alxndrTL / mamba.py

A simple and efficient Mamba implementation in pure PyTorch and MLX.
MIT License
960 stars 86 forks source link

Add Mamba language training script to pytorch examples folder #32

Closed michal1000w closed 5 months ago

michal1000w commented 5 months ago

After checking this repository I've found a lack of example training script for Language processing on a custom dataset. Because of that I've created an initial version of such script for the pytorch version. It allows for training custom Mamba model or fine tuning pretrained one on a custom text dataset. I've added it to the examples folder with comments in the code for better readibility.

alxndrTL commented 5 months ago

Thank you!