TPU inference for vLLM, with unified JAX and PyTorch support.
Source originale
github.com
Prévenir un proche