Skip to content

Latest commit

 

History

History
22 lines (17 loc) · 806 Bytes

README.md

File metadata and controls

22 lines (17 loc) · 806 Bytes

Learn-from-datasets

Learn deeplearning from public datasets & state-of-the-art papers or starter datasets by different models.

Who need this project?

This project is for deeplearning learners who find no way to advance their skills.

We will fit every project follow the list below:

  1. Write a data reader with same API.
  2. Build a model with common model architecture and fine-turn it, save the fine-turn history.
  3. Try to modify the common model architecture and fine-turn it.
  4. Recurrent the model in the paper and analysis what improve the score.
  5. Summary.

Starter datasets

  1. Natural Language Processing with Disaster Tweets

State-of-the-art results

  1. sentiment_analysis