aboutsummaryrefslogtreecommitdiff
path: root/src/gpu/cl/kernels/ClDirectConv2dKernel.h
diff options
context:
space:
mode:
authorGiorgio Arena <giorgio.arena@arm.com>2021-10-13 11:13:04 +0100
committerGiorgio Arena <giorgio.arena@arm.com>2021-10-14 14:54:48 +0000
commit945ae9e4027655267170ecc56563c362d8110d1e (patch)
treed5c695462c57ca88bc628901e4b26b739d440651 /src/gpu/cl/kernels/ClDirectConv2dKernel.h
parentde23fc381aca403c94870d7f8bc78716eb350d53 (diff)
downloadComputeLibrary-945ae9e4027655267170ecc56563c362d8110d1e.tar.gz
Implement CLDirectConv3D f32/f16
Resolve COMPMID-4660 Signed-off-by: Giorgio Arena <giorgio.arena@arm.com> Change-Id: Ibd66ec1eb6faa60086981b1e3a9c12561df3445f Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/6420 Tested-by: Arm Jenkins <bsgcomp@arm.com> Comments-Addressed: Arm Jenkins <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Diffstat (limited to 'src/gpu/cl/kernels/ClDirectConv2dKernel.h')
-rw-r--r--src/gpu/cl/kernels/ClDirectConv2dKernel.h4
1 files changed, 2 insertions, 2 deletions
diff --git a/src/gpu/cl/kernels/ClDirectConv2dKernel.h b/src/gpu/cl/kernels/ClDirectConv2dKernel.h
index 4041c7bf27..5624f3a0a7 100644
--- a/src/gpu/cl/kernels/ClDirectConv2dKernel.h
+++ b/src/gpu/cl/kernels/ClDirectConv2dKernel.h
@@ -42,9 +42,9 @@ public:
ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(ClDirectConv2dKernel);
/** Set the src, weights, biases and dst tensors info.
*
- * @note: Due to set_valid_region(), thus src/weights/biases cannot be const. Need to change this once the set_valid_region() is removed.
+ * @note: Due to set_valid_region() in NCHW, src/weights/biases cannot be const. Need to change this once the set_valid_region() is removed.
*
- * @note: DirectConvolution only works in the following configurations:
+ * @note: DirectConvolution only works in the following configurations for the NCHW data layout:
* 1x1 convolution with stride_x = 1/2/3, stride_y = 1/2/3
* 3x3 convolution with stride_x = 1/2, stride_y = 1/2
* 5x5 convolution with stride_x = 1/2, stride_y = 1/2