Upload retrieve_docs.py with huggingface_hub
Browse files- retrieve_docs.py +5 -11
retrieve_docs.py
CHANGED
|
@@ -3,6 +3,7 @@
|
|
| 3 |
import json
|
| 4 |
import faiss
|
| 5 |
import numpy as np
|
|
|
|
| 6 |
|
| 7 |
INDEX_PATH = "data/docs/docs.index"
|
| 8 |
META_PATH = "data/docs/docs_meta.json"
|
|
@@ -17,16 +18,10 @@ QUERY_TEMPLATES = {
|
|
| 17 |
"file_not_found": "Jenkins pipeline workspace file not found",
|
| 18 |
"git_authentication_error": "Jenkins git authentication failed checkout"
|
| 19 |
}
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
def get_model():
|
| 25 |
-
global _model
|
| 26 |
-
if _model is None:
|
| 27 |
-
from sentence_transformers import SentenceTransformer
|
| 28 |
-
_model = SentenceTransformer("all-MiniLM-L6-v2", cache_folder="./models")
|
| 29 |
-
return _model
|
| 30 |
|
| 31 |
def retrieve_docs(error_category: str):
|
| 32 |
|
|
@@ -39,7 +34,6 @@ def retrieve_docs(error_category: str):
|
|
| 39 |
"Jenkins pipeline error"
|
| 40 |
)
|
| 41 |
|
| 42 |
-
model = get_model()
|
| 43 |
query_embedding = model.encode([query])
|
| 44 |
distances, indices = index.search(query_embedding, TOP_K)
|
| 45 |
|
|
|
|
| 3 |
import json
|
| 4 |
import faiss
|
| 5 |
import numpy as np
|
| 6 |
+
from sentence_transformers import SentenceTransformer
|
| 7 |
|
| 8 |
INDEX_PATH = "data/docs/docs.index"
|
| 9 |
META_PATH = "data/docs/docs_meta.json"
|
|
|
|
| 18 |
"file_not_found": "Jenkins pipeline workspace file not found",
|
| 19 |
"git_authentication_error": "Jenkins git authentication failed checkout"
|
| 20 |
}
|
| 21 |
+
model = SentenceTransformer(
|
| 22 |
+
"paraphrase-MiniLM-L3-v2",
|
| 23 |
+
cache_folder="./model_cache"
|
| 24 |
+
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
|
| 26 |
def retrieve_docs(error_category: str):
|
| 27 |
|
|
|
|
| 34 |
"Jenkins pipeline error"
|
| 35 |
)
|
| 36 |
|
|
|
|
| 37 |
query_embedding = model.encode([query])
|
| 38 |
distances, indices = index.search(query_embedding, TOP_K)
|
| 39 |
|