dlk 0.0.1

dlk: Deep Learning Kit

A Deep Learning Kit

This project is WIP, only provide some basic method, and is not been tested.

What is this project?

• Provide a templete for deep learning (especially for nlp) training and deploy.
• Provide parameters search.
• Provide basic architecture search.
• Provide some basic modules and models.
• Provide basic deploy method.

Config example

TODO

• Add more documents
• distill struct
• Add disable tokenizer post process
• [o] Predict
  • make the main code.
  • test

• when the high config change the _name of the basic config, the base config should be coverd

• one Optimizer different Para Group use different Schedulers.

  • Ref https://pytorch.org/docs/stable/generated/torch.optim.lr_scheduler.CyclicLR.html#torch.optim.lr_scheduler.CyclicLR
  • add loss schedule,
    • how to get best checkpoint by loss

• main entry?

• LightGBM

• It's necessary？