Spaces:
Sleeping
Sleeping
Update PDF_Reader.py
Browse files- PDF_Reader.py +16 -5
PDF_Reader.py
CHANGED
@@ -1,17 +1,28 @@
|
|
1 |
import os
|
2 |
from together import Together
|
|
|
3 |
|
4 |
client = Together(api_key=os.environ.get("TOGETHER_API_KEY"))
|
5 |
|
6 |
-
def process_text(question,
|
|
|
|
|
7 |
response = client.chat.completions.create(
|
8 |
model="meta-llama/Llama-3-8b-chat-hf",
|
9 |
messages=[{"role": "user", "content": question}],
|
10 |
-
context=
|
11 |
)
|
12 |
return response.choices[0].message.content
|
13 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
if __name__ == "__main__":
|
15 |
-
|
16 |
-
question = "What can I learn about New York from
|
17 |
-
print(process_text(question,
|
|
|
1 |
import os
|
2 |
from together import Together
|
3 |
+
import PyPDF2
|
4 |
|
5 |
client = Together(api_key=os.environ.get("TOGETHER_API_KEY"))
|
6 |
|
7 |
+
def process_text(question, pdf_paths):
|
8 |
+
texts = [get_pdf_content(pdf) for pdf in pdf_paths]
|
9 |
+
combined_text = "\n".join(texts)
|
10 |
response = client.chat.completions.create(
|
11 |
model="meta-llama/Llama-3-8b-chat-hf",
|
12 |
messages=[{"role": "user", "content": question}],
|
13 |
+
context=combined_text
|
14 |
)
|
15 |
return response.choices[0].message.content
|
16 |
|
17 |
+
def get_pdf_content(pdf_path):
|
18 |
+
with open(pdf_path, "rb") as file:
|
19 |
+
reader = PyPDF2.PdfReader(file)
|
20 |
+
text = ""
|
21 |
+
for page in reader.pages:
|
22 |
+
text += page.extract_text()
|
23 |
+
return text
|
24 |
+
|
25 |
if __name__ == "__main__":
|
26 |
+
pdf_paths = ["doc1.pdf", "doc2.pdf"] # Adjust paths as necessary
|
27 |
+
question = "What can I learn about New York from these texts?"
|
28 |
+
print(process_text(question, pdf_paths))
|