On M1/M2, it offers a convenient single binary deployment, thanks to Rust. You can find the latest release at https://github.com/TabbyML/tabby/releases/tag/latest
(Disclaimer: I am the author)
1,499 karma · joined March 30, 2011
Socials: - cal.com/mengzhang - linkedin.com/in/mzh0 - github.com/wsxiaoys
---
On M1/M2, it offers a convenient single binary deployment, thanks to Rust. You can find the latest release at https://github.com/TabbyML/tabby/releases/tag/latest
(Disclaimer: I am the author)
Model supports list: GPT-2, GPT-J, GPT-NeoX, OPT, BLOOM, LLAMA, T5, WHISPER
( Found this library during my research on alternatives to triton/FasterTransformer in Tabby https://github.com/TabbyML/tabby)
Self-hosted AI coding assistant. An opensource / on-prem alternative to GitHub Copilot.
Project: https://github.com/TabbyML/tabby
Show HN post: https://news.ycombinator.com/item?id=35470915
We have an ambitious roadmap ahead, and we need your help! If you are interested in being an early member of a fast-growing startup team(spoiler: 100% remote , open source , and transparent compensation). Please reach out and apply at https://tabbyml.notion.site/Careers-35b1a77f3d1743d9bae06b7d...
[1] https://news.ycombinator.com/item?id=35471882
[2] https://news.ycombinator.com/item?id=35471390
I hope these can be of help and answer your questions as well!
Related research works include [1]. (Hints: combine code search and LLM).
[1]: RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation https://arxiv.org/abs/2303.12570
This also reveals Tabby's roadmap beyond other OSS work like Fauxipilot :)
In the meantime, I'm training language-specific base models using NeoX blocks. I hope to release them soon.
The Huggingface Space demo is running on a derived model of Salesforce/codegen-350M-multi
[1]: https://github.com/TabbyML/tabby/blob/main/tabby/tools/repos...
[2]: https://github.com/TabbyML/tabby/blob/main/tabby/tools/train...
Local context is definitely the key factor for small models achieving better quality than copilot (related: [1], [2]) .
One things I’d really wanna have in Sourcegraph: A Search API supports custom retrieval / ranking. Research works (e.g [2]) show simple BoW fetched context is more efficient for code completion tasks.
Disclaimer: I’m building https://github.com/TabbyML/tabby an open source alternative of copilot.
There’re recent works show a potential of beating Copilot’s performance (e.g https://arxiv.org/abs/2303.12570) with much smaller models (500M vs 10B+).
Inspired by these work, I’m building Tabby (https://github.com/TabbyML/tabby), a OSS GitHub Copilot alternative. Hopefully it could make low cost AI coding accessible to everyone.
Great time to start an awesome AI coding list!
flask: https://quick-question.fly.dev/pallets/flask
diffusers: https://quick-question.fly.dev/huggingface/diffusers
underlying it's powered by Next.JS / R3F / ThreeJS / WebGL PathTracing
Highly recommend to any team wanna build their internal machine learning platform on cloud.
1. Slot based positions instead of ticker based.
2. Cash management.
3. Performance tracking based on actual asset rather than latest asset.
Suggestions for title welcome
I understand the benefits of Immutable pattern, but it'll be nice if datasettle could provide interface with single non-frequent writer(for example, most FRED data is updated seasonally, some of them updates daily) so the pipeline pulling data from some source could be automated
Gambit Scheme: A fast scheme implementation. http://github.com/gambit/gambit
Gerbil Scheme: Provides full module and syntactic tower on top of Gambit Scheme. https://github.com/vyzo/gerbil
Full tail recursion scheme implementation by never "return" in C