Update app.py
Browse files
app.py
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
from fastapi import FastAPI, HTTPException
|
| 2 |
from pydantic import BaseModel
|
| 3 |
-
from sentence_transformers import SentenceTransformer
|
| 4 |
from transformers import pipeline
|
| 5 |
|
| 6 |
# Initialize FastAPI app
|
|
@@ -17,7 +17,7 @@ class ModifyQueryRequest(BaseModel):
|
|
| 17 |
|
| 18 |
class AnswerQuestionRequest(BaseModel):
|
| 19 |
question: str
|
| 20 |
-
context:
|
| 21 |
|
| 22 |
# Define response models (if needed)
|
| 23 |
class ModifyQueryResponse(BaseModel):
|
|
@@ -25,6 +25,7 @@ class ModifyQueryResponse(BaseModel):
|
|
| 25 |
|
| 26 |
class AnswerQuestionResponse(BaseModel):
|
| 27 |
answer: str
|
|
|
|
| 28 |
|
| 29 |
# Define API endpoints
|
| 30 |
@app.post("/modify_query", response_model=ModifyQueryResponse)
|
|
@@ -38,12 +39,26 @@ async def modify_query(request: ModifyQueryRequest):
|
|
| 38 |
@app.post("/answer_question", response_model=AnswerQuestionResponse)
|
| 39 |
async def answer_question(request: AnswerQuestionRequest):
|
| 40 |
try:
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
except Exception as e:
|
| 48 |
raise HTTPException(status_code=500, detail=str(e))
|
| 49 |
|
|
|
|
| 1 |
from fastapi import FastAPI, HTTPException
|
| 2 |
from pydantic import BaseModel
|
| 3 |
+
from sentence_transformers import SentenceTransformer, utils
|
| 4 |
from transformers import pipeline
|
| 5 |
|
| 6 |
# Initialize FastAPI app
|
|
|
|
| 17 |
|
| 18 |
class AnswerQuestionRequest(BaseModel):
|
| 19 |
question: str
|
| 20 |
+
context: dict
|
| 21 |
|
| 22 |
# Define response models (if needed)
|
| 23 |
class ModifyQueryResponse(BaseModel):
|
|
|
|
| 25 |
|
| 26 |
class AnswerQuestionResponse(BaseModel):
|
| 27 |
answer: str
|
| 28 |
+
locations: list
|
| 29 |
|
| 30 |
# Define API endpoints
|
| 31 |
@app.post("/modify_query", response_model=ModifyQueryResponse)
|
|
|
|
| 39 |
@app.post("/answer_question", response_model=AnswerQuestionResponse)
|
| 40 |
async def answer_question(request: AnswerQuestionRequest):
|
| 41 |
try:
|
| 42 |
+
res_locs = []
|
| 43 |
+
context_string = ''
|
| 44 |
+
corpus_embeddings = model.encode(request.context['context'], convert_to_tensor=True)
|
| 45 |
+
query_embeddings = model.encode(request.question, convert_to_tensor=True)
|
| 46 |
+
hits = util.semantic_search(query_embeddings, corpus_embeddings)
|
| 47 |
+
for hit in hits:
|
| 48 |
+
if hit['score'] > .5:
|
| 49 |
+
loc = hit['corpus_id']
|
| 50 |
+
res_locs.append(request.context['locations'][loc])
|
| 51 |
+
context_string += request.context['context'][loc] + ' '
|
| 52 |
+
if len(res_locs) == 0:
|
| 53 |
+
ans = "Sorry, I couldn't find any results for your query."
|
| 54 |
+
else:
|
| 55 |
+
QA_input = {
|
| 56 |
+
'question': request.question,
|
| 57 |
+
'context': context_string.replace.('\n',' ')
|
| 58 |
+
}
|
| 59 |
+
result = nlp(QA_input)
|
| 60 |
+
ans = result['answer']
|
| 61 |
+
return AnswerQuestionResponse(answer=ans, locations = res_locs)
|
| 62 |
except Exception as e:
|
| 63 |
raise HTTPException(status_code=500, detail=str(e))
|
| 64 |
|