-
Install Ollama [linux example] curl -fsSL https://ollama.com/install.sh | sh
-
Create virtual environment:
$ python3 -m venv .venv
$ source .venv/bin/activate
$ pip install -r requirements.txt
- Download and run Ollama model:
$ ollama run llama2
$ ollama list -> will list downloaded models; models can be also imported
$ vim /etc/systemd/system/ollama.service
-
under [Service] add: Environment="https_proxy=http://mycorporateproxy.local:8080"
-
save the file
-
reset ollama service: $systemctl daemon-reload $systemctl restart ollama.service