Just in case using local LLMs doesn't work well with huggingface, there are a few options:
- ScaDS.AI has a dedicated LLM server here, which is only accessible using TU Dresden VPN.
- Alternatively, we could use the LLM server from KISSKI, a service provided for free to German academics.
- I presume also ollama can be installed in docker.
You can test API connections to the services as demonstrated here:
Just in case using local LLMs doesn't work well with huggingface, there are a few options:
You can test API connections to the services as demonstrated here: