Spaces:
Sleeping
Sleeping
File size: 2,438 Bytes
e4e2851 f222f88 e4e2851 f222f88 e4e2851 f222f88 e4e2851 f222f88 e4e2851 f222f88 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 |
from functools import wraps
import torch
from huggingface_hub import HfApi
import os
import logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
class DeviceManager:
_instance = None
def __new__(cls):
if cls._instance is None:
cls._instance = super(DeviceManager, cls).__new__(cls)
cls._instance._initialized = False
return cls._instance
def __init__(self):
if self._initialized:
return
self._initialized = True
self._current_device = None
def check_zero_gpu_availability(self):
try:
# 檢查 Hugging Face Space 環境變數
if not os.environ.get('SPACE_ID'):
return False
# 檢查是否在 Spaces 環境中並且啟用了 ZeroGPU
if os.environ.get('ZERO_GPU_AVAILABLE') == '1':
return True
return False
except Exception as e:
logger.warning(f"Error checking ZeroGPU availability: {e}")
return False
def get_optimal_device(self):
if self._current_device is None:
if self.check_zero_gpu_availability():
try:
# 確保 CUDA 可用
if torch.cuda.is_available():
self._current_device = torch.device('cuda')
logger.info("Using ZeroGPU")
else:
raise RuntimeError("CUDA not available")
except Exception as e:
logger.warning(f"Failed to initialize ZeroGPU: {e}")
self._current_device = torch.device('cpu')
logger.info("Fallback to CPU due to GPU initialization failure")
else:
self._current_device = torch.device('cpu')
logger.info("Using CPU (ZeroGPU not available)")
return self._current_device
def move_to_device(self, tensor_or_model):
device = self.get_optimal_device()
try:
if hasattr(tensor_or_model, 'to'):
return tensor_or_model.to(device)
except Exception:
self._current_device = torch.device('cpu')
if hasattr(tensor_or_model, 'to'):
return tensor_or_model.to('cpu')
return tensor_or_model |