google / jetstream-pytorch

PyTorch/XLA integration with JetStream (https://github.com/google/JetStream) for LLM inference"
Apache License 2.0
32 stars 14 forks source link

Add gemma and update recent changes to multiple host #74

Closed FanhaiLu1 closed 3 months ago

FanhaiLu1 commented 3 months ago

This PR add gemma and update recent changes to multiple host. In the end, the multiple host and single host should share as much code as possible after we finished multiple host support, but right now the code are diverged.