AI-Hypercomputer / JetStream

JetStream is a throughput and memory optimized engine for LLM inference on XLA devices, starting with TPUs (and GPUs in future -- PRs welcome).
Apache License 2.0
202 stars 26 forks source link

Align Tokenizer in JetStream #40

Closed JoeZijunZhou closed 5 months ago

JoeZijunZhou commented 5 months ago