thamognya
/
internet_ml
1
1
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

working

main
Thamognya Kodi 2023-01-19 14:47:35 +07:00
parent 091e0b27e6
commit 5939e7804e
No known key found for this signature in database
GPG Key ID: 2701511C082EDAB5
2 changed files with 6 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
internet_ml/NLP/no_context/QA.py
View File

@ -47,7 +47,7 @@ def answer(
results: tuple[list[str], list[str]] = internet.Google(
query, GOOGLE_SEARCH_API_KEY, GOOGLE_SEARCH_ENGINE_ID
).google()
context: str = str(" ".join([str(string) for string in results[0]]))
context: str = str(" ".join([str(string) for string in results]))
print(f"context: {context}")
if model.startswith("openai-"):
@ -90,6 +90,6 @@ def answer(
print(
answer(
query="What is the newest pokemon game?",
model="hf-deepset/xlm-roberta-large-squad2",
model="openai-chatgpt",
)
)

9
internet_ml/tools/NLP/data/internet.py
View File

@ -100,9 +100,7 @@ class Google:
soup = BeautifulSoup(html, "html.parser")
text = soup.get_text()
normalized_text = normalizer(text)
sentences: list[str] = sentencizer(normalized_text)
sentence: str = str(" ".join(sentences))
return sentence
return normalized_text
except Exception:
error: str = ""
return error
@ -118,8 +116,9 @@ class Google:
def __get_urls_contents(self: "Google") -> None:
loop = asyncio.new_event_loop()
asyncio.set_event_loop(loop)
self.__content = loop.run_until_complete(self.__fetch_urls(self.__urls))
contents = loop.run_until_complete(self.__fetch_urls(self.__urls))
loop.close()
self.__content = contents
def __filter_irrelevant_processing(self: "Google") -> None:
with concurrent.futures.ThreadPoolExecutor(max_workers=500) as executor:
@ -137,7 +136,7 @@ class Google:
self.__get_urls_contents()
if filter_irrelevant:
self.__filter_irrelevant_processing()
results: tuple[list[str], list[str]] = (self.__content[0], self.__urls) # type: ignore
results: tuple[list[str], list[str]] = (self.__content, self.__urls) # type: ignore
return results