Sam Shah

VP Eng @Databricks. Past: Founder @SkipFlag (acq @Workday), @Accel, @LinkedIn, PhD @UMich. 🇨🇦🇺🇸

[Mike Conover @vagabondjack]

Just in time for the weekend, all the details on how we created Databricks’ Dolly on this episode of the Latent Space podcast w @FanaHOVA & @swyx.

From the nitty gritty of instruction tuning to how AI will change society, this was a great conversation.

https://t.co/XA7C3Yk4pM

[Mike Conover @vagabondjack]

With the Dolly 2.0 release yesterday we heard a lot of people saying they wanted smaller models that can fit on consumer grade GPU’s.

Here you go! Checkpoints from 2.8B and 6.9B now on the Databricks Hugging Face page! 🐑🚀

https://t.co/l4G7zPe3GZ

[Rachel Woods @rachel_l_woods]

@Databricks just released the first commercially usable, instruction-tuned open-source LLM

Might be a bigger step forward for open-source than LLaMA, Alpaca, et al

I'm sitting here reading the story of how they got there and it's just incredible:

[Ali Ghodsi @alighodsi]

Free Dolly! Introducing the first *commercially viable*, open source, instruction-following LLM. Dolly 2.0 is available for commercial applications without having to pay for API access or sharing data with 3rd parties. https://t.co/HtRVSmArFU

[Pete Skomoroch @peteskomoroch]

Databricks just released Dolly 2.0: the first open-source, instruction-following LLM licensed for commercial use.

- 12B parameter LLM based on EleutherAI pythia model family
- training code, dataset, and model weights all licensed for commercial use.

https://t.co/xbEeTZ3lMb

[Pete Skomoroch @peteskomoroch]

Dolly, a ChatGPT-like LLM anyone can train at low cost was just open sourced by @databricks. The team releasing this includes former @SkipFlag teammates @vagabondjack and Matt Hayes working with co-founder @sam_shah

[JosH100 @josh_wills]

The story of one crazy week in March (and several slightly less crazy weeks in April, May, and June) working on modeling the COVID-19 pandemic in California:

https://t.co/5uazF8oKyf

This was the best working experience of my life and an absolute privilege to be a part of it.

[Sam Shah @sam_shah]

@gregthursam @josh_wills Greg, honor was mine, your help was completely invaluable!

[Sam Shah @sam_shah]

@dauber Congrats Mike! 🍾

[Software Engineering Daily @software_daily]

COVID Modeling with Josh Wills and Sam Shah @josh_wills @sam_shah https://t.co/WW2t4MCgwS

[dj patil @dpatil]

6/ Here’s one example of @sam_shah & @josh_wills w/ help from @Werner + the @awscloud team https://t.co/Bh8VdnLCnf

It's amazing & outstanding what they have delivered

[Sam Shah @sam_shah]

@esammer I’m really sorry to hear about your dad, Eric.

[Pete Skomoroch @peteskomoroch]

Many companies struggle to get impact from machine learning projects, so @mikeloukides and I wrote this guide: "What you need to know about product management for AI" https://t.co/45kfIFYVTA

[Sam Shah @sam_shah]

@chrismessina Thanks @chrismessina!

[Sam Shah @sam_shah]

What are a company's Coronavirus cancellation and change policies? We're crowdsourcing these policies at https://t.co/eJ2L8KPVPL to give consumers one place to view every company's policy. Users can also add policies and update new information, so please contribute if you can.

[Pete Skomoroch @peteskomoroch]

If you are struggling with changing travel plans during the Coronavirus pandemic, check out https://t.co/ggS6rTkPPF

The #COVIDー19 Refund Tracker from @sam_shah gives consumers one place to view every organization’s policy, users can also add policies & update new information.

[Sam Shah @sam_shah]

@jdetle models:
"in production" = revenue generating (w/ SLA)
"production ready" = can be operated/managed like any other revenue generating software

[Sam Shah @sam_shah]

Privately owned publicly accessible spots (POPOS) are public spaces (plazas, atriums, parks), usable by all, but funded as part of a development requirement for San Francisco. Hard to find, but often are empty quiet places to work or take a meeting. #sf https://t.co/0HpDmPQdf1

[Sam Shah @sam_shah]

RecSys 2020 challenge: ~200m tweets with daily GDPR scrubbing (https://t.co/FV18OETTxz)

[Sam Shah @sam_shah]

Worker-in-the-loop, where a human corrects/validates results of an AI to an end user (vs generating labels fed back into training of the AI), will become increasingly commonplace, esp in the enterprise. Many open problems here, great post: https://t.co/4INvwjEjN2