File size: 2,290 Bytes
82549b8 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 | #!/usr/bin/env python3
"""Smoke-test Hugging Face Inference API (embeddings + chat) using your .env.
Run from the repository root (so `.env` is found):
LLM_PROVIDER=huggingface uv run python scripts/verify_huggingface_inference.py
Or set `LLM_PROVIDER=huggingface` in `.env` together with `HUGGINGFACE_API_KEY`, then:
uv run python scripts/verify_huggingface_inference.py
"""
from __future__ import annotations
import os
import sys
from pathlib import Path
# Ensure repo root is cwd so pydantic-settings loads `.env`.
_ROOT = Path(__file__).resolve().parent.parent
os.chdir(_ROOT)
if str(_ROOT) not in sys.path:
sys.path.insert(0, str(_ROOT))
from langchain_core.messages import HumanMessage
from api.config import get_settings
from rag.embedder import create_embedding_function
from rag.retriever import _create_chat_model
def main() -> int:
get_settings.cache_clear()
settings = get_settings()
if settings.llm_provider.lower() != "huggingface":
print(
"Set LLM_PROVIDER=huggingface in your environment or `.env` before running this script.\n"
f"Current LLM_PROVIDER={settings.llm_provider!r}",
file=sys.stderr,
)
return 1
if not (settings.huggingface_api_key or "").strip():
print(
"Missing token: set HUGGINGFACE_API_KEY or HF_TOKEN in `.env`.\n"
"Current LLM_PROVIDER is huggingface but no Hugging Face token resolved.",
file=sys.stderr,
)
return 1
print("Model:", settings.huggingface_model)
print("Embedding model:", settings.huggingface_embedding_model)
print("--- Embeddings ---")
emb = create_embedding_function()
vec = emb.embed_query("hello from DocuAudit local smoke test")
print(f"OK: single query embedding length = {len(vec)}")
print("--- Chat (Inference API) ---")
llm = _create_chat_model(settings)
msg = HumanMessage(content='Reply with exactly the word "ok" and nothing else.')
out = llm.invoke([msg])
text = (getattr(out, "content", None) or str(out)).strip()
print("OK: chat response (first 200 chars):", text[:200])
print("\nHugging Face path looks good for deployment smoke checks.")
return 0
if __name__ == "__main__":
raise SystemExit(main())
|