dor-rondel Ollama Model Management
This skill handles interaction with the Ollama service to update, add, or audit local LLM and embedding models.
🛠️ Service Target
- Host Address: http://127.0.0.1:11435
- Default Preloaded Models:
llama3,nomic-embed-text
🏃♂️ Core Commands
1. List Installed Models
Query the active Ollama engine inside the container to see what weights are already loaded on disk:
docker exec -it ollama ollama list
2. Download / Pull a New Model
Instruct the containerized daemon to pull down a new model directly from the Ollama registry:
docker exec -it ollama ollama pull <model-name>
3. Check Service Logs
If a model fails to load or download, inspect the execution logs directly from the active service:
docker-compose logs ollama