MeerkatML

An interactive programming tool for unstructured data, built around LLMs and FMs. Built @HazyResearch by @krandiash, @eyuboglusabri, @jundesai. See

[MeerkatML @MeerkatML]

New demo up 🚀

[Arjun Desai @jundesai]

Want to get the latest ML tools into your labeling workflows?

Check out this @MeerkatML demo using @MetaAI’s Segment Anything Model (SAM) to assist with fine-grained image labeling in real time!

[MeerkatML @MeerkatML]

Llamas are nothing without Meerkats

[Karan Goel @krandiash]

We built a data exploration dashboard that we shipped with @togethercompute's new Red Pajama LLM data release!

We embedded the entire Github subset of Red Pajama (releasing indexes + embeddings soon!).

Built in 100 lines of Python with @MeerkatML 🚀

[Dan Fu @realDanFu]

Super excited to release the RedPajama dataset - a new, fully open *1.2 trillion token* dataset following the LLaMA recipe. A first step towards creating leading, fully open-source large language models.

https://t.co/3ao0ryRjT2

[Karan Goel @krandiash]

Excited to see the RedPajama dataset released: check out the @MeerkatML data exploration dashboard we put together in a collaboration with @togethercompute as part of this release 🚀

We’ll continue to update and add to that in the RedPajama repo!

[MeerkatML @MeerkatML]

Meerkat data frames 🤝 3D objects

[Sabri Eyuboglu @EyubogluSabri]

This is a super fun dataset.

Started exploring it in a Jupyter Notebook by loading it into a Meerkat data frame. No lengthy download required. 1/3
https://t.co/64a6ijhKcJ
https://t.co/MnWQGmXuqW

[MeerkatML @MeerkatML]

We’ll be talking more about Meerkat 🤝 Medicine @StanfordAIMI 3pm PST, check it out!

[Stanford AIMI @StanfordAIMI]

@jundesai @MeerkatML @BraTS_challenge So excited for your talk today! Join us live using this webinar link today at 3pm PST!: https://t.co/DTl2ORnsky

[MeerkatML @MeerkatML]

We’ve got 🧠 MRI support! 3D volumes, segmentation overlays and more 🚀

If you’re in the medical community, hit us up, we���ve got even more cool stuff planned!

[Arjun Desai @jundesai]

Want to visualize your large medical imaging datasets? Now you can with @MeerkatML!

Check out this custom interface for visualizing *all* 3D 🧠 MRIs from @BraTS_challenge. Supports reformats, segmentation overlays 🎨, and more!

All with <10 lines of code in native Python 🐍

[Karan Goel @krandiash]

A very short blog post on 3 directions for data tools I’m personally excited about in the era of GPT-4.

We’re working on these in @MeerkatML (stay tuned for something cool coming soon!)

https://t.co/wc9710PfSn

[MeerkatML @MeerkatML]

Check out our new talk about how the 🩺 medical AI community can use Meerkat for
- 🔎 data exploration and search
- 🐛 error analysis
- 🏷️ annotation/labeling
- 👨‍🎨 generative workflows
in a way that brings together clinicians and engineers!

Talk by @EyubogluSabri + @jundesai 🚀

[Karan Goel @krandiash]

🎥 New Talk about @MeerkatML on @MedaiStanford YouTube: https://t.co/p6V5s85WuV

Data is the language of communication between domain experts (clinicians) and ML people.

Watch how @MeerkatML brings together clinicians and ML folks into a single workflow, designed around data!

[Karan Goel @krandiash]

🚀 ChatGPT / GPT-4 for querying and asking questions on codebases

Point to any GitHub repo, and get an index that is used to answer questions. Use --prompt-only mode if you can only access GPT-4 via ChatGPT to copy-paste.

Built with @MeerkatML!

https://t.co/NH0nO18wtV

[MeerkatML @MeerkatML]

Meerkat is kind of like Seaborn but for highly interactive visualizations: pass in a data frame and get a useful viz for your data!

[Karan Goel @krandiash]

🚀 Demo: 🖼️ Custom Visualizations in @MeerkatML

This is super cool: a fully interactive viz in your notebook cell in 7 lines of code.

Plug in any embeddings and works on any data type out-of-the-box: images, text, audio, <your datatype>!

This demo shows this for audio data!

[MeerkatML @MeerkatML]

Meerkat ❤️ 🤗 datasets 🔥

[Karan Goel @krandiash]

🤗 @MeerkatML now has super good @huggingface datasets integration!

Our data frames (arrow-backed) can load in any HF dataset (like audio in this gif!) and give you a much nicer interactive experience in a notebook -- we're multimodal from the ground up 🚀

[shreya rajpal @ShreyaR]

Excited to release ✨Guardrails AI✨— an open-source package to add SLAs for LLM outputs!

Guardrails supports
🌟 pydantic-style validation of LLM outputs
🌟 corrective actions (e.g. reasking LLM) when needed
🌟 structure and type guarantees (e.g. JSON)

https://t.co/4cfvbNsEaO

[MeerkatML @MeerkatML]

Meerkat meets medical imaging!

If you want to use something like this in your medical data use cases, drop us a DM!

We’re looking to help users who want to interact more with data and models, while gathering feedback about our tools 🛠️

[Arjun Desai @jundesai]

Interested in using foundation models to wrangle your medical imaging data?

Check out this @MeerkatML demo for generating and exploring chest X-rays!

🌐 https://t.co/hPt84rBTPm

#Radiology #AI

[MeerkatML @MeerkatML]

A new version of Meerkat is out!

0.4.3 includes a couple of fixes:
- add support for Python 3.7
- fix Match component support on Linux machines