rakka-inference-runtime-litellm 0.2.5

LiteLLM proxy provider for rakka-inference — implements ModelRunner against the LiteLLM unified API gateway, fronting OpenAI-compatible endpoints for any provider LiteLLM supports.