Tiny-Net PyTorch Model Creation

[rsomani95]

The TinyNet file says that timm==0.1.2 is required. However, this version does not have the efficientnet modules that are imported in the model file.

Additionally, there's also a missing import for the Swish activation. What's the correct version of timm to use this model?

Thanks!

[rsomani95]

Looks like timm==0.4.8 works, with some more additional imports from timm.models.efficientnet_blocks

[iamhankai]

Thanks for the correction.

[rsomani95]

Hi @iamhankai, I added a PR to timm making the model compatible with the latest version here: https://github.com/rwightman/pytorch-image-models/pull/1037

[iamhankai]

@rsomani95 Good job! Thanks for the contribution.

[lucasjinreal]

@iamhankai Hi, I want extend tinynet on a object detection task, but the tinynet constructions coulpled with input resolution, how to take a suitable one for detection task like we want train with input at 640?

[iamhankai]

@jinfagang The simple solution is use the resolution ratio in classification task.