from fastapi import FastAPI, Request, Depends from fastapi.responses import HTMLResponse, StreamingResponse from fastapi.staticfiles import StaticFiles from modules.pmbl import PMBL import torch print(f"CUDA available: {torch.cuda.is_available()}") print(f"CUDA device count: {torch.cuda.device_count()}") if torch.cuda.is_available(): print(f"CUDA device name: {torch.cuda.get_device_name(0)}") app = FastAPI(docs_url=None, redoc_url=None) app.mount("/static", StaticFiles(directory="static"), name="static") app.mount("/templates", StaticFiles(directory="templates"), name="templates") async def get_pmbl(): pmbl = PMBL("./PMB-7b.Q6_K.gguf", gpu_layers=50) try: yield pmbl finally: await pmbl.close() @app.head("/") @app.get("/") def index() -> HTMLResponse: with open("templates/index.html") as f: return HTMLResponse(content=f.read()) @app.post("/chat") async def chat(request: Request, pmbl: PMBL = Depends(get_pmbl)): try: data = await request.json() user_input = data["user_input"] mode = data["mode"] history = await pmbl.get_chat_history(mode, user_input) response_generator = pmbl.generate_response(user_input, history, mode) return StreamingResponse(response_generator, media_type="text/plain") except Exception as e: print(f"[SYSTEM] Error: {str(e)}") return {"error": str(e)} @app.post("/sleep") async def sleep(pmbl: PMBL = Depends(get_pmbl)): try: await pmbl.sleep_mode() return {"message": "Sleep mode completed successfully"} except Exception as e: print(f"[SYSTEM] Error: {str(e)}") return {"error": str(e)}