Skip to content

Instantly share code, notes, and snippets.

@Artem-B

Artem-B/ConvFwd_Add_Add_ReluFwd_eng15_k5=1_k6=0_k7=1_k10=1.log

Created May 11, 2023 17:24
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save Artem-B/c96d0ef86c333d02efa9f9dbf3b6a47b to your computer and use it in GitHub Desktop.
Save Artem-B/c96d0ef86c333d02efa9f9dbf3b6a47b to your computer and use it in GitHub Desktop.
Download ZIP
ConvFwd_Add_Add_ReluFwd_eng15_k5=1_k6=0_k7=1_k10=1
Raw
ConvFwd_Add_Add_ReluFwd_eng15_k5=1_k6=0_k7=1_k10=1.log
Do cudnn execution plan with plan tag: ConvFwd_Add_Add_ReluFwd_eng15_k5=1_k6=0_k7=1_k10=1
Workspace size in bytes: 409856
VariantPack: CUDNN_BACKEND_VARIANT_PACK_DESCRIPTOR : has 5 data pointers
I0509 11:27:06.030540 1504356 cuda_dnn.cc:4319]
Tensor_x: CUDNN_BACKEND_TENSOR_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Id: 120 Alignment: 32 nDims 4 VectorCount: 1 vectorDimension -1 Dim [ 1,64,7,9 ] Str [ 4032,63,9,1 ] isVirtual: 0 isByValue: 0 reorder_type: CUDNN_TENSOR_REORDERING_NONE
Tensor_y: CUDNN_BACKEND_TENSOR_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Id: 121 Alignment: 32 nDims 4 VectorCount: 1 vectorDimension -1 Dim [ 1,64,3,4 ] Str [ 768,12,4,1 ] isVirtual: 0 isByValue: 0 reorder_type: CUDNN_TENSOR_REORDERING_NONE
Tensor_z: CUDNN_BACKEND_TENSOR_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Id: 122 Alignment: 32 nDims 4 VectorCount: 1 vectorDimension -1 Dim [ 1,64,3,4 ] Str [ 768,12,4,1 ] isVirtual: 0 isByValue: 0 reorder_type: CUDNN_TENSOR_REORDERING_NONE
Tensor_w: CUDNN_BACKEND_TENSOR_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Id: 119 Alignment: 32 nDims 4 VectorCount: 1 vectorDimension -1 Dim [ 64,64,3,3 ] Str [ 576,9,3,1 ] isVirtual: 0 isByValue: 0 reorder_type: CUDNN_TENSOR_REORDERING_NONE
Tensor_b: CUDNN_BACKEND_TENSOR_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Id: 98 Alignment: 32 nDims 4 VectorCount: 1 vectorDimension -1 Dim [ 1,64,1,1 ] Str [ 64,1,1,1 ] isVirtual: 0 isByValue: 0 reorder_type: CUDNN_TENSOR_REORDERING_NONE
Tensor_conv: CUDNN_BACKEND_TENSOR_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Id: 67 Alignment: 32 nDims 4 VectorCount: 1 vectorDimension -1 Dim [ 1,64,3,4 ] Str [ 768,12,4,1 ] isVirtual: 1 isByValue: 0 reorder_type: CUDNN_TENSOR_REORDERING_NONE
Tensor_add: CUDNN_BACKEND_TENSOR_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Id: 65 Alignment: 32 nDims 4 VectorCount: 1 vectorDimension -1 Dim [ 1,64,3,4 ] Str [ 768,12,4,1 ] isVirtual: 1 isByValue: 0 reorder_type: CUDNN_TENSOR_REORDERING_NONE
Tensor_bias: CUDNN_BACKEND_TENSOR_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Id: 66 Alignment: 32 nDims 4 VectorCount: 1 vectorDimension -1 Dim [ 1,64,3,4 ] Str [ 768,12,4,1 ] isVirtual: 1 isByValue: 0 reorder_type: CUDNN_TENSOR_REORDERING_NONE
Conv: CUDNN_BACKEND_CONVOLUTION_DESCRIPTOR : Datatype: CUDNN_DATA_FLOAT Mode: 1 Num Dimensions: 2 PadLower [ 0,0 ] PadUpper [,0,0 ] Dilation [,1,1 ] Stride [,2,2]
Add: CUDNN_BACKEND_POINTWISE_DESCRIPTOR : Mode: CUDNN_POINTWISE_ADD Math precision CUDNN_DATA_FLOAT
BiasAdd: CUDNN_BACKEND_POINTWISE_DESCRIPTOR : Mode: CUDNN_POINTWISE_ADD Math precision CUDNN_DATA_FLOAT
Act: CUDNN_BACKEND_POINTWISE_DESCRIPTOR : Mode: CUDNN_POINTWISE_RELU_FWD Math precision CUDNN_DATA_FLOAT
ConvOp: CUDNN_BACKEND_OPERATION : OpMode: 10 X 0x16f318a4b698 Y 0x16f318a4b518 W 0x16f318a4b4d8 B (nil) T (nil) DW (nil) DY (nil) DX (nil) C 0x16f318a4a098 A Mtrix (nil) B Mtrix (nil) C Mtrix (nil) P (nil) MatMul (nil) Reduction (nil) alphabetaType 4 Alpha: 1 1 Alpha2: 1 1 Beta: 0 0
AddOp: CUDNN_BACKEND_OPERATION : OpMode: 13 X 0x16f318a4b518 Y 0x16f318a4b458 W (nil) B 0x16f318a4b5d8 T (nil) DW (nil) DY (nil) DX (nil) C (nil) A Mtrix (nil) B Mtrix (nil) C Mtrix (nil) P 0x16f318a4b418 MatMul (nil) Reduction (nil) alphabetaType 4 Alpha: 1 1 Alpha2: 0 0 Beta: 0 0
BiasAddOp: CUDNN_BACKEND_OPERATION : OpMode: 13 X 0x16f318a4b458 Y 0x16f318a4b498 W (nil) B 0x16f318a4b558 T (nil) DW (nil) DY (nil) DX (nil) C (nil) A Mtrix (nil) B Mtrix (nil) C Mtrix (nil) P 0x16f318a4a258 MatMul (nil) Reduction (nil) alphabetaType 4 Alpha: 1 1 Alpha2: 1 1 Beta: 0 0
ActOp: CUDNN_BACKEND_OPERATION : OpMode: 13 X 0x16f318a4b498 Y 0x16f318a4b618 W (nil) B (nil) T (nil) DW (nil) DY (nil) DX (nil) C (nil) A Mtrix (nil) B Mtrix (nil) C Mtrix (nil) P 0x16f318a4b318 MatMul (nil) Reduction (nil) alphabetaType 4 Alpha: 1 1 Alpha2: 1 1 Beta: 0 0
OpGraph: CUDNN_BACKEND_OPERATIONGRAPH_DESCRIPTOR has 4operations.
Tag: ConvFwd_Add_Add_ReluFwd_
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment