Spaces:
Sleeping
Sleeping
| import os | |
| import sys | |
| from pathlib import Path | |
| import pytest | |
| ROOT = Path(__file__).resolve().parent.parent | |
| sys.path.insert(0, str(ROOT)) | |
| sys.path.insert(0, str(ROOT / "examples")) | |
| from mrag.config import Config # noqa: E402 | |
| def corpus(tmp_path_factory): | |
| import make_demo_corpus | |
| return make_demo_corpus.build(tmp_path_factory.mktemp("corpus")) | |
| def model_cfg(data_dir) -> Config: | |
| """Real models. Text: WordLlama (weights ship in the wheel) unless MRAG_TEXT_BACKEND says otherwise. | |
| CLIP: MRAG_CLIP_PRETRAINED (tag or local .pt), default laion400m_e32.""" | |
| return Config.load( | |
| data_dir=str(data_dir), | |
| text_backend=os.environ.get("MRAG_TEXT_BACKEND", "wordllama"), | |
| llm_backend="extractive", | |
| ) | |
| def rag(corpus, tmp_path_factory): | |
| from mrag.pipeline import MultimodalRAG | |
| r = MultimodalRAG(model_cfg(tmp_path_factory.mktemp("index"))) | |
| stats = r.ingest([corpus]) | |
| assert stats["errors"] == [], stats | |
| yield r | |
| r.close() | |