mirror of
https://github.com/pjreddie/darknet.git
synced 2023-08-10 21:13:14 +03:00
Use half_float16 instead of float32 if defined both CUDNN and CUDNN_HALF. Use Tensor Cores.
This commit is contained in:
@ -242,6 +242,8 @@ struct layer{
|
||||
float * weights_gpu;
|
||||
float * weight_updates_gpu;
|
||||
|
||||
float * weights_gpu16;
|
||||
|
||||
float * biases_gpu;
|
||||
float * bias_updates_gpu;
|
||||
|
||||
|
Reference in New Issue
Block a user