dotnet / machinelearning

ML.NET is an open source and cross-platform machine learning framework for .NET.
https://dot.net/ml
MIT License
8.92k stars 1.86k forks source link

[Tokenizers] Port CodeGenTokenizer & byte-level BPE algorithm #6992

Closed ericstj closed 3 weeks ago

ericstj commented 4 months ago

Port Codegen Tokenizer to enable Phi-2 models

Reference: https://huggingface.co/docs/transformers/main/en/model_doc/codegen https://github.com/huggingface/transformers/blob/0549000c5bf6c7249f411917f2a6f0b6d0f06da1/src/transformers/models/codegen/tokenization_codegen.py#L98

Paper: https://arxiv.org/abs/2203.13474 https://arxiv.org/pdf/2203.13474.pdf

tarekgh commented 3 weeks ago

This is already done!