when i was running neuraltalk2 on my raspberry pi, i have some problem on cpu model. According to Google, the model may have some problem on x86 or arm, and have to convert it. But, when i following the step on Google "th>> torch.load()" appeared an error ,cannot find nn.Sequential which require CUDA TOOL and cunn . i was confused, why a cpu model require gpu tool
thx for reply
when i was running neuraltalk2 on my raspberry pi, i have some problem on cpu model. According to Google, the model may have some problem on x86 or arm, and have to convert it. But, when i following the step on Google "th>> torch.load()" appeared an error ,cannot find nn.Sequential which require CUDA TOOL and cunn . i was confused, why a cpu model require gpu tool thx for reply