What's Changed
Bug fixes 🐛
- fix(vllm): images and videos are base64 by default by @mudler in #3867
- fix(dependencies): pin pytorch version by @mudler in #3872
- fix(dependencies): move deps that brings pytorch by @mudler in #3873
- fix(vllm): do not set videos if we don't have any by @mudler in #3885
Exciting New Features 🎉
- feat(templates): extract text from multimodal requests by @mudler in #3866
- feat(templates): add sprig to multimodal templates by @mudler in #3868
🧠 Models
- models(gallery): add llama-3_8b_unaligned_beta by @mudler in #3818
- models(gallery): add llama3.1-flammades-70b by @mudler in #3819
- models(gallery): add llama3.1-gutenberg-doppel-70b by @mudler in #3820
- models(gallery): add llama-3.1-8b-arliai-formax-v1.0-iq-arm-imatrix by @mudler in #3821
- models(gallery): add supernova-medius by @mudler in #3822
- models(gallery): add hermes-3-llama-3.1-70b-lorablated by @mudler in #3823
- models(gallery): add hermes-3-llama-3.1-8b-lorablated by @mudler in #3824
- models(gallery): add eva-qwen2.5-14b-v0.1-i1 by @mudler in #3825
- models(gallery): add cursorcore-qw2.5-7b-i1 by @mudler in #3826
- models(gallery): add cursorcore-qw2.5-1.5b-lc-i1 by @mudler in #3827
- models(gallery): add cursorcore-ds-6.7b-i1 by @mudler in #3828
- models(gallery): add cursorcore-yi-9b by @mudler in #3829
- models(gallery): add edgerunner-command-nested-i1 by @mudler in #3830
- models(gallery): add llama-3.2-chibi-3b by @mudler in #3843
- models(gallery): add llama-3.2-3b-reasoning-time by @mudler in #3844
- models(gallery): add ml-ms-etheris-123b by @mudler in #3845
- models(gallery): add doctoraifinetune-3.1-8b-i1 by @mudler in #3846
- models(gallery): add astral-fusion-neural-happy-l3.1-8b by @mudler in #3848
- models(gallery): add tsunami-0.5x-7b-instruct-i1 by @mudler in #3849
- models(gallery): add mahou-1.5-llama3.1-70b-i1 by @mudler in #3850
- models(gallery): add llama-3.1-nemotron-70b-instruct-hf by @mudler in #3854
- models(gallery): add qevacot-7b-v2 by @mudler in #3855
- models(gallery): add l3.1-etherealrainbow-v1.0-rc1-8b by @mudler in #3856
- models(gallery): add phi-3.5-mini-titanfusion-0.2 by @mudler in #3857
- models(gallery): add mn-lulanum-12b-fix-i1 by @mudler in #3859
- models(gallery): add apollo2-9b by @mudler in #3860
- models(gallery): add theia-llama-3.1-8b-v1 by @mudler in #3861
- models(gallery): add tor-8b by @mudler in #3862
- models(gallery): add darkens-8b by @mudler in #3863
- models(gallery): add baldur-8b by @mudler in #3864
- models(gallery): add meissa-qwen2.5-7b-instruct by @mudler in #3865
- models(gallery): add phi-3 vision by @mudler in #3890
👒 Dependencies
- chore(deps): Bump docs/themes/hugo-theme-relearn from
d5a0ee0toe1a1f01by @dependabot in #3798 - chore(deps): Bump mxschmitt/action-tmate from 3.18 to 3.19 by @dependabot in #3799
- chore(deps): Bump sentence-transformers from 3.1.1 to 3.2.0 in /backend/python/sentencetransformers by @dependabot in #3801
- chore(deps): Bump langchain from 0.3.2 to 0.3.3 in /examples/langchain/langchainpy-localai-example by @dependabot in #3803
- chore(deps): Bump llama-index from 0.11.16 to 0.11.17 in /examples/langchain-chroma by @dependabot in #3804
- chore(deps): Bump python from 3.12-bullseye to 3.13-bullseye in /examples/langchain by @dependabot in #3805
- chore(deps): Bump openai from 1.51.1 to 1.51.2 in /examples/functions by @dependabot in #3806
- chore(deps): Bump llama-index from 0.11.16 to 0.11.17 in /examples/chainlit by @dependabot in #3807
- chore(deps): Bump langchain from 0.3.1 to 0.3.3 in /examples/langchain-chroma by @dependabot in #3809
- chore(deps): Bump openai from 1.51.1 to 1.51.2 in /examples/langchain/langchainpy-localai-example by @dependabot in #3808
- chore(deps): Bump yarl from 1.13.1 to 1.15.1 in /examples/langchain/langchainpy-localai-example by @dependabot in #3816
- chore(deps): Bump chromadb from 0.5.11 to 0.5.13 in /examples/langchain-chroma by @dependabot in #3811
- chore(deps): Bump langchain from 0.3.2 to 0.3.3 in /examples/functions by @dependabot in #3802
- chore(deps): Bump debugpy from 1.8.6 to 1.8.7 in /examples/langchain/langchainpy-localai-example by @dependabot in #3814
- chore(deps): Bump aiohttp from 3.10.9 to 3.10.10 in /examples/langchain/langchainpy-localai-example by @dependabot in #3812
- chore(deps): Bump openai from 1.51.1 to 1.51.2 in /examples/langchain-chroma by @dependabot in #3810
- chore(deps): Bump charset-normalizer from 3.3.2 to 3.4.0 in /examples/langchain/langchainpy-localai-example by @dependabot in #3834
- chore(deps): Bump langchain-community from 0.3.1 to 0.3.2 in /examples/langchain/langchainpy-localai-example by @dependabot in #3831
- chore(deps): Bump yarl from 1.15.1 to 1.15.2 in /examples/langchain/langchainpy-localai-example by @dependabot in #3832
- chore(deps): Bump numpy from 2.1.1 to 2.1.2 in /examples/langchain/langchainpy-localai-example by @dependabot in #3833
- chore(deps): Bump docs/themes/hugo-theme-relearn from
e1a1f01to007cc20by @dependabot in #3835 - chore(deps): Bump gradio from 3.48.0 to 5.0.0 in /backend/python/openvoice in the pip group by @dependabot in #3880
- chore(deps): bump llama-cpp to cda0e4b648dde8fac162b3430b14a99597d3d74f by @mudler in #3884
Other Changes
- docs: ⬆️ update docs version mudler/LocalAI by @localai-bot in #3796
- chore: dependabot ignore generated grpc go package by @dave-gray101 in #3795
- chore: ⬆️ Update ggerganov/llama.cpp to
edc265661cd707327297b6ec4d83423c43cb50a5by @localai-bot in #3797 - chore: ⬆️ Update ggerganov/llama.cpp to
d4c19c0f5cdb1e512573e8c86c79e8d0238c73c4by @localai-bot in #3817 - chore: ⬆️ Update ggerganov/llama.cpp to
a89f75e1b7b90cb2d4d4c52ca53ef9e9b466aa45by @localai-bot in #3837 - chore: ⬆️ Update ggerganov/whisper.cpp to
06a1da9daff94c1bf1b1d38950628264fe443f76by @localai-bot in #3836 - Update integrations.md with LLPhant by @f-lombardo in #3838
- fix(llama.cpp): consider also native builds by @mudler in #3839
- chore: ⬆️ Update ggerganov/whisper.cpp to
b6049060dd2341b7816d2bce7dc7451c1665828eby @localai-bot in #3842 - chore: ⬆️ Update ggerganov/llama.cpp to
755a9b2bf00fbae988e03a47e852b66eaddd113aby @localai-bot in #3841 - chore(deps): bump grpcio to 1.67.0 by @mudler in #3851
- chore: ⬆️ Update ggerganov/llama.cpp to
9e041024481f6b249ab8918e18b9477f873b5a5eby @localai-bot in #3853 - chore: ⬆️ Update ggerganov/whisper.cpp to
d3f7137cc9befa6d74dc4085de2b664b97b7c8bbby @localai-bot in #3852 - fix(mamba): pin torch version by @mudler in #3871
- chore: ⬆️ Update ggerganov/llama.cpp to
99bd4ac28c32cd17c0e337ff5601393b033dc5fcby @localai-bot in #3869 - chore: ⬆️ Update ggerganov/whisper.cpp to
a5abfe6a90495f7bf19fe70d016ecc255e97359cby @localai-bot in #3870 - chore(deps): pin packaging by @mudler in #3875
- chore(deps): pin numpy by @mudler in #3876
- chore(deps): pin networkx by @mudler in #3878
- chore: ⬆️ Update ggerganov/llama.cpp to
afd9909a6481402844aecefa8a8908afdd7f52f1by @localai-bot in #3879 - Revert "chore(deps): Bump gradio from 3.48.0 to 5.0.0 in /backend/python/openvoice in the pip group" by @mudler in #3881
- chore: ⬆️ Update ggerganov/llama.cpp to
45f097645efb11b6d09a5b4adbbfd7c312ac0126by @localai-bot in #3889
New Contributors
- @f-lombardo made their first contribution in #3838
Full Changelog: v2.22.0...v2.22.1