uv run img-ask -h
python -m venv .venv
source .venv/bin/activate
pip install .
img-ask -h
If you don't end up with a runnable img-ask
command on your path, first make sure your virtual environment is activated.
If all else fails, you can always pip install ollama
and then python main.py -h
.
usage: img-ask [-h] [-s] [-m MODEL] prompt [image ...]
Prompt ollama with a collection of images
positional arguments:
prompt
image
options:
-h, --help show this help message and exit
-s, --streaming print each inference result as it is produced, instead
of all results at the end. (default: False)
-m MODEL, --model MODEL
the ollama vision model to use (default: moondream)
see 'https://ollama.com/search?c=vision' for
compatible models.