Fixed app v2

app.py

@@ -1,58 +1,63 @@
-import uvicorn
 from fastapi import FastAPI, HTTPException
 from pydantic import BaseModel
 from llama_cpp import Llama
-import json
 
 # Initialize FastAPI app
 app = FastAPI()
 
-# Define request model for job description comparison
-class CompareRequest(BaseModel):
+# Load the SecurityLLM model
+try:
+    llm = Llama.from_pretrained(
+        repo_id="QuantFactory/SecurityLLM-GGUF",
+        filename="SecurityLLM.Q5_K_M.gguf",  # Ensure the file path is correct
+    )
+except Exception as e:
+    raise RuntimeError(f"Failed to load model: {e}")
+
+# Define request model for CV matching
+class CVMatchRequest(BaseModel):
     cv_text: str
     job_descriptions: str
 
 # Define response model
-class AnalysisResponse(BaseModel):
+class CVMatchResponse(BaseModel):
     results: list
 
-# Endpoint to compare job descriptions with the CV text
-@app.post("/compare/", response_model=AnalysisResponse)
-async def compare_job_cv(request: CompareRequest):
+# Define the route for CV and job description matching
+@app.post("/match_cv_to_jobs", response_model=CVMatchResponse)
+async def match_cv_to_jobs(request: CVMatchRequest):
     try:
-        # Load the Llama model within the endpoint to control its lifecycle
-        with Llama.from_pretrained(
-            repo_id="HuggingFaceTB/SmolLM2-360M-Instruct-GGUF",
-            filename="smollm2-360m-instruct-q8_0.gguf"  # Replace with the correct path to your GGUF file
-        ) as llm:
-            
-            # Split job descriptions by line
-            descriptions = request.job_descriptions.strip().split("\n")
-            results = []
+        # Split job descriptions by line
+        descriptions = request.job_descriptions.strip().split("\n")
+        results = []
+        
+        for description in descriptions:
+            # Create a prompt to compare the CV with each job description
+            prompt = (
+                f"Compare the following job description with this resume. Job Description: {description}. "
+                f"Resume: {request.cv_text}. Provide a match score and a brief analysis."
+            )
             
-            for description in descriptions:
-                # Create chat messages for each job description
-                messages = [
-                    {"role": "user", "content": f"Compare the following job description with this resume. Job Description: {description}. Resume: {request.cv_text}. Give a match score and brief analysis."}
+            # Generate response from the model
+            response = llm.create_chat_completion(
+                messages=[
+                    {
+                        "role": "user",
+                        "content": prompt
+                    }
                 ]
-                
-                # Generate response using Llama
-                response = llm.create_chat_completion(messages=messages)
-                response_content = response["choices"][0]["message"]["content"]
-                
-                # Parse response content for a score and summary
-                try:
-                    response_data = json.loads(response_content)
-                    results.append(response_data)
-                except json.JSONDecodeError:
-                    results.append({
-                        "Job Description": description,
-                        "Analysis": response_content  # Use raw response if JSON parsing fails
-                    })
+            )
+            
+            # Extract and return the analysis text
+            analysis_text = response["choices"][0]["message"]["content"]
+            results.append({
+                "Job Description": description,
+                "Analysis": analysis_text
+            })
         
-        return AnalysisResponse(results=results)
+        return CVMatchResponse(results=results)
     
     except Exception as e:
         raise HTTPException(status_code=500, detail=str(e))
 
-# Run the app with: uvicorn main:app --reload
+# To run the app, use: uvicorn app:app --reload

requirements.txt

@@ -2,4 +2,5 @@ fastapi
 uvicorn
 transformers
 git+https://github.com/abetlen/llama-cpp-python.git
-nest_asyncio
+nest_asyncio
+pydantic