Making the embedding work with ollama on local network

[jdrees]

I've been trying to get the docker .env file to use my ollama instance on my local network to to the embedding but I can't get it to work. It always tries to find models on huggingface and I don't know which variables to set in order for the container to start using the correct parameters. The ollama instance works as I'm already using it as the LLM. Just the embedding doesn't work.
I've tried:
EMBEDDINGS_MODEL_NAME=nomic-embed-text
EMBEDDINGS_MODEL_PROVIDER=ollama
LLM_MODE=ollama
OLLAMA_EMBEDDING_MODEL=nomic-embed-text
LLM_TOKENIZER=nomic-embed-text
OLLAMA_API_BASE="http://192.168.x.x:11434"
OLLAMA_LLM_MODEL=llama3.1:8b

But it still always give me errors that it can't access or find the models on huggingface.
What would be the correct set of config values in order to get it to work?

[3x3cut0r]

Hi, maybe the you need to set OLLAMA_EMBEDDING_API_BASE="http://192.168.x.x:11434/". Normally, this should be automatically set to OLLAMA_API_BASE if empty, but it does not seem to work correctly.
Look at my README for all working OLLAMA parameters.