Add A100 PyTorch/CUDA RT-DETR benchmark submissions (5 variants)#3
Merged
Conversation
5 new submissions for RT-DETR ResNet variants on NVIDIA A100-PCIE-40GB (MIG 7g.40gb slice) with torch 2.6.0+cu124, libreyolo 1.0.0 at the existing pinned commit 1c70efb05a78d1a6e82f29478283883fc9bf38f9. Variants: r18, r34, r50, r50m, r101. Metadata changes: - support-matrix.json: append the 5 rtdetr-r* model ids. - website/src/data/metadata/models.json: add rtdetr-r34 and rtdetr-r50m entries. The other three (rtdetr-r18, rtdetr-r50, rtdetr-r101) were already present. The r50m variant is LibreYOLO-specific (half-width hybrid encoder + early decoder exit at layer 3 of 6), not in the original RT-DETR paper. Local: scripts/validate_submission.py and scripts/build_verified_results.py both exit 0. generated/verified-results.v1.json regenerated and committed. Measured COCO val2017 mAP@50-95 (paper reference in parens): rtdetr-r18 0.4556 (~0.465) rtdetr-r34 0.4822 (~0.485) rtdetr-r50 0.5274 (~0.531) rtdetr-r50m 0.5079 (LibreYOLO variant, no paper reference) rtdetr-r101 0.5392 (~0.543)
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
5 new submissions for RT-DETR ResNet variants on NVIDIA A100-PCIE-40GB (MIG 7g.40gb slice) with torch 2.6.0+cu124, libreyolo 1.0.0 at the existing pinned commit 1c70efb05a78d1a6e82f29478283883fc9bf38f9.
Variants: r18, r34, r50, r50m, r101.
Metadata changes:
The r50m variant is LibreYOLO-specific (half-width hybrid encoder + early decoder exit at layer 3 of 6), not in the original RT-DETR paper.
Local: scripts/validate_submission.py and scripts/build_verified_results.py both exit 0. generated/verified-results.v1.json regenerated and committed.
Measured COCO val2017 mAP@50-95 (paper reference in parens):
rtdetr-r18 0.4556 (~0.465)
rtdetr-r34 0.4822 (~0.485)
rtdetr-r50m 0.5079 (LibreYOLO variant, no paper reference)
rtdetr-r50 0.5274 (~0.531)
rtdetr-r101 0.5392 (~0.543)