Image Classification

We currenent release the code and models for:

• ImageNet-1K pretraining

• Large resolution fine-tuning

Update

• :fire: 03/12/2024: Pretrained models on ImageNet-1K are released.

Model Zoo

See MODEL_ZOO.

Usage

Normal Training

Simply run the training scripts in exp as followed:

bash ./exp/videomamba_tiny/run224.sh

If the training was interrupted abnormally, you can simply rerun the script for auto-resuming. Sometimes the checkpoint may not be saved properly, you should set the resumed model via --reusme ${OUTPUT_DIR}/ckpt/checkpoint.pth.

Training w/ SD

Simply run the training scripts in exp_distill as followed:

bash ./exp_distill/videomamba_middle/run224.sh

For teacher_model, we use a smaller model by default.

Large Resolution Fine-tuning

Simply run the training scripts in exp as followed:

bash ./exp/videomamba_tiny/run448.sh

Please set pretrained model via --finetune.

Evaluation

Simply add --eval in the training scripts.

It will evaluate the last model by default. You can set other models via --resume.

Generate curves

You can generate the training curves as followed:

python3 generate_tensoboard.py

Note that you should install tensorboardX.