Skip to content

Latest commit

 

History

History
8 lines (5 loc) · 271 Bytes

mlsd-preprocessing.md

File metadata and controls

8 lines (5 loc) · 271 Bytes

Preprocessing Text:

  • Normalization -> Tokenization [Pre-Tokenization -> Tokenizer Model -> Post-processing] -> Token to ids (lookup table, hashing)

Preprocessing Images:

Preprocessing Videos:

Decode frames -> sample frames -> Resize -> Scale, normalize