[Bugfix] Nms_ir data_race solved #2600
Merged
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
vinx13
reviewed
Feb 15, 2019
| @@ -151,8 +151,7 @@ def calculate_overlap(out_tensor, box_a_idx, box_b_idx): | |||
| with ib.if_scope(tvm.all(nms_topk_node > 0, nms_topk < p_valid_count[b])): | |||
| with ib.for_range(0, p_valid_count[b] - nkeep) as l: | |||
| with ib.if_scope(i < 6): | |||
| p_out[(base_idx + (l + nkeep) * 6 + i)] = \ | |||
| p_data[(base_idx + (l + nkeep) * 6 + i)] | |||
| p_out[(base_idx + (l + nkeep) * 6 + i)] = -1.0 | |||
There was a problem hiding this comment.
What's the meaning of -1? Are the last (valid_count - nkeep) bboxes dropped by this line?
There was a problem hiding this comment.
Yes, it's based on the implementation of the GluonCV SSD, plz reference PR #2353 .
vinx13
reviewed
Feb 16, 2019
| @@ -169,6 +168,9 @@ def calculate_overlap(out_tensor, box_a_idx, box_b_idx): | |||
| base_idx + offset_i + 2) | |||
| with ib.if_scope(iou >= nms_threshold): | |||
| p_out[base_idx + offset_i] = -1.0 | |||
| ib.emit(tvm.make.Call(None, 'tvm_storage_sync', | |||
There was a problem hiding this comment.
this makes sense to me, could you also add this to proposal op?
There was a problem hiding this comment.
Sure, I’ll do that.
vinx13
approved these changes
Feb 18, 2019
masahi
reviewed
Feb 18, 2019
masahi
approved these changes
Feb 18, 2019
wweic
pushed a commit
to neo-ai/tvm
that referenced
this pull request
Feb 20, 2019
* nms data race solved * tst_topi_vision reference results are gonna be updated in PR apache#2353 * proposal nms_ir updated
wweic
pushed a commit
to neo-ai/tvm
that referenced
this pull request
Feb 20, 2019
* nms data race solved * tst_topi_vision reference results are gonna be updated in PR apache#2353 * proposal nms_ir updated
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.
Bug reproduced on Mali GPU. Previously was tested on Intel graphics, Nvidia GPUs, worked find with no erros. @vinx13 Please lemme know if you think this also applies to faster-r-cnn, thanks!
@vinx13 @masahi Please help review