Update app.py

app.py

@@ -1,9 +1,8 @@
 import gradio as gr
 import PyPDF2
 import docx
-from io import BytesIO
+import requests
 import json
-from mistral_inference import MistralInference  # Hypothetical API wrapper for Mistral model
 
 # Function to extract text from PDF
 def extract_text_from_pdf(file):
@@ -28,60 +27,54 @@ def process_uploaded_file(file):
     else:
         raise ValueError("Unsupported file format. Please upload a PDF or Word document.")
 
-# Hypothetical Mistral API wrapper (replace with actual API call)
-class MistralInference:
-    def __init__(self, model="mistralai/Mistral-7B-Instruct-v0.3", provider="together"):
-        self.model = model
-        self.provider = provider
+# Function to call Together API for Mistral inference
+def analyze_with_mistral(resume_text, job_description):
+    # Replace with your Together API key
+    TOGETHER_API_KEY = "your-together-api-key-here"
+    url = "https://api.together.xyz/inference"
+    
+    # Construct the prompt
+    prompt = f"""
+    Analyze the following resume against the job description for ATS compatibility.
+    Provide a detailed breakdown of ATS parameters (keywords, formatting, skills match, experience relevance, education) 
+    and assign a score out of 100 for each, along with an overall score. Return the result in JSON format.
 
-    def analyze(self, resume_text, job_description):
-        # Simulated prompt to Mistral model
-        prompt = f"""
-        Analyze the following resume against the job description for ATS compatibility.
-        Provide a detailed breakdown of ATS parameters (keywords, formatting, skills match, experience relevance, education) 
-        and assign a score out of 100 for each, along with an overall score. Return the result in JSON format.
+    Resume:
+    {resume_text}
 
-        Resume:
-        {resume_text}
-
-        Job Description:
-        {job_description}
-        """
-        
-        # Simulated response (replace with actual API call)
-        response = {
-            "ats_analysis": {
-                "keywords": {
-                    "score": 85,
-                    "details": "Key terms like 'Python', 'Machine Learning' found, missing 'AWS'."
-                },
-                "formatting": {
-                    "score": 90,
-                    "details": "Simple layout, readable by ATS, no complex tables."
-                },
-                "skills_match": {
-                    "score": 80,
-                    "details": "Strong match for programming skills, lacks cloud skills."
-                },
-                "experience_relevance": {
-                    "score": 75,
-                    "details": "Relevant experience in data science, but duration slightly short."
-                },
-                "education": {
-                    "score": 95,
-                    "details": "Matches required degree in Computer Science."
-                },
-                "overall_score": 85
-            }
-        }
-        return json.dumps(response, indent=2)
+    Job Description:
+    {job_description}
+    """
+    
+    # API request payload
+    payload = {
+        "model": "mistralai/Mistral-7B-Instruct-v0.3",
+        "prompt": prompt,
+        "max_tokens": 1000,
+        "temperature": 0.7,
+        "top_p": 0.9,
+    }
+    
+    headers = {
+        "Authorization": f"Bearer {TOGETHER_API_KEY}",
+        "Content-Type": "application/json",
+    }
+    
+    # Make the API call
+    response = requests.post(url, json=payload, headers=headers)
+    
+    if response.status_code == 200:
+        result = response.json()
+        # Assuming the API returns the generated text in 'output' field
+        return result.get("output", json.dumps({"error": "No output from API"}, indent=2))
+    else:
+        return json.dumps({"error": f"API request failed with status {response.status_code}: {response.text}"}, indent=2)
 
 # Main function to analyze resume
 def analyze_resume(file, job_description):
     try:
         resume_text = process_uploaded_file(file)
-        mistral = MistralInference()
-        result = mistral.analyze(resume_text, job_description)
+        result = analyze_with_mistral(resume_text, job_description)
         return result
     except Exception as e:
         return json.dumps({"error": str(e)}, indent=2)
@@ -108,7 +101,4 @@ with gr.Blocks(fill_height=True, title="Smart ATS Resume Analyzer") as demo:
         outputs=output
     )
 
-    # Load Mistral model (hypothetical, adjust based on actual integration)
-    gr.load("models/mistralai/Mistral-7B-Instruct-v0.3", accept_token=button, provider="together")
-
 demo.launch()