39 lines (33 loc) · 885 Bytes

Try local VLM models on Ollama

Check if Ollama is running:
```
curl http://localhost:11434
```
Install dependencies
```
pip install opencv-python unidecode
```

Run the Python script 1.3.1.image_processing.py with the following content:

import ollama

response = ollama.chat(
    model='gemma3:12b',
    messages=[{
        'role': 'user',
        'content': 'Kas matoma paveikslėlyje?',
        'images': ['image.png']
    }]
)

print(response)

Run the script:
```
python3 ./1.3.1.image_processing.py
```

Let's try to run video processing with Ollama

Check if Ollama is running:
```
curl http://localhost:11434
```
Run the Python script 1.3.2.video_processing.py. Produce input video input.mp4.

Return README.md