simple-rag: models

Declared by version 0.1.0: 2.

Model	Type	Provider	Input	Output
qwen/qwen3-8b	instruct	huggingface	chat	chat
sentence-transformers/all-minilm-l6-v2	text-embedding	huggingface	text	vector/384

models.yaml

models:

  - name: qwen/qwen3-8b
    provider: huggingface
    type: instruct
    input: chat
    output: chat
    extra_env:
      LOAD_IN_4BIT: "1"
    description: >
      Small instruction model, for text summary.

  - name: sentence-transformers/all-minilm-l6-v2
    provider: huggingface
    type: text-embedding
    input: text
    output: vector/384
    parameters:
      vector_size: 384
    description: >
      Matches RAG_EMBEDDING_MODEL already set on the open-webui service
      in docker-compose.core.yaml -- no other config change needed for
      open-webui to use it.