Skip to content

Download ollama model #12

Download ollama model

Download ollama model #12

Workflow file for this run

name: Download ollama model
on:
workflow_dispatch:
inputs:
model_name:
required: true
description: Specify model to be pulled by ollama such as llama3, phi3, etc.
cpu:
description: Specify a floating-point number of CPU cores, default to 2.0.
memory:
description: Specify minimum memory size required, default to 2048 MB.
timeout:
description: Specify execution timeout, default to 600 seconds.
keep_warm:
description: Specify to keep number of containers warm and running at all times, default to 1.
workflow_call:
jobs:
download:
runs-on: ubuntu-latest
env:
MODAL_USERNAME: ${{ secrets.MODAL_USERNAME }}
MODAL_TOKEN_ID: ${{ secrets.MODAL_TOKEN_ID }}
MODAL_TOKEN_SECRET: ${{ secrets.MODAL_TOKEN_SECRET }}
MODEL: ${{ github.event.inputs.model_name }}
CPU: ${{ github.event.inputs.cpu || '2.0' }}
MEMORY: ${{ github.event.inputs.memory || '2048' }}
TIMEOUT: ${{ github.event.inputs.timeout || '600' }}
KEEP_WARM: ${{ github.event.inputs.keep_warm || '1' }}
steps:
- name: Checkout repo
uses: actions/checkout@v4
- name: Setup Ollama
run: curl -fsSL https://ollama.com/install.sh | sh
- name: Pull a model
run: ollama pull ${{ github.event.inputs.model_name }}
- name: Copy model file to the current directory
run: |
./copy-model-file.sh /usr/share/ollama/.ollama/models/blobs ${{ github.event.inputs.model_name }}
ls -l
- name: Setup Python
uses: actions/setup-python@v4
with:
python-version: "3.11"
- name: Install Python packages
run: pip install modal-client
- name: Set APP_NAME variable
run: |
APP_NAME="${{ github.event.inputs.model_name }}"
echo "APP_NAME=${APP_NAME//:/-}" >> $GITHUB_ENV
- name: Deploy FastAPI app
run: |
echo "${APP_NAME}"
modal deploy modal_server
- name: Test FastAPI app
run: curl "https://${{ secrets.MODAL_USERNAME }}--${APP_NAME}-fastapi-app.modal.run/v1/models"