Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -7,6 +7,7 @@ from langchain.prompts import PromptTemplate
|
|
7 |
from pathlib import Path
|
8 |
from langchain_huggingface import ChatHuggingFace, HuggingFaceEndpoint
|
9 |
from langchain_core.output_parsers import JsonOutputParser
|
|
|
10 |
|
11 |
llm = HuggingFaceEndpoint(
|
12 |
repo_id="mistralai/Mistral-7B-Instruct-v0.3",
|
@@ -34,26 +35,7 @@ def summarize(file, n_words):
|
|
34 |
with open(file_path, 'r', encoding='utf-8') as f:
|
35 |
text = f.read()
|
36 |
|
37 |
-
|
38 |
-
Please carefully read the following document:
|
39 |
-
|
40 |
-
<document>
|
41 |
-
{TEXT}
|
42 |
-
</document>
|
43 |
-
|
44 |
-
identify the MOST used language in the document, return detected language in json format with key "language" and value is the detected language
|
45 |
-
'''
|
46 |
-
|
47 |
-
prompt_detect = PromptTemplate(
|
48 |
-
template=template_detect,
|
49 |
-
input_variables=['TEXT']
|
50 |
-
)
|
51 |
-
|
52 |
-
language_detect = prompt_detect | llm | JsonOutputParser()
|
53 |
-
formatted_prompt = prompt_detect.format(TEXT=text)
|
54 |
-
language = language_detect.invoke(formatted_prompt)
|
55 |
-
|
56 |
-
lang = language["language"]
|
57 |
template_translate = '''
|
58 |
Please carefully read the following document:
|
59 |
<document>
|
|
|
7 |
from pathlib import Path
|
8 |
from langchain_huggingface import ChatHuggingFace, HuggingFaceEndpoint
|
9 |
from langchain_core.output_parsers import JsonOutputParser
|
10 |
+
from langdetect import detect
|
11 |
|
12 |
llm = HuggingFaceEndpoint(
|
13 |
repo_id="mistralai/Mistral-7B-Instruct-v0.3",
|
|
|
35 |
with open(file_path, 'r', encoding='utf-8') as f:
|
36 |
text = f.read()
|
37 |
|
38 |
+
lang = detect(text)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
39 |
template_translate = '''
|
40 |
Please carefully read the following document:
|
41 |
<document>
|