Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,37 +2,66 @@ import gradio as gr
|
|
| 2 |
from huggingface_hub import InferenceClient
|
| 3 |
import PyPDF2
|
| 4 |
import io
|
| 5 |
-
from docx import Document #
|
| 6 |
|
|
|
|
| 7 |
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
| 8 |
|
| 9 |
-
def extract_text_from_pdf(
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
text
|
| 14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
|
| 16 |
-
def extract_text_from_docx(
|
| 17 |
-
|
| 18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
def parse_cv(file, job_description):
|
|
|
|
| 21 |
if file is None:
|
| 22 |
return "Please upload a CV file."
|
| 23 |
-
|
| 24 |
file_ext = file.name.split(".")[-1].lower()
|
| 25 |
-
|
| 26 |
-
|
|
|
|
|
|
|
|
|
|
| 27 |
if file_ext == "pdf":
|
| 28 |
text = extract_text_from_pdf(file_bytes)
|
| 29 |
elif file_ext == "docx":
|
| 30 |
text = extract_text_from_docx(file_bytes)
|
| 31 |
else:
|
| 32 |
return "Unsupported file format. Please upload a PDF or DOCX file."
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
|
| 34 |
-
prompt = f"Analyze the following CV against the job description provided. Provide a summary, an assessment of fit, and a score from 0 to 10.\n\nJob Description:\n{job_description}\n\nCandidate CV:\n{text}"
|
| 35 |
-
response = client.text_generation(prompt, max_tokens=512)
|
| 36 |
return response
|
| 37 |
|
| 38 |
def respond(
|
|
@@ -43,32 +72,40 @@ def respond(
|
|
| 43 |
temperature,
|
| 44 |
top_p,
|
| 45 |
):
|
|
|
|
|
|
|
|
|
|
| 46 |
messages = [{"role": "system", "content": system_message}]
|
| 47 |
|
| 48 |
-
for
|
| 49 |
-
if
|
| 50 |
-
messages.append({"role": "user", "content":
|
| 51 |
-
if
|
| 52 |
-
messages.append({"role": "assistant", "content":
|
| 53 |
|
| 54 |
messages.append({"role": "user", "content": message})
|
| 55 |
response = ""
|
| 56 |
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
|
|
|
|
|
|
|
|
|
| 67 |
|
|
|
|
| 68 |
demo = gr.Blocks()
|
| 69 |
|
| 70 |
with demo:
|
| 71 |
gr.Markdown("## AI-powered CV Analyzer and Chatbot")
|
|
|
|
| 72 |
with gr.Tab("Chatbot"):
|
| 73 |
chat_interface = gr.ChatInterface(
|
| 74 |
respond,
|
|
@@ -87,7 +124,9 @@ with demo:
|
|
| 87 |
)
|
| 88 |
|
| 89 |
with gr.Tab("CV Analyzer"):
|
| 90 |
-
gr.Markdown(
|
|
|
|
|
|
|
| 91 |
file_input = gr.File(label="Upload CV", type="file")
|
| 92 |
job_desc_input = gr.Textbox(label="Job Description", lines=5)
|
| 93 |
output_text = gr.Textbox(label="CV Analysis Report", lines=10)
|
|
@@ -96,4 +135,4 @@ with demo:
|
|
| 96 |
analyze_button.click(parse_cv, inputs=[file_input, job_desc_input], outputs=output_text)
|
| 97 |
|
| 98 |
if __name__ == "__main__":
|
| 99 |
-
demo.launch()
|
|
|
|
| 2 |
from huggingface_hub import InferenceClient
|
| 3 |
import PyPDF2
|
| 4 |
import io
|
| 5 |
+
from docx import Document # Make sure you have installed python-docx
|
| 6 |
|
| 7 |
+
# Initialize the client for Hugging Face inference.
|
| 8 |
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
| 9 |
|
| 10 |
+
def extract_text_from_pdf(pdf_file_bytes):
|
| 11 |
+
"""Extract text from a PDF file given as bytes."""
|
| 12 |
+
try:
|
| 13 |
+
pdf_reader = PyPDF2.PdfReader(io.BytesIO(pdf_file_bytes))
|
| 14 |
+
text = ""
|
| 15 |
+
for page in pdf_reader.pages:
|
| 16 |
+
page_text = page.extract_text()
|
| 17 |
+
if page_text:
|
| 18 |
+
text += page_text + "\n"
|
| 19 |
+
return text.strip() or "No text could be extracted from the PDF."
|
| 20 |
+
except Exception as e:
|
| 21 |
+
return f"Error reading PDF: {str(e)}"
|
| 22 |
|
| 23 |
+
def extract_text_from_docx(docx_file_bytes):
|
| 24 |
+
"""Extract text from a DOCX file given as bytes."""
|
| 25 |
+
try:
|
| 26 |
+
doc = Document(io.BytesIO(docx_file_bytes))
|
| 27 |
+
text = "\n".join(para.text for para in doc.paragraphs)
|
| 28 |
+
return text.strip() or "No text could be extracted from the DOCX file."
|
| 29 |
+
except Exception as e:
|
| 30 |
+
return f"Error reading DOCX: {str(e)}"
|
| 31 |
|
| 32 |
def parse_cv(file, job_description):
|
| 33 |
+
"""Analyze a CV (PDF or DOCX) against a job description and generate a report."""
|
| 34 |
if file is None:
|
| 35 |
return "Please upload a CV file."
|
| 36 |
+
|
| 37 |
file_ext = file.name.split(".")[-1].lower()
|
| 38 |
+
try:
|
| 39 |
+
file_bytes = file.read()
|
| 40 |
+
except Exception as e:
|
| 41 |
+
return f"Error reading the uploaded file: {str(e)}"
|
| 42 |
+
|
| 43 |
if file_ext == "pdf":
|
| 44 |
text = extract_text_from_pdf(file_bytes)
|
| 45 |
elif file_ext == "docx":
|
| 46 |
text = extract_text_from_docx(file_bytes)
|
| 47 |
else:
|
| 48 |
return "Unsupported file format. Please upload a PDF or DOCX file."
|
| 49 |
+
|
| 50 |
+
if text.startswith("Error reading"):
|
| 51 |
+
return text # Return error from extraction if any.
|
| 52 |
+
|
| 53 |
+
prompt = (
|
| 54 |
+
f"Analyze the following CV against the provided job description. "
|
| 55 |
+
f"Provide a summary, an assessment of fit, and a score from 0 to 10.\n\n"
|
| 56 |
+
f"Job Description:\n{job_description}\n\n"
|
| 57 |
+
f"Candidate CV:\n{text}"
|
| 58 |
+
)
|
| 59 |
+
|
| 60 |
+
try:
|
| 61 |
+
response = client.text_generation(prompt, max_tokens=512)
|
| 62 |
+
except Exception as e:
|
| 63 |
+
return f"Error during CV analysis: {str(e)}"
|
| 64 |
|
|
|
|
|
|
|
| 65 |
return response
|
| 66 |
|
| 67 |
def respond(
|
|
|
|
| 72 |
temperature,
|
| 73 |
top_p,
|
| 74 |
):
|
| 75 |
+
"""
|
| 76 |
+
Chatbot response generator that interacts with a conversational model.
|
| 77 |
+
"""
|
| 78 |
messages = [{"role": "system", "content": system_message}]
|
| 79 |
|
| 80 |
+
for user_msg, bot_msg in history:
|
| 81 |
+
if user_msg:
|
| 82 |
+
messages.append({"role": "user", "content": user_msg})
|
| 83 |
+
if bot_msg:
|
| 84 |
+
messages.append({"role": "assistant", "content": bot_msg})
|
| 85 |
|
| 86 |
messages.append({"role": "user", "content": message})
|
| 87 |
response = ""
|
| 88 |
|
| 89 |
+
try:
|
| 90 |
+
for message_chunk in client.chat_completion(
|
| 91 |
+
messages,
|
| 92 |
+
max_tokens=max_tokens,
|
| 93 |
+
stream=True,
|
| 94 |
+
temperature=temperature,
|
| 95 |
+
top_p=top_p,
|
| 96 |
+
):
|
| 97 |
+
token = message_chunk.choices[0].delta.content
|
| 98 |
+
response += token
|
| 99 |
+
yield response
|
| 100 |
+
except Exception as e:
|
| 101 |
+
yield f"Error during chat generation: {str(e)}"
|
| 102 |
|
| 103 |
+
# Build the Gradio interface
|
| 104 |
demo = gr.Blocks()
|
| 105 |
|
| 106 |
with demo:
|
| 107 |
gr.Markdown("## AI-powered CV Analyzer and Chatbot")
|
| 108 |
+
|
| 109 |
with gr.Tab("Chatbot"):
|
| 110 |
chat_interface = gr.ChatInterface(
|
| 111 |
respond,
|
|
|
|
| 124 |
)
|
| 125 |
|
| 126 |
with gr.Tab("CV Analyzer"):
|
| 127 |
+
gr.Markdown(
|
| 128 |
+
"### Upload your CV (PDF or DOCX) and provide the job description to receive a professional analysis and suitability score."
|
| 129 |
+
)
|
| 130 |
file_input = gr.File(label="Upload CV", type="file")
|
| 131 |
job_desc_input = gr.Textbox(label="Job Description", lines=5)
|
| 132 |
output_text = gr.Textbox(label="CV Analysis Report", lines=10)
|
|
|
|
| 135 |
analyze_button.click(parse_cv, inputs=[file_input, job_desc_input], outputs=output_text)
|
| 136 |
|
| 137 |
if __name__ == "__main__":
|
| 138 |
+
demo.launch()
|