Spaces:
Build error
Build error
File size: 3,499 Bytes
915cbc7 |
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 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 |
from fastapi.routing import APIRouter
from pydantic import BaseModel
from Pipeline.HallucinationPipeline import HallucinationPipeline
from Pipeline.CorrectionLLMs import DeepseekAPI
from nltk.tokenize import sent_tokenize
from utils import detectionProcess
import os
deepseek_apikey="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"
router=APIRouter(prefix="/api")
pipeline=HallucinationPipeline("Razor2507/Roberta-Base-Finetuned","cpu")
deepseek=DeepseekAPI(api_key=deepseek_apikey)
# {'predictions': [0],
# 'corrected_summary': [],
# 'sent_predicted': [array([2, 0, 0, 0])],
# 'factual_score': [0.21776947937905788],
# 'contradiction_score': [0.7815729230642319]}
# Detection Endpoint
class DetectionRequest(BaseModel):
article:str
summary:str
arbiter:str
@router.post("/detect")
def detect(data:DetectionRequest):
try:
article=data.article.strip().replace("\n"," ").replace("\t"," ")
summary=data.summary.strip().replace("\n"," ").replace("\t"," ")
arbiter=True if data.arbiter=="on" else False
print("Arbiter : ",arbiter)
result=detectionProcess(article=article,summary=summary,pipeline=pipeline,arbiter=arbiter)
result["status"]=200
return result
except Exception as e:
print(e)
return {"status":404}
# Correction Endpoint
class correctionRequest(BaseModel):
article:str
tag_summary:str
model:str
@router.post("/correct")
def correct(data:correctionRequest):
try:
if data.model=="mistral":
pass
elif data.model=="gemini":
pass
elif data.model=="deepseek":
correction=deepseek.correct(premise=data.article, summary=data.tag_summary)
print(correction)
result=detectionProcess(article=data.article,summary=correction,pipeline=pipeline)
result["corrected_summary"]=correction
result["status"]=200
return result
except Exception as e:
print(e)
return {"status":404}
@router.get("/test")
def keyTest():
print("Testing ",os.getenv("deepseek_apikey"))
return {"msg":"testing_works"}
# @router.post("/detect")
# def detect(data:DetectionRequest):
# article=data.article.strip().replace("\n","").replace("\t"," ")
# summary=data.summary.strip().replace("\n","").replace("\t"," ")
# result=pipeline.process([[article,summary]],correct_the_summary=False)
# all_sentences=sent_tokenize(summary)
# print(result)
# summary=pipeline.addTags(all_sentences,result["sent_predicted"][0],len(all_sentences))
# score=str(result["factual_score"][0])
# sentenceLabels=list(result["sent_predicted"][0])
# labelCounts=[sentenceLabels.count(0),sentenceLabels.count(2)]
# prompt=f"""
# Here is a summary with hallucinated parts marked using <xx> tags.
# Please correct only the text inside the <xx> tags to make it factually accurate based on the original article. Leave the rest of the summary unchanged and remove the <xx> tags after correction.
# Return the summary with hallucinated parts fixed and you can remove those <xx></xx> tags. Don't remove that entire sentence.
# Original Article:
# {data.article}
# Summary:
# {summary}
# """
# return {"summary":summary,"score":score,"counts":labelCounts,"copy_prompt":prompt} |