You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
pred_depth, confidence, output_dict = model.inference({'input': rgb})
我将模型和上述代码中的rgb通过to(device)移到GPU:1后会在代码的其他部分报错:
Metric3D/mono/model/decode_heads/RAFTDepthNormalDPTDecoder5.py", line 215, in compute_depth_expectation
depth = torch.sum(prob * depth_values, 1)
~~~~~^~~~~~~~~~~~~~
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:1 and cuda:0!
也就是说,目前代码中的数据都是全部默认放置在GPU:0中的?
The text was updated successfully, but these errors were encountered:
pred_depth, confidence, output_dict = model.inference({'input': rgb})
我将模型和上述代码中的rgb通过to(device)移到GPU:1后会在代码的其他部分报错:
Metric3D/mono/model/decode_heads/RAFTDepthNormalDPTDecoder5.py", line 215, in compute_depth_expectation
depth = torch.sum(prob * depth_values, 1)
~~~~~^~~~~~~~~~~~~~
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:1 and cuda:0!
也就是说,目前代码中的数据都是全部默认放置在GPU:0中的?
The text was updated successfully, but these errors were encountered: