You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 
Jonathan Yan 9f70ebac64 Cast/ReLU dynamic shapelsgpu op supports int32 and int64 5 years ago
..
activation_gpu_kernel.cc Cast/ReLU dynamic shapelsgpu op supports int32 and int64 5 years ago
activation_gpu_kernel.h Cast/ReLU dynamic shapelsgpu op supports int32 and int64 5 years ago
activation_grad_kernel.cc support ops elu and elugrad on gpu 5 years ago
activation_grad_kernel.h gpu suppor dynamic shape 5 years ago
adam_gpu_kernel.cc mindspore path adjust 5 years ago
adam_gpu_kernel.h Unified code style 5 years ago
batchnorm_grad_gpu_kernel.cc Add 2 reserve output for BatchNormGrad v1 5 years ago
batchnorm_grad_gpu_kernel.h gpu suppor dynamic shape 5 years ago
bias_add_grad_gpu_kenel.cc mindspore path adjust 5 years ago
bias_add_grad_gpu_kenel.h gpu suppor dynamic shape 5 years ago
binary_cross_entropy_gpu_kernel.cc fixBCE 5 years ago
binary_cross_entropy_gpu_kernel.h fix bugs when shape is empty 5 years ago
binary_cross_entropy_grad_kernel.cc fixBCE 5 years ago
binary_cross_entropy_grad_kernel.h fix bugs when shape is empty 5 years ago
combine_momentum_gpu_kernel.cc gpu add combine mom fusion 5 years ago
combine_momentum_gpu_kernel.h gpu nhwc 5 years ago
conv2d_gpu_kernel.cc mindspore path adjust 5 years ago
conv2d_gpu_kernel.h gpu fix padding and dataset_heleper 5 years ago
conv2d_grad_filter_gpu_kernel.cc mindspore path adjust 5 years ago
conv2d_grad_filter_gpu_kernel.h gpu fix padding and dataset_heleper 5 years ago
conv2d_grad_input_gpu_kernel.cc mindspore path adjust 5 years ago
conv2d_grad_input_gpu_kernel.h gpu fix padding and dataset_heleper 5 years ago
ctcloss_gpu_kernel.cc add GPU CTCLoss 5 years ago
ctcloss_gpu_kernel.h add GPU CTCLoss 5 years ago
dropout_gpu_kernel.cc mindspore path adjust 5 years ago
dropout_gpu_kernel.h modify some bug and add test case for gpu dropout op 5 years ago
dropout_grad_kernel.cc mindspore path adjust 5 years ago
dropout_grad_kernel.h Unified code style 5 years ago
flatten_gpu_kernel.cc gpu reshepe add type 5 years ago
flatten_gpu_kernel.h gpu suppor dynamic shape 5 years ago
flatten_grad_gpu_kernel.cc mindspore path adjust 5 years ago
flatten_grad_gpu_kernel.h gpu suppor dynamic shape 5 years ago
ftrl_gpu_kernel.cc mindspore path adjust 5 years ago
ftrl_gpu_kernel.h Unified code style 5 years ago
fused_adam_weight_decay.cc mindspore path adjust 5 years ago
fused_adam_weight_decay.h fix gpu adam fusion kernel 5 years ago
fused_batch_norm_ex_gpu_kernel.cc add FusedBatchNoramEx gpu kernel 5 years ago
fused_batch_norm_ex_gpu_kernel.h gpu suppor dynamic shape 5 years ago
fused_batch_norm_gpu_kernel.cc gpu add fusion: replace momentum cast 5 years ago
fused_batch_norm_gpu_kernel.h gpu suppor dynamic shape 5 years ago
fused_batch_norm_grad_ex_gpu_kernel.cc add FusedBatchNormGradEx gpu kernel 5 years ago
fused_batch_norm_grad_ex_gpu_kernel.h gpu suppor dynamic shape 5 years ago
fused_batchnorm_grad_gpu_kernel.cc gpu add fusion: replace momentum cast 5 years ago
fused_batchnorm_grad_gpu_kernel.h gpu suppor dynamic shape 5 years ago
fused_scale_momentum_gpu_kernel.cc gpu resnet50 fusion 5 years ago
fused_scale_momentum_gpu_kernel.h gpu add combine mom fusion 5 years ago
fused_weightdecay_scale_momentum_gpu_kernel.cc gpu resnet50 fusion 5 years ago
fused_weightdecay_scale_momentum_gpu_kernel.h gpu add combine mom fusion 5 years ago
gelu_grad_kernel.cc mindspore path adjust 5 years ago
gelu_grad_kernel.h Unified code style 5 years ago
gelu_kernel.cc mindspore path adjust 5 years ago
gelu_kernel.h Unified code style 5 years ago
im2col_gpu_kernel.cc support img2col for resnet50_thor GPU 5 years ago
im2col_gpu_kernel.h gpu suppor dynamic shape 5 years ago
kl_div_loss_gpu_kernel.cc fixBCE 5 years ago
kl_div_loss_gpu_kernel.h fix BCE and KLDiv 5 years ago
kl_div_loss_grad_kernel.cc fixBCE 5 years ago
kl_div_loss_grad_kernel.h add KLDiv loss 5 years ago
l2_loss_gpu_kernel.cc add support to op L2Loss on gpu 5 years ago
l2_loss_gpu_kernel.h add support to op L2Loss on gpu 5 years ago
l2normalize_gpu_kernel.cc add l2normalize gpu kernel. 5 years ago
l2normalize_gpu_kernel.h support int64 shape 5 years ago
l2normalize_grad_gpu_kernel.cc new add l2normalize_grad gpu ops. 5 years ago
l2normalize_grad_gpu_kernel.h support int64 shape 5 years ago
layer_norm_gpu_kernel.cc mindspore path adjust 5 years ago
layer_norm_gpu_kernel.h support int64 shape 5 years ago
layer_norm_grad_gpu_kernel.cc mindspore path adjust 5 years ago
layer_norm_grad_gpu_kernel.h support int64 shape 5 years ago
lstm_gpu_kernel.cc mindspore path adjust 5 years ago
lstm_gpu_kernel.h gpu suppor dynamic shape 5 years ago
lstm_grad_data_gpu_kernel.cc mindspore path adjust 5 years ago
lstm_grad_data_gpu_kernel.h gpu suppor dynamic shape 5 years ago
lstm_grad_weight_gpu_kernel.cc mindspore path adjust 5 years ago
lstm_grad_weight_gpu_kernel.h gpu suppor dynamic shape 5 years ago
maxpool_with_argmax_gpu_kernel.cc add maxpool_with_argmax/grad cuda kernel 5 years ago
maxpool_with_argmax_gpu_kernel.h support int64 shape 5 years ago
maxpool_with_argmax_grad_gpu_kernel.cc add maxpool_with_argmax/grad cuda kernel 5 years ago
maxpool_with_argmax_grad_gpu_kernel.h support int64 shape 5 years ago
mirror_pad_gpu_kernel.cc Reworked mirrorPad 5 years ago
mirror_pad_gpu_kernel.h Reworked mirrorPad 5 years ago
mirror_pad_grad_gpu_kernel.cc Reworked mirrorPad 5 years ago
mirror_pad_grad_gpu_kernel.h Reworked mirrorPad 5 years ago
momentum_gpu_kernel.cc gpu add fusion: replace momentum cast 5 years ago
momentum_gpu_kernel.h GPU momentum supports use_nesterov 5 years ago
pad_gpu_kernel.cc Reworked mirrorPad 5 years ago
pad_gpu_kernel.h support int64 shape 5 years ago
pooling_gpu_kernel.cc mindspore path adjust 5 years ago
pooling_gpu_kernel.h gpu suppor dynamic shape 5 years ago
pooling_grad_gpu_kernel.cc mindspore path adjust 5 years ago
pooling_grad_gpu_kernel.h gpu suppor dynamic shape 5 years ago
rmsprop_gpu_kernel.cc mindspore path adjust 5 years ago
rmsprop_gpu_kernel.h Unified code style 5 years ago
roi_align_gpu_kernel.cc code warning clean 5 years ago
roi_align_gpu_kernel.h support int64 shape 5 years ago
roi_align_grad_gpu_kernel.cc code warning clean 5 years ago
roi_align_grad_gpu_kernel.h support int64 shape 5 years ago
sgd_gpu_kernel.cc gpu ops code and test case. 5 years ago
sgd_gpu_kernel.h gpu ops code and test case. 5 years ago
sigmoid_cross_entropy_with_logits_gpu_kernel.cc mindspore path adjust 5 years ago
sigmoid_cross_entropy_with_logits_gpu_kernel.h Unified code style 5 years ago
sigmoid_cross_entropy_with_logits_grad_gpu_kernel.cc mindspore path adjust 5 years ago
sigmoid_cross_entropy_with_logits_grad_gpu_kernel.h bugfix:SigmoidCrossEntropyWithLogitsGrad need multiply dout 5 years ago
smooth_l1_loss_gpu_kernel.cc mindspore path adjust 5 years ago
smooth_l1_loss_gpu_kernel.h modified documentation and gpu kernel for smoothL1Loss 5 years ago
smooth_l1_loss_grad_gpu_kernel.cc mindspore path adjust 5 years ago
smooth_l1_loss_grad_gpu_kernel.h modified documentation and gpu kernel for smoothL1Loss 5 years ago
softmax_cross_entropy_with_logits_gpu_kernel.cc mindspore path adjust 5 years ago
softmax_cross_entropy_with_logits_gpu_kernel.h gpu suppor dynamic shape 5 years ago
softmax_gpu_kernel.cc mindspore path adjust 5 years ago
softmax_gpu_kernel.h gpu suppor dynamic shape 5 years ago
softmax_grad_gpu_kernel.cc mindspore path adjust 5 years ago
softmax_grad_gpu_kernel.h gpu suppor dynamic shape 5 years ago
softplus_gpu_kernel.cc new add softplus and softplus grad gpu ops. 5 years ago
softplus_gpu_kernel.h new add softplus and softplus grad gpu ops. 5 years ago
softplus_grad_gpu_kernel.cc new add softplus and softplus grad gpu ops. 5 years ago
softplus_grad_gpu_kernel.h new add softplus and softplus grad gpu ops. 5 years ago
sparse_apply_proximal_adagrad_kernel.cc add GPU SparseApplyProximalAdagrad 5 years ago
sparse_apply_proximal_adagrad_kernel.h add GPU SparseApplyProximalAdagrad 5 years ago
sparse_ftrl_gpu_kernel.cc sparse apply ftrl gpu kernel 5 years ago
sparse_ftrl_gpu_kernel.h Adapt changes to GPU framework for ScatterAdd/Update and SparseFtrl 5 years ago
sparse_softmax_cross_entropy_with_logits_gpu_kernel.cc mindspore path adjust 5 years ago
sparse_softmax_cross_entropy_with_logits_gpu_kernel.h gpu suppor dynamic shape 5 years ago