version 0.2 iteration plan

[mydmdm]

Estimated Release Date: 2/5 Release Manager: @suiguoxin Schedule:

• Design Review: 1/19
• Coding: 1/26
• Testing: 2/2

Features

• [x] P0 Feature Planning @iofu728 @lunaqiu ETA: 1.16
• [ ] P0 Interface Definition: Engine < Core < Wrapper < Applications #52 @SiyunZhao ETA: 1.16
• [ ] P0 Layered refactor @SiyunZhao @iofu728
• [ ] fixed/customized/accurate/target/max compression ratio
  • [ ] P1 exp: target comp ratio v.s. real comp ratio on specific data
  • [ ] P0 doc @SiyunZhao
• [ ] P1 Support customized compression spec, such as user specified segment boundary and compression ratio
  • [ ] Support preserving essential characters
  • [ ] list different mappings and design interface
• [ ] bug fix TBD After Interface Refactoring
  • [ ] P1 word level compression #4
  • [ ] P1 #50
• [ ] P1 Support more / faster engines #41, including llama_cpp, FasterTransformer, vLLM ETA: TBD
  • [ ] survey which engines to support
• [x] P1 Support more models, small LMs e.g., Phi2 ETA: 2 days #67
• [ ] P2 Documentation and examples
• [x] PR (Ch, 1000 words) @lunaqiu @iofu728 1.17

Backlog