| LibreHRNetw48-pose weights |
| ========================== |
|
|
| These weights derive from the official HRNet pose model zoo in |
| https://github.com/leoxiaobin/deep-high-resolution-net.pytorch at commit |
| 6f69e4676ad8d43d0d61b64b1b9726f0c369e7b1. |
|
|
| The upstream repository is MIT-licensed, Copyright (c) 2019 Leo Xiao. Its |
| adapted implementation files also carry: Copyright (c) Microsoft. Written by |
| Bin Xiao. The official model zoo attaches no separate per-file weight license, |
| so this mirror relies on the MIT license implied by the releasing project and |
| does not describe that basis as a checkpoint-specific grant. |
|
|
| Official source: pose_hrnet_w48_384x288.pth |
| Source SHA-256: 95e0fec3194826d5e3f806ea89be68bbb84517b114c3a32b3058c56610b5ef61 |
|
|
| LibreYOLO wraps the original state dict with model-family, size, task, class, |
| input-shape, and COCO-17 keypoint metadata. Learned tensor keys, values, and |
| dtypes are unchanged. |
|
|
| Converted file: LibreHRNetw48-pose.pt |
| Converted SHA-256: ab85504f323c43babdcbed6550f759d323efd0ffc284a0593043e108cd3c4df5 |
|
|
| See LICENSE for the verbatim upstream MIT license. |
|
|
