Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Draft ONLY] nn.Graph asymmetric input/output debug runnable #6236

Closed
wants to merge 18 commits into from

Conversation

chengtbf
Copy link
Contributor

@chengtbf chengtbf commented Sep 10, 2021

if (node->GetTaskType() == TaskType::kSrcSubsetTick) {
CHECK(rank_id2src_tick.emplace(node->machine_id(), node).second);
} else if (node->GetTaskType() == TaskType::kDstSubsetTick) {
CHECK(rank_id2dst_tick.emplace(node->machine_id(), node).second);
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

修复 BUG 1 : dst tick 也需要保证 output 和 callback notify 的顺序。

const auto& op_name = nn_graph->outputs_op_names().at(i);
auto* mut_blob = eager_blob_object->mut_blob();
if (IsEmptyComponentEagerBlobObject(eager_blob_object)) {
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

修复 BUG 2: 跳过没有本 rank 分量的 push / pull cb 构建 和 send

bool IsEmptyComponentEagerBlobObject(
const std::shared_ptr<vm::EagerBlobObject>& eager_blob_object) {
Blob* blob = eager_blob_object->mut_blob();
if (blob && blob->shape().elem_cnt() > 0 && blob->dptr()) {
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

修复判断本 rank 分量是否存在的逻辑

@chengtbf
Copy link
Contributor Author

本 PR 已通过

修复

@chengtbf chengtbf closed this Sep 12, 2021
@chengtbf chengtbf deleted the dev_cc_fix_src_tick branch September 12, 2021 17:25
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

流水并行出错记录: header_size == rhs->blob_desc().ByteSizeOfBlobHeader() 报错
1 participant