Mat Miller
@matdmiller.bsky.social
AI, Engineer, Developer, Traveler, Technologist, @fastdotai Fellow
I like Anthropic's "contextual retrieval" suggestion for creating chunks simonwillison.net/2024/Sep/20/...
Introducing Contextual Retrieval
Here's an interesting new embedding/RAG technique, described by Anthropic but it should work for any embedding model against any other LLM. One of the big challenges in implementing semantic search …
simonwillison.net
Looks like @graze.social exists already! Their programming language is a nested JSON object of filters and Boolean operations: www.graze.social/docs/editor-... bsky.app/profile/graz...
![{
"filter": {
"and": [
{
"attribute_compare": [
"embed.external.uri",
"==",
"https://www.youtube.com/watch?v=E8Ew6K0W3RY"
]
},
{
"regex_matches": ["text", "\\bthe\\b"]
},
{
"regex_negation_matches": ["text", "\\bunwanted_term\\b"]
},
{
"user_network": ["devingaffney.com", "in", "follows"]
},
{
"text_similarity": [
"text",
{
"anchor_text": "This is an important update",
"model_name": "all-MiniLM-L6-v2"
},
">=",
0.3
]
},](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:kft6lu4trxowqmter2b6vg6z/bafkreicria2jounrzuq7xuhmkc6ynhqzafqdfv4umokjpaj3jduvosuxna)
Graze Social@graze.social · 2y ago👀
🔥 Host your apps for FREE and simplify your infrastructure! 🚀 Tired of complex cloud hosting and the recurring bills that come with it? Host locally on hardware you already own and access anywhere with a browser. It's simpler than you think! Learn how in my latest post. 🔗👇
🔥 Host your apps for FREE and simplify your infrastructure! 🚀 Tired of complex cloud hosting and the recurring bills that come with it? Host locally on hardware you already own and access anywhere with a browser. It's simpler than you think! Learn how in my latest post. 🔗👇
This week we released ModernBERT, the first encoder to reach SOTA on most common benchmarks across language understanding, retrieval, and code, while running twice as fast as DeBERTaV3 on short context and three times faster than NomicBERT & GTE on long context.
I'll get straight to the point. We trained 2 new models. Like BERT, but modern. ModernBERT. Not some hypey GenAI thing, but a proper workhorse model, for retrieval, classification, etc. Real practical stuff. It's much faster, more accurate, longer context, and more useful. 🧵
I figured out a prompting pattern for getting Claude to produce fully self-contained Python scripts that execute with "uv run" using PEP 723 inline script dependencies - and now I can one-shot useful Python utilities with it simonwillison.net/2024/Dec/19/...
Building Python tools with a one-shot prompt using uv run and Claude Projects
I’ve written a lot about how I’ve been using Claude to build one-shot HTML+JavaScript applications via Claude Artifacts. I recently started using a similar pattern to create one-shot Python utilities,...
simonwillison.net
HOT 🔥 fastest, most precise, and most capable hand control setup ever... Less than $450 and fully open-source 🤯 by @huggingface, @therobotstudio, @NepYope This tendon-driven technology will disrupt robotics! Retweet to accelerate its democratization 🚀 A thread 🧵
Hey @runme.dev - love your notebook-style script execution! Any plans for Windows native support? Many of us in corporate environments need this without the WSL dependency. As a Jupyter notebooks fan, I can immediately see the value of this but a many of my use cases today would be Windows native.
Excited to share that we have just released genomenotebook 1.0 together with @trichomedoctor.bsky.social ! If you like to analyze genomics data in Jupyter notebooks this is for you. You can get beautiful genome browsers directly in your notebooks: dbikard.github.io/genomenotebo...
genomenotebook
A genome browser in your Jupyter notebook
dbikard.github.io
Non-commercial license, as usual from MSR. 🙄 And of course they don't even give you the option to pay for it.
Tommi Somersuo@tommis.fi · 2y ago
Chat with any open source repo easily. Gitingest (free online tool) turns any GitHub repository into a single markdown file for pasting. Claude artifacts makes this 300k token output pretty easy to work with.
This is a neat new variant on RAG - no vectors, not even full-text search, instead showing the model a header hierarchy and giving it a tool to read the relevant sections My notes here: simonwillison.net/2024/Dec/6/r...
Roaming RAG – make the model find the answers
Neat new RAG technique (with a snappy name) from John Berryman: > The big idea of Roaming RAG is to craft a simple LLM application so that the LLM assistant …
simonwillison.net
John Berryman@jnbrymn.bsky.social · 2y agoTired of wrestling with complex RAG pipelines? 🚀 Enter Roaming RAG: a simpler way to make your LLMs find answers in well-structured docs. No vector databases, no headaches—just rich, structured context. 👉 Read how it works: arcturus-labs.com/blog/2024/11...
As R&D staff @ answer.ai, I work a lot on boosting productivity with AI. A common theme that always comes up is the combination of human+AI. This combination proved to be powerful in our new project ShellSage, which is an AI terminal buddy that learns and teaches with you. A 🧵
I can't begin to describe how life-changing this new project, ShellSage, has been for me over the last few weeks. ShellSage is an LLM that lives in your terminal. It can see what directory you're in, what commands you've typed, what output you got, & your previous AI Q&A's.🧵
Nathan Cooper@ncoop57.bsky.social · 2y agoAs R&D staff @ answer.ai, I work a lot on boosting productivity with AI. A common theme that always comes up is the combination of human+AI. This combination proved to be powerful in our new project ShellSage, which is an AI terminal buddy that learns and teaches with you. A 🧵
For anyone interested in fine-tuning or aligning LLMs, I’m running this free and open course called smol course. It’s not a big deal, it’s just smol. 🧵>>
Amazing demo by Vaibhav Srivastav of structured data extraction running on an LLM that executes entirely in the browser (Chrome only for the moment since it uses WebGPU). https://simonwillison.net/2024/Nov/29/structured-generation-smollm2-webgpu/
Structured Generation w/ SmolLM2 running in browser & WebGPU
Extraordinary demo by Vaibhav Srivastav. Here's Hugging Face's [SmolLM2-1.7B-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct) running directly in a web browser (using WebGPU, so requires Chrome [for the moment](https://github.com/gpuweb/gpuweb/wiki/Implementation-Status)) demonstrating structured text extraction, converting a text …
simonwillison.net
After seeing all those toxic replies to @danielvanstrien.bsky.social , I created a small script to auto-block people. Let's use AI to moderate the replies. gist.github.com/tcapelle/00a... cc. @eugeneyan.bsky.social
A dataset of 1 million or 2 million Bluesky posts is completely irrelevant to training large language models. The primary usecase for the datasets that people are losing their shit over isn't ChatGPT, it's social science research and developing systems that improve Bluesky.
Jeremy Howard @howard.fm · 2y agoDid you know that 99% of email today is spam? Your inbox isn’t 99% spam because AI is used to filter it. The same 99% will happen here too, but if AI researchers continue to get perma-banned for making available the datasets needed to filter it, it’s going to make this platform unusable.
I find strange that engineering the gradient radically to learn better is not a whole field.
TIL how to implement GitHub OAuth for an otherwise completely static website (HTML+JavaScript on GitHub Pages) using Cloudflare Workers: https://til.simonwillison.net/cloudflare/workers-github-oauth
✨ More Bluesky tools for power-users, curious people, and nerds. ✨ One thing that makes this place a better option for anyone wanting a more conscious social experience is its "opensource-ness." It gives us a level of freedom we haven't had since the MySpace era. I missed that. 🧵 1/10
Did you know that 99% of email today is spam? Your inbox isn’t 99% spam because AI is used to filter it. The same 99% will happen here too, but if AI researchers continue to get perma-banned for making available the datasets needed to filter it, it’s going to make this platform unusable.
It's pretty sad to see the negative sentiment towards Hugging Face on this platform due to a dataset put by one of the employees. I want to write a small piece. 🧵 Hugging Face empowers everyone to use AI to create value and is against monopolization of AI it's a hosting platform above all.
Feels good to be mentioned on HN for engineers learning AI 🥰 Helping others is a big reason I write. Here's a list on ML/AI: ## Building AI systems • Patterns for Building LLM-based Systems: eugeneyan.com/writing/llm-... • What We’ve Learned From A Year of Building with LLMs: applied-llms.org
Small yet mighty! 💫 We are releasing SmolVLM: a new 2B small vision language made for on-device use, fine-tunable on consumer GPU, immensely memory efficient 🤠 We release three checkpoints under Apache 2.0: SmolVLM-Instruct, SmolVLM-Synthetic and SmolVLM-Base huggingface.co/collections/...
Thanks to @simonwillison.net for improving this > forked this to add script inline metadata for the dependency and use input() and getpass() to accept the credentials - now you can run it without having to install it first with uv See comment in new gist: gist.github.com/simonw/848a3...
Run like this: uv run https://gist.githubusercontent.com/simonw/848a3b91169a789bc084a459aa7ecf83/raw/44fe7e0b326832e88beb83748b50104e5e7f70d0/follow_theirs.py
Run like this: uv run https://gist.githubusercontent.com/simonw/848a3b91169a789bc084a459aa7ecf83/raw/44fe7e0b326832e88beb83748b50104e5e7f70d0/follow_theirs.py - follow_theirs.py
gist.github.com
Hamel Husain@hamel.bsky.social · 2y agoI wrote this code to follow the same people someone else is following. I figured that would fix my feed esp if someone is having a good experience I can just "have what they are having" gist.github.com/hamelsmu/fb9...
It is their implementation of PEP 723 - it's really handy peps.python.org/pep-0723/
PEP 723 – Inline script metadata | peps.python.org
This PEP specifies a metadata format that can be embedded in single-file Python scripts to assist launchers, IDEs and other external tools which may need to interact with such scripts.
peps.python.org