0

我正在尝试在 Nvidia Jetson Nano 2 上导入 pytorch，但我面临以下问题。我使用的 pytorch 版本是 1.10.0，带有 torchvision 0.11.1 和 torchtext 0.6.0（如果相关）

我尝试了以下方法：

  • 更改导入顺序
  • 导出 $LD_PRELOAD='/home/msc/.local/lib/python3.6/site-packages/torch/lib/libgomp-d22c30c5.so.1' 后跟源 ~/.bashrc。
          2 import math
          3 import numpy as np
    ----> 4 import torch
          5 import torch.nn as nn
          6 import torch.nn.functional as F
    
    ~/.local/lib/python3.6/site-packages/torch/__init__.py in <module>
        194     # See Note [Global dependencies]
        195     if USE_GLOBAL_DEPS:
    --> 196         _load_global_deps()
        197     from torch._C import *  # noqa: F403
        198 
    
    ~/.local/lib/python3.6/site-packages/torch/__init__.py in _load_global_deps()
        147     lib_path = os.path.join(os.path.dirname(here), 'lib', lib_name)
        148 
    --> 149     ctypes.CDLL(lib_path, mode=ctypes.RTLD_GLOBAL)
        150 
        151 
    
    /usr/lib/python3.6/ctypes/__init__.py in __init__(self, name, mode, handle, use_errno, use_last_error)
        346 
        347         if handle is None:
    --> 348             self._handle = _dlopen(self._name, mode)
        349         else:
        350             self._handle = handle
    
    /home/msc/.local/lib/python3.6/site-packages/torch/lib/libgomp-d22c30c5.so.1: cannot allocate memory in static TLS block

但是，错误仍然存​​在。我真的很感激任何帮助！

这是磁盘空间的描述，以防万一。

    Filesystem      Size  Used Avail Use% Mounted on
    /dev/mmcblk0p1   30G   24G  4.7G  84% /
    none            947M     0  947M   0% /dev
    tmpfs           986M   92K  986M   1% /dev/shm
    tmpfs           986M   21M  966M   3% /run
    tmpfs           5.0M  4.0K  5.0M   1% /run/lock
    tmpfs           986M     0  986M   0% /sys/fs/cgroup
    tmpfs           198M  8.0K  198M   1% /run/user/1000

谢谢！

4

0 回答 0