@ -6,7 +6,7 @@ keywords: ImageNet10, ImageNet, Ultralytics, CI tests, sanity checks, training p
# ImageNet10 Dataset
The [ImageNet10](https://github.com/ultralytics/yolov5/releases/download/v1.0/imagenet10.zip) dataset is a small-scale subset of the [ImageNet](https://www.image-net.org/) database, developed by [Ultralytics](https://ultralytics.com) and designed for CI tests, sanity checks, and fast testing of training pipelines. This dataset is composed of the first image in the training set and the first image from the validation set of the first 10 classes in ImageNet. Although significantly smaller, it retains the structure and diversity of the original ImageNet dataset.
The [ImageNet10](https://github.com/ultralytics/assets/releases/download/v0.0.0/imagenet10.zip) dataset is a small-scale subset of the [ImageNet](https://www.image-net.org/) database, developed by [Ultralytics](https://ultralytics.com) and designed for CI tests, sanity checks, and fast testing of training pipelines. This dataset is composed of the first image in the training set and the first image from the validation set of the first 10 classes in ImageNet. Although significantly smaller, it retains the structure and diversity of the original ImageNet dataset.
## Key Features
@ -80,7 +80,7 @@ We would like to acknowledge the ImageNet team, led by Olga Russakovsky, Jia Den
### What is the ImageNet10 dataset and how is it different from the full ImageNet dataset?
The [ImageNet10](https://github.com/ultralytics/yolov5/releases/download/v1.0/imagenet10.zip) dataset is a compact subset of the original [ImageNet](https://www.image-net.org/) database, created by Ultralytics for rapid CI tests, sanity checks, and training pipeline evaluations. ImageNet10 comprises only 20 images, representing the first image in the training and validation sets of the first 10 classes in ImageNet. Despite its small size, it maintains the structure and diversity of the full dataset, making it ideal for quick testing but not for benchmarking models.
The [ImageNet10](https://github.com/ultralytics/assets/releases/download/v0.0.0/imagenet10.zip) dataset is a compact subset of the original [ImageNet](https://www.image-net.org/) database, created by Ultralytics for rapid CI tests, sanity checks, and training pipeline evaluations. ImageNet10 comprises only 20 images, representing the first image in the training and validation sets of the first 10 classes in ImageNet. Despite its small size, it maintains the structure and diversity of the full dataset, making it ideal for quick testing but not for benchmarking models.
### How can I use the ImageNet10 dataset to test my deep learning model?
@ -124,4 +124,4 @@ The ImageNet10 dataset has several key features:
### Where can I download the ImageNet10 dataset?
You can download the ImageNet10 dataset from the [Ultralytics GitHub releases page](https://github.com/ultralytics/yolov5/releases/download/v1.0/imagenet10.zip). For more detailed information about its structure and applications, refer to the [ImageNet10 Dataset](imagenet10.md) page.
You can download the ImageNet10 dataset from the [Ultralytics GitHub releases page](https://github.com/ultralytics/assets/releases/download/v0.0.0/imagenet10.zip). For more detailed information about its structure and applications, refer to the [ImageNet10 Dataset](imagenet10.md) page.
[TensorRT](https://developer.nvidia.com/tensorrt), developed by NVIDIA, is an advanced software development kit (SDK) designed for high-speed deep learning inference. It's well-suited for real-time applications like object detection.
Select a pretrained model to start training from. Here we select [YOLOv5s](https://github.com/ultralytics/yolov5/blob/master/models/yolov5s.yaml), the smallest and fastest model available. See our README [table](https://github.com/ultralytics/yolov5#pretrained-checkpoints) for a full comparison of all models. We will train this model with Multi-GPU on the [COCO](https://github.com/ultralytics/yolov5/blob/master/data/scripts/get_coco.sh) dataset.
@ -28,7 +28,7 @@ We've put together a full guide for users looking to get the best results on the
Larger models like YOLOv5x and [YOLOv5x6](https://github.com/ultralytics/yolov5/releases/tag/v5.0) will produce better results in nearly all cases, but have more parameters, require more CUDA memory to train, and are slower to run. For **mobile** deployments we recommend YOLOv5s/m, for **cloud** deployments we recommend YOLOv5l/x. See our README [table](https://github.com/ultralytics/yolov5#pretrained-checkpoints) for a full comparison of all models.
- **Start from Pretrained weights.** Recommended for small to medium-sized datasets (i.e. [VOC](https://github.com/ultralytics/yolov5/blob/master/data/VOC.yaml), [VisDrone](https://github.com/ultralytics/yolov5/blob/master/data/VisDrone.yaml), [GlobalWheat](https://github.com/ultralytics/yolov5/blob/master/data/GlobalWheat2020.yaml)). Pass the name of the model to the `--weights` argument. Models download automatically from the [latest YOLOv5 release](https://github.com/ultralytics/yolov5/releases).
@ -127,7 +127,7 @@ Organize your train and val images and labels according to the example below. YO
Select a pretrained model to start training from. Here we select [YOLOv5s](https://github.com/ultralytics/yolov5/blob/master/models/yolov5s.yaml), the second-smallest and fastest model available. See our README [table](https://github.com/ultralytics/yolov5#pretrained-checkpoints) for a full comparison of all models.
@ -185,7 +185,7 @@ Training results are automatically logged with [Tensorboard](https://www.tensorf
This directory contains train and val statistics, mosaics, labels, predictions and augmented mosaics, as well as metrics and charts including precision-recall (PR) curves and confusion matrices.
Results file `results.csv` is updated after each epoch, and then plotted as `results.png` (below) after training completes. You can also plot any `results.csv` file manually:
@ -195,7 +195,7 @@ from utils.plots import plot_results
plot_results("path/to/results.csv") # plot 'results.csv' as 'results.png'
Glenn Jocher
5 months ago
committed by
GitHub