`ultralytics 8.0.110` new `profile` and `fraction` train args (#2880)

docker/Dockerfile

@@ -4,6 +4,7 @@
 
 # Start FROM PyTorch image https://hub.docker.com/r/pytorch/pytorch or nvcr.io/nvidia/pytorch:23.03-py3
 FROM pytorch/pytorch:2.0.0-cuda11.7-cudnn8-runtime
+RUN pip install --no-cache nvidia-tensorrt --index-url https://pypi.ngc.nvidia.com
 
 # Downloads to user config dir
 ADD https://ultralytics.com/assets/Arial.ttf https://ultralytics.com/assets/Arial.Unicode.ttf /root/.config/Ultralytics/
@@ -29,7 +30,7 @@ ADD https://github.com/ultralytics/assets/releases/download/v0.0.0/yolov8n.pt /u
 
 # Install pip packages
 RUN python3 -m pip install --upgrade pip wheel
-RUN pip install --no-cache -e . albumentations comet tensorboard thop pycocotools
+RUN pip install --no-cache -e . albumentations comet thop pycocotools onnx onnx-simplifier onnxruntime-gpu
 
 # Set environment variables
 ENV OMP_NUM_THREADS=1

docs/reference/nn/modules/conv.md

@@ -7,6 +7,11 @@ description: Explore convolutional neural network modules & techniques such as L
 :::ultralytics.nn.modules.conv.Conv
 <br><br>
 
+# Conv2
+---
+:::ultralytics.nn.modules.conv.Conv2
+<br><br>
+
 # LightConv
 ---
 :::ultralytics.nn.modules.conv.LightConv

docs/reference/yolo/data/augment.md

@@ -87,6 +87,11 @@ description: Use Ultralytics YOLO Data Augmentation transforms with Base, MixUp,
 :::ultralytics.yolo.data.augment.classify_transforms
 <br><br>
 
+# hsv2colorjitter
+---
+:::ultralytics.yolo.data.augment.hsv2colorjitter
+<br><br>
+
 # classify_albumentations
 ---
 :::ultralytics.yolo.data.augment.classify_albumentations

docs/reference/yolo/utils/benchmarks.md

@@ -2,6 +2,11 @@
 description: Improve your YOLO's performance and measure its speed. Benchmark utility for YOLOv5.
 ---
 
+# ProfileModels
+---
+:::ultralytics.yolo.utils.benchmarks.ProfileModels
+<br><br>
+
 # benchmark
 ---
 :::ultralytics.yolo.utils.benchmarks.benchmark

docs/reference/yolo/utils/torch_utils.md

@@ -57,11 +57,21 @@ description: Optimize your PyTorch models with Ultralytics YOLO's torch_utils fu
 :::ultralytics.yolo.utils.torch_utils.get_num_gradients
 <br><br>
 
+# model_info_for_loggers
+---
+:::ultralytics.yolo.utils.torch_utils.model_info_for_loggers
+<br><br>
+
 # get_flops
 ---
 :::ultralytics.yolo.utils.torch_utils.get_flops
 <br><br>
 
+# get_flops_with_torch_profiler
+---
+:::ultralytics.yolo.utils.torch_utils.get_flops_with_torch_profiler
+<br><br>
+
 # initialize_weights
 ---
 :::ultralytics.yolo.utils.torch_utils.initialize_weights

ultralytics/__init__.py

@@ -1,6 +1,6 @@
 # Ultralytics YOLO 🚀, AGPL-3.0 license
 
-__version__ = '8.0.109'
+__version__ = '8.0.110'
 
 from ultralytics.hub import start
 from ultralytics.vit.rtdetr import RTDETR

ultralytics/yolo/data/augment.py

@@ -778,11 +778,8 @@ def v8_transforms(dataset, imgsz, hyp):
         if flip_idx is None and hyp.fliplr > 0.0:
             hyp.fliplr = 0.0
             LOGGER.warning("WARNING ⚠️ No 'flip_idx' array defined in data.yaml, setting augmentation 'fliplr=0.0'")
-        elif flip_idx:
-            if len(flip_idx) != kpt_shape[0]:
-                raise ValueError(f'data.yaml flip_idx={flip_idx} length must be equal to kpt_shape[0]={kpt_shape[0]}')
-            elif flip_idx[0] != 0:
-                raise ValueError(f'data.yaml flip_idx={flip_idx} must be zero-index (start from 0)')
+        elif flip_idx and (len(flip_idx) != kpt_shape[0]):
+            raise ValueError(f'data.yaml flip_idx={flip_idx} length must be equal to kpt_shape[0]={kpt_shape[0]}')
 
     return Compose([
         pre_transform,