Genius Patrick e5f8711dfe | 1 year ago | |
---|---|---|
.. | ||
README.md | 1 year ago | |
convit_base_ascend.yaml | 1 year ago | |
convit_base_plus_ascend.yaml | 1 year ago | |
convit_small_ascend.yaml | 1 year ago | |
convit_small_plus_ascend.yaml | 1 year ago | |
convit_tiny_ascend.yaml | 1 year ago | |
convit_tiny_gpu.yaml | 1 year ago | |
convit_tiny_plus_ascend.yaml | 1 year ago |
ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases
ConViT combines the strengths of convolutional architectures and Vision Transformers (ViTs).
ConViT introduces gated positional self-attention (GPSA), a form of positional self-attention
that can be equipped with a “soft” convolutional inductive bias.
ConViT initializes the GPSA layers to mimic the locality of convolutional layers,
then gives each attention head the freedom to escape locality by adjusting a gating parameter
regulating the attention paid to position versus content information.
ConViT, outperforms the DeiT (Touvron et al., 2020) on ImageNet,
while offering a much improved sample efficiency.[1]
Figure 1. Architecture of ConViT [1]
Our reproduced model performance on ImageNet-1K is reported as follows.
Model | Context | Top-1 (%) | Top-5 (%) | Params (M) | Recipe | Download |
---|---|---|---|---|---|---|
convit_tiny | D910x8-G | 73.66 | 91.72 | 5.71 | yaml | weights |
convit_tiny_plus | D910x8-G | 77.00 | 93.60 | 9.97 | yaml | weights |
convit_small | D910x8-G | 81.63 | 95.59 | 27.78 | yaml | weights |
convit_small_plus | D910x8-G | 81.80 | 95.42 | 48.98 | yaml | weights |
convit_base | D910x8-G | 82.10 | 95.52 | 86.54 | yaml | weights |
convit_base_plus | D910x8-G | 81.96 | 95.04 | 153.13 | yaml | weights |
Please refer to the installation instruction in MindCV.
Please download the ImageNet-1K dataset for model training and validation.
It is easy to reproduce the reported results with the pre-defined training recipe. For distributed training on multiple Ascend 910 devices, please run
# distributed training on multiple GPU/Ascend devices
mpirun -n 8 python train.py --config configs/convit/convit_tiny_ascend.yaml --data_dir /path/to/imagenet
If the script is executed by the root user, the
--allow-run-as-root
parameter must be added tompirun
.
Similarly, you can train the model on multiple GPU devices with the above mpirun
command.
For detailed illustration of all hyper-parameters, please refer to config.py.
Note: As the global batch size (batch_size x num_devices) is an important hyper-parameter, it is recommended to keep the global batch size unchanged for reproduction or adjust the learning rate linearly to a new global batch size.
If you want to train or finetune the model on a smaller dataset without distributed training, please run:
# standalone training on a CPU/GPU/Ascend device
python train.py --config configs/convit/convit_tiny_ascend.yaml --data_dir /path/to/dataset --distribute False
To validate the accuracy of the trained model, you can use validate.py
and parse the checkpoint path with --ckpt_path
.
python validate.py -c configs/convit/convit_tiny_ascend.yaml --data_dir /path/to/imagenet --ckpt_path /path/to/ckpt
Please refer to the deployment tutorial in MindCV.
[1] d’Ascoli S, Touvron H, Leavitt M L, et al. Convit: Improving vision transformers with soft convolutional inductive biases[C]//International Conference on Machine Learning. PMLR, 2021: 2286-2296.
No Description
Jupyter Notebook Python Markdown Text
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》