Update main.py

main.py

@@ -1,29 +1,46 @@
 from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel
 from llama_cpp import Llama
 from huggingface_hub import hf_hub_download
 
 app = FastAPI()
 
-# Downloads your specific model
+# Enable CORS so your website can call this API
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+# Download the model from your repo
+print("Fetching Aevis Model...")
 model_path = hf_hub_download(
     repo_id="devray11/Aevis-Medical-SLM",
     filename="DeepSeek-R1-Distill-Llama-8B.Q4_K_M.gguf"
 )
 
-# Initializes the AI engine
-llm = Llama(model_path=model_path, n_ctx=2048)
+# Load model (Optimized for 2GB RAM / CPU)
+llm = Llama(model_path=model_path, n_ctx=1024, n_threads=2)
 
 class Query(BaseModel):
     prompt: str
 
 @app.post("/generate")
 async def generate(query: Query):
-    # Formats the prompt exactly like your training data
+    # Prompt format based on your training
     fmt_prompt = f"### Instruction:\n{query.prompt}\n\n### Response:\n"
-    output = llm(fmt_prompt, max_tokens=512, stop=["###"])
+    
+    output = llm(
+        fmt_prompt, 
+        max_tokens=400, 
+        stop=["###"], 
+        echo=False
+    )
+    
     return {"response": output["choices"][0]["text"]}
 
 @app.get("/")
 def health():
-    return {"status": "Aevis API is Online"}
+    return {"status": "Aevis API is Online and Healthy"}