Implement ConditionallySpeculatable for ConvolutionOp #2228
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
|
Converting to draft while I fix conflicts. |
Relevant constraints from the spec: ``` (C10) dim(lhs, input_batch_dimension) % batch_group_count = 0. (C11) dim(lhs, input_feature_dimension) % feature_group_count = 0. (C14) dim(rhs, kernel_input_feature_dimension) = dim(lhs, input_feature_dimension) / feature_group_count. (C15) dim(rhs, kernel_output_feature_dimension) % batch_group_count = 0. (C16) dim(rhs, kernel_output_feature_dimension) % feature_group_count = 0. (C25) dim(result, result_dim) is defined as: * dim(lhs, input_batch_dimension) / batch_group_count if result_dim = output_batch_dimension. * dim(rhs, kernel_output_feature_dimension) if result_dim = output_feature_dimension. * num_windows otherwise, where: * output_spatial_dimensions[spatial_dim] = result_dim. * lhs_dim = input_spatial_dimensions[spatial_dim]. * rhs_dim = kernel_spatial_dimensions[spatial_dim]. * dilated_input_shape[lhs_dim] = dim(lhs, lhs_dim) = 0 ? 0 : (dim(lhs, lhs_dim) - 1) * lhs_dilation[spatial_dim] + 1. * padded_input_shape[lhs_dim] = padding[spatial_dim, 0] + dilated_input_shape[lhs_dim] + padding[spatial_dim, 1]. * dilated_window_shape[lhs_dim] = dim(rhs, rhs_dim) = 0 ? 0 : (dim(rhs, rhs_dim) - 1) * rhs_dilation[spatial_dim] + 1. * is_empty_window[lhs_dim] = padded_input_shape[lhs_dim] = 0 || dilated_window_shape[lhs_dim] > padded_input_shape[lhs_dim]. * num_windows = is_empty_window[lhs_dim] ? 0 : floor((padded_input_shape[lhs_dim] - dilated_window_shape[lhs_dim]) / window_strides[spatial_dim]) + 1. ``` Because of (C14), input_feature_dimension and kernel_input_feature_dimension must be static. input_batch_dimension must be static if batch_group_count > 1 (C10) or if output_batch_dimension is static (C25, first bullet). kernel_output_feature_dimension must be static if batch_group_count > 1 (C15) or feature_group_count > 1 (C16) or if output_feature_dimension is static (C25, second bullet). Because of (C25), each spatial dimension in the output can depend on the spatial dimensions in the inputs (input + kernel), so if it is static in the output, it must be static in the inputs, otherwise mismatches could occur at runtime.
|
Fixed the conflicts, this is ready for review. |
ghpvnist
requested changes
Apr 17, 2024
ghpvnist
approved these changes
Apr 17, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Relevant constraints from the spec:
Because of (C14), input_feature_dimension and kernel_input_feature_dimension must be static. input_batch_dimension must be static if batch_group_count > 1 (C10) or if output_batch_dimension is static (C25, first bullet). kernel_output_feature_dimension must be static if batch_group_count > 1 (C15) or feature_group_count > 1 (C16) or if output_feature_dimension is static (C25, second bullet).
Because of (C25), each spatial dimension in the output can depend on the spatial dimensions in the inputs (input + kernel), so if it is static in the output, it must be static in the inputs, otherwise mismatches could occur at runtime.