Merge branch 'mha-train-develop' into attn-train-develop-qloop-mask

8dd7156d · ltqin · d5f629e7 · b5a3ea2d · 8dd7156d · 8dd7156d
Commit 8dd7156d authored Jul 25, 2023 by ltqin
20 changed files
--- a/include/ck/tensor_operation/gpu/device/device_batchnorm_forward.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batchnorm_forward.hpp
--- a/include/ck/tensor_operation/gpu/device/device_batchnorm_infer.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batchnorm_infer.hpp
--- a/include/ck/tensor_operation/gpu/device/device_cgemm.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_cgemm.hpp
--- a/include/ck/tensor_operation/gpu/device/device_contraction_multiple_d.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_contraction_multiple_d.hpp
--- a/include/ck/tensor_operation/gpu/device/device_conv_bwd_data.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_bwd_data.hpp
--- a/include/ck/tensor_operation/gpu/device/device_conv_fwd.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_fwd.hpp
--- a/include/ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation.hpp
--- a/include/ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation_add.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation_add.hpp
--- a/include/ck/tensor_operation/gpu/device/device_elementwise_base.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_elementwise_base.hpp
--- a/include/ck/tensor_operation/gpu/device/device_elementwise_normalization.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_elementwise_normalization.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_bias_e_permute.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_bias_e_permute.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_layernorm.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_layernorm.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_multiple_r.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_multiple_r.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_reduce.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_reduce.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_splitk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_splitk.hpp
--- a/include/ck/tensor_operation/gpu/device/device_grouped_contraction_multiple_d.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_grouped_contraction_multiple_d.hpp
--- a/include/ck/tensor_operation/gpu/device/device_grouped_conv_bwd_data_multiple_d.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_grouped_conv_bwd_data_multiple_d.hpp
--- a/include/ck/tensor_operation/gpu/device/device_grouped_conv_bwd_weight.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_grouped_conv_bwd_weight.hpp