我正在尝试从“ https://github.com/reedscot/cvpr2016”中获取火炬的预训练网络(#####预训练模型:Char-CNN-RNN用于鸟类)但是当我尝试使用它时我收到以下错误
THCudaCheck FAIL文件= / usr / local / torch / extra / cunn / lib / THCUNN / generic / Threshold.cu line = 34 error = 8:无效的设备功能 / usr / local / torch / install / bin / luajit:/usr/local/torch/install/share/lua/5.1/nn/Container.lua:67: 在nn.Sequential的2个模块中: /usr/local/torch/install/share/lua/5.1/nn/THNN.lua:110:cuda运行时错误(8):/ usr / local / torch / extra / cunn / lib / THCUNN / generic处的设备功能无效/Threshold.cu:34 堆栈回溯: [C]:在函数“ v”中 /usr/local/torch/install/share/lua/5.1/nn/THNN.lua:110:在函数'Threshold_updateOutput'中 /usr/local/torch/install/share/lua/5.1/nn/Threshold.lua:20:在函数中 [C]:在函数“ xpcall”中 /usr/local/torch/install/share/lua/5.1/nn/Container.lua:63:在函数'rethrowErrors'中 /usr/local/torch/install/share/lua/5.1/nn/Sequential.lua:44:在函数'func'中 /usr/local/torch/install/share/lua/5.1/nngraph/gmodule.lua:345:在'neteval'函数中 /usr/local/torch/install/share/lua/5.1/nngraph/gmodule.lua:380:在函数'forward'中 test.lua:130:在函数'extract_txt'中 test.lua:136:在主要块中 [C]:在函数“ dofile”中 ... ocal / torch / install / lib / luarocks / rocks / trepl / scm-1 / bin / th:150:在主要块中 [C]:位于0x00406670
CUDA版本有什么问题吗?