search

THU-MIG / RepViT

RepViT: Revisiting Mobile CNN From ViT Perspective [CVPR 2024] and RepViT-SAM: Towards Real-Time Segmenting Anything
https://arxiv.org/abs/2307.09283
Apache License 2.0
681 stars 55 forks source link

您好，请问如何将蒸馏模式改成直接训练模式？ #37

Closed Castrol68 closed 4 months ago

Castrol68 commented 4 months ago

您好，使用readme里面的训练命令会默认使用蒸馏模式（加载了教师模型），请问如何修改才能使用直接训练模式（不使用教师模型）？

jameslahm commented 4 months ago

Thanks for your interest. You could pass '--distillation-type none' as the command line arguments to disable the distillation, like below

python -m torch.distributed.launch --nproc_per_node=8 --master_port 12346 --use_env main.py --model repvit_m0_9 --data-path ~/imagenet --dist-eval --distillation-type none