File size: 1,389 Bytes
f18435c | 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 | # ingest_docs.py
import os
import json
import faiss
from sentence_transformers import SentenceTransformer
RAW_DOCS_DIR = "data/docs/raw"
INDEX_PATH = "data/docs/docs.index"
META_PATH = "data/docs/docs_meta.json"
CHUNK_SIZE = 400
def chunk_text(text, size):
chunks = []
for i in range(0, len(text), size):
chunk = text[i:i+size].strip()
if chunk:
chunks.append(chunk)
return chunks
def main():
model = SentenceTransformer("paraphrase-MiniLM-L3-v2", cache_folder="./model_cache")
documents = []
metadata = []
for fname in os.listdir(RAW_DOCS_DIR):
path = os.path.join(RAW_DOCS_DIR, fname)
with open(path, "r", encoding="utf-8") as f:
text = f.read()
chunks = chunk_text(text, CHUNK_SIZE)
for chunk in chunks:
documents.append(chunk)
metadata.append({
"source_file": fname,
"source": f"https://www.jenkins.io/doc/"
})
embeddings = model.encode(documents)
index = faiss.IndexFlatL2(embeddings.shape[1])
index.add(embeddings)
os.makedirs("data/docs", exist_ok=True)
faiss.write_index(index, INDEX_PATH)
with open(META_PATH, "w", encoding="utf-8") as f:
json.dump(metadata, f, indent=2)
print(f"Ingested {len(documents)} document chunks.")
if __name__ == "__main__":
main()
|