fix broadcast and reduce gradients distributeddataparallel and tensor hooks
This commit is contained in:
parent
ef84c08841
commit
e706b81007
12 changed files with 180 additions and 62 deletions
|
|
@ -81,11 +81,6 @@ extern "C" {
|
|||
else if ((strcmp(target_device, "cpu") == 0) && (strcmp(tensor->device, "cuda") == 0)) {
|
||||
cuda_to_cpu(tensor);
|
||||
}
|
||||
|
||||
else {
|
||||
fprintf(stderr, "Could not send tensor to device %d", device_id);
|
||||
exit(1);
|
||||
}
|
||||
}
|
||||
|
||||
Tensor* add_tensor(Tensor* tensor1, Tensor* tensor2) {
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue