Auto-Tune with Templates and AutoTVM¶
AutoTVM offers a way to tune models and operators by providing a template schedule, and searcing the parameter space defined by the template. These how-tos demonstrate how to write template schedules and optimize them for a variety of different hardware platforms.
![Tuning High Performance Convolution on NVIDIA GPUs](../../_images/sphx_glr_tune_conv2d_cuda_thumb.png)
Tuning High Performance Convolution on NVIDIA GPUs
Tuning High Performance Convolution on NVIDIA GPUs
![Auto-tuning a Convolutional Network for NVIDIA GPU](../../_images/sphx_glr_tune_relay_cuda_thumb.png)
Auto-tuning a Convolutional Network for NVIDIA GPU
Auto-tuning a Convolutional Network for NVIDIA GPU
![Auto-tuning a Convolutional Network for Mobile GPU](../../_images/sphx_glr_tune_relay_mobile_gpu_thumb.png)
Auto-tuning a Convolutional Network for Mobile GPU
Auto-tuning a Convolutional Network for Mobile GPU