Create main.py

main.py

@@ -0,0 +1,60 @@
+import os
+from fastapi import FastAPI, HTTPException
+from pydantic import BaseModel
+from huggingface_hub import hf_hub_download
+from llama_cpp import Llama
+
+app = FastAPI()
+
+# --- CONFIGURATION ---
+REPO_ID = "CoolShotSystems/Axiom-3.1-Sovereign"
+FILENAME = "Meta-Llama-3.1-8B.Q4_K_M.gguf"
+
+# Global Model Variable
+axiom_model = None
+
+@app.on_event("startup")
+def load_model():
+    global axiom_model
+    print("📡 DOWNLOADING AXIOM TO HF SPACE...")
+    try:
+        # Download (If public space, no token needed for public model. 
+        # If model is private, we need HF_TOKEN secret)
+        model_path = hf_hub_download(
+            repo_id=REPO_ID,
+            filename=FILENAME,
+            token=os.environ.get("HF_TOKEN") 
+        )
+        print("🧠 LOADING INTO 16GB RAM...")
+        axiom_model = Llama(
+            model_path=model_path,
+            n_ctx=2048,
+            n_threads=2 # Standard for HF Free Tier
+        )
+        print("✅ AXIOM ONLINE")
+    except Exception as e:
+        print(f"❌ ERROR: {e}")
+
+class ChatRequest(BaseModel):
+    messages: list
+
+@app.get("/")
+def home():
+    return {"status": "Axiom Space Online"}
+
+@app.post("/v1/chat/completions")
+async def chat(request: ChatRequest):
+    if not axiom_model:
+        raise HTTPException(status_code=503, detail="Model loading...")
+    
+    prompt = "<|begin_of_text|>"
+    for msg in request.messages:
+        role = msg['role']
+        content = msg['content']
+        prompt += f"<|start_header_id|>{role}<|end_header_id|>\n\n{content}<|eot_id|>"
+    prompt += "<|start_header_id|>assistant<|end_header_id|>\n\n"
+    
+    output = axiom_model(
+        prompt, max_tokens=512, stop=["<|eot_id|>"], echo=False
+    )
+    return {"choices": [{"message": {"role": "assistant", "content": output['choices'][0]['text']}}]}