Login

Kye Gomez/zeta

Build high-performance AI models with modular building blocks

Issues looking for funding

[BUG] RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument weight in method wrapper_CUDA__native_layer_norm)
bug

#197 opened in kyegomez/zeta

4
Fund
Should I use linear layers for the input and output of FlashAttention?
bug

#247 opened in kyegomez/zeta

2
Fund
[BUG] Why is the backpropagation calculation so slow when I use the mamba network?
bug

#220 opened in kyegomez/zeta

5
Fund
kyegomez/zeta
Creatorkyegomez
Stars378
LicenseApache License 2.0
RepositoryGitHub
Websitezeta.apac.ai