CondadosAI/sportcv-third-party

sportcv third-party weights

Unmodified copies of third-party weights that sportcv loads, mirrored so that each one can be pinned by sha256 at a fixed commit of this repository. sportcv checks every file's hash on download and on every load (core/weights.py).

Nothing here was trained or changed by CondadosAI. The terms are the upstream authors' own.

File	Upstream	Licence	sha256
rf-detr-seg-m-ft.pth	Roboflow RF-DETR-Seg Medium, https://storage.googleapis.com/rfdetr/rf-detr-seg-m-ft.pth (rfdetr 1.11.2, MD5 a49af1562c3719227ad43d0ca53b4c7a)	Apache-2.0	3ad325094735f431aee9962a8d204d68eb5bfc393d53e7e836e70998fef5ea58
rf-detr-seg-l-ft.pth	Roboflow RF-DETR-Seg Large, https://storage.googleapis.com/rfdetr/rf-detr-seg-l-ft.pth (MD5 275f7b094909544ed2841c94a677d07e)	Apache-2.0	ca7b7c630ba22496067cc4f034c4e70c8e47fd7adcea04c3a49aa8c1755cbe6b
rf-detr-seg-xl-ft.pth	Roboflow RF-DETR-Seg XLarge, https://storage.googleapis.com/rfdetr/rf-detr-seg-xl-ft.pth (MD5 3693b35d0eea86ebb3e0444f4a611fba)	Apache-2.0	77d67f2b8033fbb25e23f29c22c02a8f677b3939687e9a1a5fdde3765f33f423
rtmpose-l_simcc-body7_pt-body7_420e-384x288-3f5a1437_20230504.onnx	OpenMMLab RTMPose-l, body7, 384x288, COCO-17 keypoints: end2end.onnx from rtmposev1/onnx_sdk/rtmpose-l_simcc-body7_pt-body7_420e-384x288-3f5a1437_20230504.zip in Tau-J/RTMPose at commit cd4d7095f5cfc9cfc4f46289bee91ea4a1e1d9fd (zip sha256 fed40dcfafdbadaebaa9681569f6594220313bd90f29d8ff039256076745f349)	Apache-2.0 (MMPose code); trained on body7 (AI Challenger, COCO, CrowdPose, MPII, sub-JHMDB, Halpe, PoseTrack18), whose data terms apply	62425affc6b9edba8f8a65f55f5e278fec5d905d0c90578e0f8503bdf3954587
osnet_x1_0_imagenet.pth	torchreid OSNet x1.0 ImageNet initialisation (KaiyangZhou/deep-person-reid), Google Drive id 1LaG1EJpHrxdAxKnSCJ_i0u-nbxSAeiFY, the URL in torchreid's osnet.py	MIT (deep-person-reid); trained on ImageNet, whose terms apply	fe2d63f9157c28a4a8d8ca29bec12d5b2988ac0346d712025789ea9174968e79

RF-DETR is © Roboflow, Inc.; RTMPose is © OpenMMLab. RF-DETR and RTMPose are distributed under the Apache License 2.0, included as LICENSE; OSNet's MIT notice: Copyright (c) 2018 Kaiyang Zhou. Mirrored on 6 Oct 2026.