream examples primarily target CUDA-enabled NVIDIA systems.

If you actually meant Ollama instead of oLLM, let me know—I can provide the current Gentoo installation instructions for Ollama.

Comments

No comments yet.

Title *

Slug

Excerpt

Content *

HTML is supported.

SEO Meta Data

Meta Title

Meta Description