Spaces:
Sleeping
Sleeping
File size: 3,632 Bytes
676fec3 65c4b7f 676fec3 c802035 676fec3 65c4b7f 676fec3 65c4b7f 676fec3 65c4b7f 676fec3 65c4b7f 676fec3 65c4b7f 676fec3 65c4b7f 676fec3 65c4b7f 676fec3 65c4b7f 676fec3 65c4b7f 676fec3 65c4b7f |
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 |
import gradio as gr
from huggingface_hub import InferenceClient
import fitz # PyMuPDF
client = InferenceClient("opennyaiorg/Aalap-Mistral-7B-v0.1-bf16")
def extract_text_from_pdf(pdf_file):
document = fitz.open(pdf_file.name)
text = ""
for page_num in range(len(document)):
page = document.load_page(page_num)
text += page.get_text()
return text
def summarize_pdf(pdf_file, max_tokens, temperature, top_p):
text = extract_text_from_pdf(pdf_file)
response = ""
messages = [{"role": "user", "content": f"Summarize the following text: {text}"}]
for message in client.chat_completion(
messages,
max_tokens=max_tokens,
stream=True,
temperature=temperature,
top_p=top_p,
):
token = message.choices[0].delta.content
response += token
yield response
def ner_pdf(pdf_file, max_tokens, temperature, top_p):
text = extract_text_from_pdf(pdf_file)
response = ""
messages = [{"role": "user", "content": f"Extract named entities from the following text: {text}"}]
for message in client.chat_completion(
messages,
max_tokens=max_tokens,
stream=True,
temperature=temperature,
top_p=top_p,
):
token = message.choices[0].delta.content
response += token
yield response
def qa_pdf(pdf_file, question, max_tokens, temperature, top_p):
text = extract_text_from_pdf(pdf_file)
response = ""
messages = [{"role": "user", "content": f"Answer the question '{question}' based on the following text: {text}"}]
for message in client.chat_completion(
messages,
max_tokens=max_tokens,
stream=True,
temperature=temperature,
top_p=top_p,
):
token = message.choices[0].delta.content
response += token
yield response
with gr.Blocks() as demo:
gr.Markdown("# NLP Tasks on PDF Documents")
with gr.Tab("Summarization"):
pdf_file = gr.File(label="Upload PDF")
summarize_button = gr.Button("Summarize")
summary_output = gr.Textbox(label="Summary")
summarize_button.click(summarize_pdf, inputs=[pdf_file, gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"), gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"), gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")], outputs=summary_output)
with gr.Tab("Named Entity Recognition (NER)"):
pdf_file = gr.File(label="Upload PDF")
ner_button = gr.Button("Extract Entities")
ner_output = gr.JSON(label="Entities")
ner_button.click(ner_pdf, inputs=[pdf_file, gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"), gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"), gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")], outputs=ner_output)
with gr.Tab("Question Answering"):
pdf_file = gr.File(label="Upload PDF")
question_input = gr.Textbox(label="Enter your question")
qa_button = gr.Button("Get Answer")
qa_output = gr.Textbox(label="Answer")
qa_button.click(qa_pdf, inputs=[pdf_file, question_input, gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"), gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"), gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")], outputs=qa_output)
if __name__ == "__main__":
demo.launch()
|