Self hosted version supports only Ollama.
vLLM is also OpenAPI compatible but it seems hardcoded to accept only ollama/llama2 as the only option when loading available models.
Suggestion: Please consider removing the hardcoded Ollama only option for custom provider.
Thank you
Self hosted version supports only Ollama.
vLLM is also OpenAPI compatible but it seems hardcoded to accept only ollama/llama2 as the only option when loading available models.
Suggestion: Please consider removing the hardcoded Ollama only option for custom provider.
Thank you