Hello, why is paddle training better than mmclass framework when the same dataset has been used? My accuracy on paddle is 0.9871, but on mmclass it is 22.4299
Use the command ‘python tools/train.py configs/mobilenet_v3/mobilenet_v3_small_imagenet.py ’
Data set format
Branch
main branch (mmpretrain version)
Describe the bug
Hello, why is paddle training better than mmclass framework when the same dataset has been used? My accuracy on paddle is 0.9871, but on mmclass it is 22.4299 Use the command ‘python tools/train.py configs/mobilenet_v3/mobilenet_v3_small_imagenet.py ’ Data set format![Snipaste_2023-06-21_10-31-29](https://github.com/open-mmlab/mmpretrain/assets/62111301/512416d9-83fa-499f-906a-f92c18f9e36d)
Environment
{'sys.platform': 'win32', 'Python': '3.8.16 (default, Mar 2 2023, 03:18:16) [MSC v.1916 64 bit ' '(AMD64)]', 'CUDA available': True, 'numpy_random_seed': 2147483648, 'GPU 0': 'NVIDIA GeForce GTX 1050 Ti', 'CUDA_HOME': 'C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6', 'NVCC': 'Cuda compilation tools, release 11.6, V11.6.55', 'MSVC': '用于 x64 的 Microsoft (R) C/C++ 优化编译器 19.35.32216.1 版', 'GCC': 'n/a', 'PyTorch': '1.10.1', 'TorchVision': '0.11.2', 'OpenCV': '4.7.0', 'MMEngine': '0.7.4', 'MMCV': '2.0.0', 'MMPreTrain': '1.0.0rc8+09a3fe7'}
Other information
No response