ultralytics-ascend/ultralytics/utils
Laughing e65f59003b
Remove half when profiling ONNX models (#16405)
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: UltralyticsAssistant <web@ultralytics.com>
2024-09-21 13:54:48 +02:00
..
callbacks Ruff Docstring formatting (#15793) 2024-08-25 04:27:55 +08:00
__init__.py New PERSISTENT_CACHE (#16373) 2024-09-19 21:06:57 +02:00
autobatch.py Ruff Docstring formatting (#15793) 2024-08-25 04:27:55 +08:00
benchmarks.py Remove half when profiling ONNX models (#16405) 2024-09-21 13:54:48 +02:00
checks.py Disable FP16 val on AMP fail and improve AMP checks (#16306) 2024-09-17 18:44:56 +02:00
dist.py Fix HUB session with DDP training (#13103) 2024-06-23 20:00:34 +02:00
downloads.py Ruff Docstring formatting (#15793) 2024-08-25 04:27:55 +08:00
errors.py ultralytics 8.0.239 Ultralytics Actions and hub-sdk adoption (#7431) 2024-01-10 03:16:08 +01:00
files.py Ruff Docstring formatting (#15793) 2024-08-25 04:27:55 +08:00
instance.py Ruff Docstring formatting (#15793) 2024-08-25 04:27:55 +08:00
loss.py Fix torch.amp.autocast('cuda') warnings (#14633) 2024-07-23 21:58:39 +02:00
metrics.py PyUpgrade 3.8 updates (#15941) 2024-09-01 17:45:38 +02:00
ops.py ultralytics 8.2.85 YOLOv10 max_det arg fix (#15917) 2024-09-01 02:04:22 +02:00
patches.py Patch torch.load(..., weights_only=False) to reduce warnings (#14638) 2024-07-23 20:35:58 +02:00
plotting.py Dynamic pose line thickness (#16362) 2024-09-19 16:49:58 +02:00
tal.py Ruff Docstring formatting (#15793) 2024-08-25 04:27:55 +08:00
torch_utils.py Remove extra get_cpu_info return (#16382) 2024-09-20 11:27:20 +02:00
triton.py Add HUB-SDK Docs reference section (#7781) 2024-01-27 00:21:31 +01:00
tuner.py ultralytics 8.2.87 Ray shutdown() workers after tuning (#15976) 2024-09-03 12:43:48 +02:00