This repository has been archived by the owner on Jan 24, 2024. It is now read-only.
add group convolution, depthwise convolution and tests #235
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
wenming2014
force-pushed
the
wy
branch
2 times, most recently
from
894ecf8 to
306e443
Compare
Superjomn
reviewed
Sep 25, 2020
cinn/hlir/op/nn.cc
Outdated
| CHECK_EQ(padding.size(), 2) << "The size of padding in conv2d op is not 2! Please check."; | ||
| CHECK_EQ(stride.size(), 2) << "The size of stride in conv2d op is not 2! Please check."; | ||
| CHECK_GE(inputs_shape[0].size(), 3) << "The first input tensor's shape size of conv2d op is < 3! Please check."; | ||
| CHECK(data_format == "NCHW" || data_format == "NHWC") << "only support NCHW/NHWC data_format.\n"; |
cinn/hlir/pe/nn.cc
Outdated
| CHECK_EQ(4, output_shapes[2].size()) << "The size of output_shapes[2] of Conv2d op is not 4! Please check."; | ||
| std::vector<Expr> output_shape{ | ||
| Expr(output_shapes[2][0]), Expr(output_shapes[2][1]), Expr(output_shapes[2][2]), Expr(output_shapes[2][3])}; | ||
| CHECK_EQ(4, input->shape.size()) << "Input's dimension of Conv2d_NCHW op is not 4! Please check."; |
There was a problem hiding this comment.
这里不需要把 4 放到前面，不同于 4==a ，这里顺序不会有 bug，但可读性还是放到后面好一些
cinn/hlir/pe/nn.cc
Outdated
| std::vector<Expr> input_pad_shape; | ||
| if (output_shapes.size() == 3) { | ||
| // already computed by infer_shape | ||
| CHECK_EQ(4, output_shapes[0].size()) << "The size of output_shapes[0] of Conv2d op is not 4! Please check."; |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
add group convolution and nhwc data_format support for the former convolution op
add depthwise convolution op and PEs
add tests(#138 )