xclele/yolov10
1
0
Fork 0
mirror of https://github.com/THU-MIG/yolov10.git synced 2025-05-23 13:34:23 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

ultralytics 8.1.28 avoid * ops on bool Tensors for RT-DETR OpenVINO export (#8937)

Browse Source 
Co-authored-by: UltralyticsAssistant <web@ultralytics.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
This commit is contained in:
Kayzwer 2024-03-15 08:20:10 +08:00 committed by GitHub
parent bc34ae560b
commit 2d513a9e4b
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 2 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ultralytics/__init__.py
View File

@ -1,6 +1,6 @@
# Ultralytics YOLO 🚀, AGPL-3.0 license # Ultralytics YOLO 🚀, AGPL-3.0 license
__version__ = "8.1.27" __version__ = "8.1.28"
from ultralytics.data.explorer.explorer import Explorer from ultralytics.data.explorer.explorer import Explorer
from ultralytics.models import RTDETR, SAM, YOLO, YOLOWorld from ultralytics.models import RTDETR, SAM, YOLO, YOLOWorld

2
ultralytics/models/rtdetr/model.py
View File

@ -7,8 +7,6 @@ hybrid encoder and IoU-aware query selection for enhanced detection accuracy.
For more information on RT-DETR, visit: https://arxiv.org/pdf/2304.08069.pdf For more information on RT-DETR, visit: https://arxiv.org/pdf/2304.08069.pdf
""" """
from pathlib import Path
from ultralytics.engine.model import Model from ultralytics.engine.model import Model
from ultralytics.nn.tasks import RTDETRDetectionModel from ultralytics.nn.tasks import RTDETRDetectionModel

2
ultralytics/nn/modules/head.py
View File

@ -395,7 +395,7 @@ class RTDETRDecoder(nn.Module):
anchors.append(torch.cat([grid_xy, wh], -1).view(-1, h * w, 4)) # (1, h*w, 4) anchors.append(torch.cat([grid_xy, wh], -1).view(-1, h * w, 4)) # (1, h*w, 4)
anchors = torch.cat(anchors, 1) # (1, h*w*nl, 4) anchors = torch.cat(anchors, 1) # (1, h*w*nl, 4)
valid_mask = ((anchors > eps) * (anchors < 1 - eps)).all(-1, keepdim=True) # 1, h*w*nl, 1 valid_mask = ((anchors > eps) & (anchors < 1 - eps)).all(-1, keepdim=True) # 1, h*w*nl, 1
anchors = torch.log(anchors / (1 - anchors)) anchors = torch.log(anchors / (1 - anchors))
anchors = anchors.masked_fill(~valid_mask, float("inf")) anchors = anchors.masked_fill(~valid_mask, float("inf"))
return anchors, valid_mask return anchors, valid_mask