v0.15.3
xorbitsai/inferencev0.15.3Sep 30, 2024by XprobeBot
AI Summary
This update adds support for Jina-embedding-v3, DeepCache with Stable Diffusion models, and Minicpm-reranker, alongside fixes for UI bugs and vLLM defaults.
Key Highlights
- Added support for Jina-embedding-v3.
- Added DeepCache support for Stable Diffusion models.
- Added support for Minicpm-reranker model.
New Features
- Support jina-embedding-v3
- Support deepcache with sd models
- support minicpm-reranker model
- add vllm restart check and support internvl multi-image chat
Full Release Notes
# What's new in 0.15.3 (2024-09-30) These are the changes in inference v0.15.3. ## New features * Feat: Support jina-embedding-v3 by @amumu96 in https://github.com/xorbitsai/inference/pull/2379 * FEAT: Support deepcache with sd models by @frostyplanet in https://github.com/xorbitsai/inference/pull/2313 * FEAT: support minicpm-reranker model by @hwzhuhao in https://github.com/xorbitsai/inference/pull/2383 * FEAT: add vllm restart check and support internvl multi-image chat by @amumu96 in https://github.com/xorbitsai/inference/pull/2384 ## Bug fixes * BUG: [UI] Fix 'Model Format' bug on model registration page. by @yiboyasss in https://github.com/xorbitsai/inference/pull/2353 * BUG: Fix default value of max_model_len for vLLM backend. by @zjuyzj in https://github.com/xorbitsai/inference/pull/2385 ## New Contributors * @zjuyzj made their first contribution in https://github.com/xorbitsai/inference/pull/2385 * @hwzhuhao made their first contribution in https://github.com/xorbitsai/inference/pull/2383 **Full Changelog**: https://github.com/xorbitsai/inference/compare/v0.15.2...v0.15.3