vllm-docker 0.6.3

Breaking change

This image no longer provides a GPU image since the upstream image is better. This repo going forward will only publish cpu, tpu and other images that are not being published by vLLM.

What's Changed

• vLLM 0.6.2 by @nstogner in #29
• update to v0.6.2 by @samos123 in #31
• update to v0.6.3 by @samos123 in #32
• enable caching of gha by @samos123 in #33
• reduce pip install size by @samos123 in #34
• Remove GPU image and point to upstream image instead by @samos123 in #35

New Contributors

• @nstogner made their first contribution in #29

Full Changelog: v0.6.2...v0.6.3