-
-
Notifications
You must be signed in to change notification settings - Fork 1.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat: Openvino runtime for transformer backend and streaming support for Openvino and CUDA #1892
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Manage different BitsAndBytes options with the quantization: parameter in yaml
First draft of OpenVINO integration in transformer backend
✅ Deploy Preview for localai canceled.
|
Example of model definition for Openchat
|
richiejp
reviewed
Mar 26, 2024
mudler
previously approved these changes
Mar 26, 2024
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Great add, thanks @fakezeta !
Signed-off-by: Ettore Di Giacinto <[email protected]>
Signed-off-by: Ettore Di Giacinto <[email protected]>
mudler
approved these changes
Mar 26, 2024
mudler
approved these changes
Mar 26, 2024
whisper.cpp also support |
truecharts-admin
added a commit
to truecharts/charts
that referenced
this pull request
Apr 9, 2024
…2.1 by renovate (#20490) This PR contains the following updates: | Package | Update | Change | |---|---|---| | [docker.io/localai/localai](https://togithub.com/mudler/LocalAI) | minor | `v2.11.0-cublas-cuda11-ffmpeg-core` -> `v2.12.1-cublas-cuda11-ffmpeg-core` | | [docker.io/localai/localai](https://togithub.com/mudler/LocalAI) | minor | `v2.11.0-cublas-cuda11-core` -> `v2.12.1-cublas-cuda11-core` | | [docker.io/localai/localai](https://togithub.com/mudler/LocalAI) | minor | `v2.11.0-cublas-cuda12-ffmpeg-core` -> `v2.12.1-cublas-cuda12-ffmpeg-core` | | [docker.io/localai/localai](https://togithub.com/mudler/LocalAI) | minor | `v2.11.0-cublas-cuda12-core` -> `v2.12.1-cublas-cuda12-core` | | [docker.io/localai/localai](https://togithub.com/mudler/LocalAI) | minor | `v2.11.0-ffmpeg-core` -> `v2.12.1-ffmpeg-core` | | [docker.io/localai/localai](https://togithub.com/mudler/LocalAI) | minor | `v2.11.0` -> `v2.12.1` | --- > [!WARNING] > Some dependencies could not be looked up. Check the Dependency Dashboard for more information. --- ### Release Notes <details> <summary>mudler/LocalAI (docker.io/localai/localai)</summary> ### [`v2.12.1`](https://togithub.com/mudler/LocalAI/releases/tag/v2.12.1) [Compare Source](https://togithub.com/mudler/LocalAI/compare/v2.12.0...v2.12.1) I'm happy to announce the v2.12.1 LocalAI release is out! ##### 🌠 Landing page and Swagger Ever wondered what to do after LocalAI is up and running? Integration with a simple web interface has been started, and you can see now a landing page when hitting the LocalAI front page: ![Screenshot from 2024-04-07 14-43-26](https://togithub.com/mudler/LocalAI/assets/2420543/e7aea8de-4385-45ae-b52e-db8154495493) You can also now enjoy Swagger to try out the API calls directly: ![swagger](https://togithub.com/mudler/LocalAI/assets/2420543/6405ab11-2908-45ff-b635-38e4456251d6) ##### 🌈 AIO images changes Now the default model for CPU images is https://huggingface.co/NousResearch/Hermes-2-Pro-Mistral-7B-GGUF - pre-configured for functions and tools API support! If you are an Intel-GPU owner, the Intel profile for AIO images is now available too! ##### 🚀 OpenVINO and transformers enhancements Now there is support for OpenVINO and transformers got token streaming support now thanks to [@​fakezeta](https://togithub.com/fakezeta)! To try OpenVINO, you can use the example available in the documentation: https://localai.io/features/text-generation/#examples ##### 🎈 Lot of small improvements behind the scenes! Thanks for our outstanding community, we have enhanced several areas: - The build time of LocalAI was speed up significantly! thanks to [@​cryptk](https://togithub.com/cryptk) for the efforts in enhancing the build system - [@​thiner](https://togithub.com/thiner) worked hardly to get Vision support for AutoGPTQ - ... and much more! see down below for a full list, be sure to star LocalAI and give it a try! ##### 📣 Spread the word! First off, a massive thank you (again!) to each and every one of you who've chipped in to squash bugs and suggest cool new features for LocalAI. Your help, kind words, and brilliant ideas are truly appreciated - more than words can say! And to those of you who've been heros, giving up your own time to help out fellow users on Discord and in our repo, you're absolutely amazing. We couldn't have asked for a better community. Just so you know, LocalAI doesn't have the luxury of big corporate sponsors behind it. It's all us, folks. So, if you've found value in what we're building together and want to keep the momentum going, consider showing your support. A little shoutout on your favorite social platforms using @​LocalAI_OSS and @​mudler_it or joining our sponsors can make a big difference. Also, if you haven't yet joined our Discord, come on over! Here's the link: https://discord.gg/uJAeKSAGDy Every bit of support, every mention, and every star adds up and helps us keep this ship sailing. Let's keep making LocalAI awesome together! Thanks a ton, and here's to more exciting times ahead with LocalAI! ##### What's Changed ##### Bug fixes 🐛 - fix: downgrade torch by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1902 - fix(aio): correctly detect intel systems by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1931 - fix(swagger): do not specify a host by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1930 - fix(tools): correctly render tools response in templates by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1932 - fix(grammar): respect JSONmode and grammar from user input by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1935 - fix(hermes-2-pro-mistral): add stopword for toolcall by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1939 - fix(functions): respect when selected from string by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1940 - fix: use exec in entrypoint scripts to fix signal handling by [@​cryptk](https://togithub.com/cryptk) in [mudler/LocalAI#1943 - fix(hermes-2-pro-mistral): correct stopwords by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1947 - fix(welcome): stable model list by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1949 - fix(ci): manually tag latest images by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1948 - fix(seed): generate random seed per-request if -1 is set by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1952 - fix regression [#​1971](https://togithub.com/mudler/LocalAI/issues/1971) by [@​fakezeta](https://togithub.com/fakezeta) in [mudler/LocalAI#1972 ##### Exciting New Features 🎉 - feat(aio): add intel profile by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1901 - Enhance autogptq backend to support VL models by [@​thiner](https://togithub.com/thiner) in [mudler/LocalAI#1860 - feat(assistant): Assistant and AssistantFiles api by [@​christ66](https://togithub.com/christ66) in [mudler/LocalAI#1803 - feat: Openvino runtime for transformer backend and streaming support for Openvino and CUDA by [@​fakezeta](https://togithub.com/fakezeta) in [mudler/LocalAI#1892 - feat: Token Stream support for Transformer, fix: missing package for OpenVINO by [@​fakezeta](https://togithub.com/fakezeta) in [mudler/LocalAI#1908 - feat(welcome): add simple welcome page by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1912 - fix(build): better CI logging and correct some build failure modes in Makefile by [@​cryptk](https://togithub.com/cryptk) in [mudler/LocalAI#1899 - feat(webui): add partials, show backends associated to models by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1922 - feat(swagger): Add swagger API doc by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1926 - feat(build): adjust number of parallel make jobs by [@​cryptk](https://togithub.com/cryptk) in [mudler/LocalAI#1915 - feat(swagger): update by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1929 - feat: first pass at improving logging by [@​cryptk](https://togithub.com/cryptk) in [mudler/LocalAI#1956 - fix(llama.cpp): set better defaults for llama.cpp by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1961 ##### 📖 Documentation and examples - docs(aio-usage): update docs to show examples by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1921 ##### 👒 Dependencies - ⬆️ Update docs version mudler/LocalAI by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1903 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1904 - ⬆️ Update M0Rf30/go-tiny-dream by [@​M0Rf30](https://togithub.com/M0Rf30) in [mudler/LocalAI#1911 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1913 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1914 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1923 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1924 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1928 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1933 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1934 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1937 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1941 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1953 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1958 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1959 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1964 ##### Other Changes - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1927 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1960 - fix(hermes-2-pro-mistral): correct dashes in template to suppress newlines by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1966 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1969 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1970 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1973 ##### New Contributors - [@​thiner](https://togithub.com/thiner) made their first contribution in [mudler/LocalAI#1860 **Full Changelog**: mudler/LocalAI@v2.11.0...v2.12.1 ### [`v2.12.0`](https://togithub.com/mudler/LocalAI/releases/tag/v2.12.0) [Compare Source](https://togithub.com/mudler/LocalAI/compare/v2.11.0...v2.12.0) I'm happy to announce the v2.12.0 LocalAI release is out! ##### 🌠 Landing page and Swagger Ever wondered what to do after LocalAI is up and running? Integration with a simple web interface has been started, and you can see now a landing page when hitting the LocalAI front page: ![Screenshot from 2024-04-07 14-43-26](https://togithub.com/mudler/LocalAI/assets/2420543/e7aea8de-4385-45ae-b52e-db8154495493) You can also now enjoy Swagger to try out the API calls directly: ![swagger](https://togithub.com/mudler/LocalAI/assets/2420543/6405ab11-2908-45ff-b635-38e4456251d6) ##### 🌈 AIO images changes Now the default model for CPU images is https://huggingface.co/NousResearch/Hermes-2-Pro-Mistral-7B-GGUF - pre-configured for functions and tools API support! If you are an Intel-GPU owner, the Intel profile for AIO images is now available too! ##### 🚀 OpenVINO and transformers enhancements Now there is support for OpenVINO and transformers got token streaming support now thanks to [@​fakezeta](https://togithub.com/fakezeta)! To try OpenVINO, you can use the example available in the documentation: https://localai.io/features/text-generation/#examples ##### 🎈 Lot of small improvements behind the scenes! Thanks for our outstanding community, we have enhanced several areas: - The build time of LocalAI was speed up significantly! thanks to [@​cryptk](https://togithub.com/cryptk) for the efforts in enhancing the build system - [@​thiner](https://togithub.com/thiner) worked hardly to get Vision support for AutoGPTQ - ... and much more! see down below for a full list, be sure to star LocalAI and give it a try! ##### 📣 Spread the word! First off, a massive thank you (again!) to each and every one of you who've chipped in to squash bugs and suggest cool new features for LocalAI. Your help, kind words, and brilliant ideas are truly appreciated - more than words can say! And to those of you who've been heros, giving up your own time to help out fellow users on Discord and in our repo, you're absolutely amazing. We couldn't have asked for a better community. Just so you know, LocalAI doesn't have the luxury of big corporate sponsors behind it. It's all us, folks. So, if you've found value in what we're building together and want to keep the momentum going, consider showing your support. A little shoutout on your favorite social platforms using @​LocalAI_OSS and @​mudler_it or joining our sponsors can make a big difference. Also, if you haven't yet joined our Discord, come on over! Here's the link: https://discord.gg/uJAeKSAGDy Every bit of support, every mention, and every star adds up and helps us keep this ship sailing. Let's keep making LocalAI awesome together! Thanks a ton, and here's to more exciting times ahead with LocalAI! ##### What's Changed ##### Bug fixes 🐛 - fix: downgrade torch by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1902 - fix(aio): correctly detect intel systems by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1931 - fix(swagger): do not specify a host by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1930 - fix(tools): correctly render tools response in templates by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1932 - fix(grammar): respect JSONmode and grammar from user input by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1935 - fix(hermes-2-pro-mistral): add stopword for toolcall by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1939 - fix(functions): respect when selected from string by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1940 - fix: use exec in entrypoint scripts to fix signal handling by [@​cryptk](https://togithub.com/cryptk) in [mudler/LocalAI#1943 - fix(hermes-2-pro-mistral): correct stopwords by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1947 - fix(welcome): stable model list by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1949 - fix(ci): manually tag latest images by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1948 - fix(seed): generate random seed per-request if -1 is set by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1952 - fix regression [#​1971](https://togithub.com/mudler/LocalAI/issues/1971) by [@​fakezeta](https://togithub.com/fakezeta) in [mudler/LocalAI#1972 ##### Exciting New Features 🎉 - feat(aio): add intel profile by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1901 - Enhance autogptq backend to support VL models by [@​thiner](https://togithub.com/thiner) in [mudler/LocalAI#1860 - feat(assistant): Assistant and AssistantFiles api by [@​christ66](https://togithub.com/christ66) in [mudler/LocalAI#1803 - feat: Openvino runtime for transformer backend and streaming support for Openvino and CUDA by [@​fakezeta](https://togithub.com/fakezeta) in [mudler/LocalAI#1892 - feat: Token Stream support for Transformer, fix: missing package for OpenVINO by [@​fakezeta](https://togithub.com/fakezeta) in [mudler/LocalAI#1908 - feat(welcome): add simple welcome page by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1912 - fix(build): better CI logging and correct some build failure modes in Makefile by [@​cryptk](https://togithub.com/cryptk) in [mudler/LocalAI#1899 - feat(webui): add partials, show backends associated to models by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1922 - feat(swagger): Add swagger API doc by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1926 - feat(build): adjust number of parallel make jobs by [@​cryptk](https://togithub.com/cryptk) in [mudler/LocalAI#1915 - feat(swagger): update by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1929 - feat: first pass at improving logging by [@​cryptk](https://togithub.com/cryptk) in [mudler/LocalAI#1956 - fix(llama.cpp): set better defaults for llama.cpp by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1961 ##### 📖 Documentation and examples - docs(aio-usage): update docs to show examples by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1921 ##### 👒 Dependencies - ⬆️ Update docs version mudler/LocalAI by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1903 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1904 - ⬆️ Update M0Rf30/go-tiny-dream by [@​M0Rf30](https://togithub.com/M0Rf30) in [mudler/LocalAI#1911 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1913 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1914 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1923 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1924 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1928 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1933 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1934 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1937 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1941 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1953 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1958 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1959 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1964 ##### Other Changes - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1927 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1960 - fix(hermes-2-pro-mistral): correct dashes in template to suppress newlines by [@​mudler](https://togithub.com/mudler) in [mudler/LocalAI#1966 - ⬆️ Update ggerganov/whisper.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1969 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1970 - ⬆️ Update ggerganov/llama.cpp by [@​localai-bot](https://togithub.com/localai-bot) in [mudler/LocalAI#1973 ##### New Contributors - [@​thiner](https://togithub.com/thiner) made their first contribution in [mudler/LocalAI#1860 **Full Changelog**: mudler/LocalAI@v2.11.0...v2.12.0 </details> --- ### Configuration 📅 **Schedule**: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined). 🚦 **Automerge**: Enabled. ♻ **Rebasing**: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox. 🔕 **Ignore**: Close this PR and you won't be reminded about these updates again. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR has been generated by [Renovate Bot](https://togithub.com/renovatebot/renovate). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiIzNy4yODEuMiIsInVwZGF0ZWRJblZlciI6IjM3LjI4MS4yIiwidGFyZ2V0QnJhbmNoIjoibWFzdGVyIiwibGFiZWxzIjpbImF1dG9tZXJnZSIsInVwZGF0ZS9kb2NrZXIvZ2VuZXJhbC9ub24tbWFqb3IiXX0=-->
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This PR add OpenVINO Runtime through Optimum Intel with acceleration for CPU (Intel and ARM)/iGPU/dGPU/NPU
Detailed list of supported Hardware from OpenVINO site.
Dynamic quantization and automatic conversion of Transformers models is not implemented because is slow so the model shall be converted and quantized first: instruction in Optimum Documentation.
I leave here two model for testing, one with 4bit quantization and another with 8bit.
Performance on iGPU UHD 770 (32 EU) with int8 quantized model above is >10tk/s.
Notes for Reviewers
Regression tested properly for CUDA, could not perform extensive testing on XPU due to missing hardware.
Signed commits