Reynold Xin(@rxin) 's Twitter Profileg
Reynold Xin

@rxin

Cofounder @Databricks

ID:17712257

linkhttp://www.rxin.org calendar_today28-11-2008 16:48:59

2,5K Tweets

11,8K Followers

728 Following

Julia Neagu(@JuliaANeagu) 's Twitter Profile Photo

Databricks Mosaic Research DBRX outperforms OpenAI GPT-4 on realistic, domain-specific benchmark datasets. For example, on a customer support summarization use-case👇👇👇
Still neck and neck but it shows that open models can be the no-brainer choice for actual enterprise applications.

@DbrxMosaicAI DBRX outperforms @OpenAI GPT-4 on realistic, domain-specific benchmark datasets. For example, on a customer support summarization use-case👇👇👇 Still neck and neck but it shows that open models can be the no-brainer choice for actual enterprise applications.
account_circle
Ali Ghodsi(@alighodsi) 's Twitter Profile Photo

Today we released an open source model, DBRX, that beats all previous open source models on the standard benchmarks. The model itself is a Mixture of Experts (MoE), that's roughly twice the brains (132B) but half the cost (36B) of Llama2-70B. Making it both smart and cheap. Since…

account_circle
Mihir Patel(@mvpatel2000) 's Twitter Profile Photo

🚨 Announcing DBRX-Medium 🧱, a new SoTA open weights 36b active 132T total parameter MoE trained on 12T tokens (~3e24 flops). Dbrx achieves 150 tok/sec while clearing a wide variety of benchmarks. Deep dive below! 1/N

🚨 Announcing DBRX-Medium 🧱, a new SoTA open weights 36b active 132T total parameter MoE trained on 12T tokens (~3e24 flops). Dbrx achieves 150 tok/sec while clearing a wide variety of benchmarks. Deep dive below! 1/N
account_circle
Databricks(@databricks) 's Twitter Profile Photo

Databricks Ventures is excited to share its latest investment — @Glean! 🎉

We’re thrilled to accelerate their momentum as the industry’s leading AI-powered work assistant, & bring the secure deployment of & knowledge building to all enterprises👇bit.ly/3wqsQ5s

account_circle
Metronome(@getmetronome) 's Twitter Profile Photo

🎉We’re excited to announce our $43M Series B funding led by NEA, with continued support from a16z and General Catalyst.

metronome.com/blog/series-b?…

account_circle
Ram Parameswaran(@_ram_) 's Twitter Profile Photo

A good listen about the foundational product (spark) of a foundational software company (databricks). well done Reynold Xin and great interview Sudip Chakrabarti

sudipchakrabarti.substack.com/p/from-spark-t…

account_circle
Reynold Xin(@rxin) 's Twitter Profile Photo

This is the first time that I talked extensively about Apache Spark's history, design philosophies, and the evolutions. Over the years, there have been a lot of misconceptions (why it came into existence, memory vs disk). This should explain all of them.

sudipchakrabarti.substack.com/p/from-spark-t…

account_circle
Databricks(@databricks) 's Twitter Profile Photo

Databricks Vector Search is in public preview!

Augment your apps and synchronize source data, eliminating the need to write and maintain separate data pipelines.

Find out why our customers already love it and how you can easily get started 👇 dbricks.co/3R9m2A1

account_circle
Reynold Xin(@rxin) 's Twitter Profile Photo

Today, more than 80% of the table metadata updates on Databricks are AI-assisted. It took 2 engineers, 1 month, and less than $1000 in compute cost to develop the custom LLM for this task. databricks.com/blog/creating-…

account_circle
Ali Ghodsi(@alighodsi) 's Twitter Profile Photo

The founders of Databricks put together this strategy blog on where we think data platforms are headed in the future. We're moving Databricks quickly in this direction. This is very exciting and is the outcome of the MosaicML acquisition we did earlier this year!…

account_circle
Adi Polak(@AdiPolak) 's Twitter Profile Photo

“ Solid understanding of Spark and ability to write, debug, and optimize Spark code.  “

I was today years old when I discovered that openAI relies heavily on Spark.

“ Solid understanding of Spark and ability to write, debug, and optimize Spark code.  “ I was today years old when I discovered that openAI relies heavily on Spark.
account_circle
Madhav Singhal(@madhavsinghal_) 's Twitter Profile Photo

At Replit, we extensively use Spark on Databricks for all our training data work.

We run highly customized transformations on code like parsing, deduping, PII redaction, code filtering, tokenization, and more, written with low-level Spark primitives.

High quality data is key.

account_circle
Databricks(@databricks) 's Twitter Profile Photo

The trade-offs of Query Optimization is a *perfect* problem for . Chief Architect Reynold Xin walking us through new prediction optimizations in Databricks!

The trade-offs of Query Optimization is a *perfect* problem for #AI. Chief Architect @rxin walking us through new prediction optimizations in Databricks! #LakehouseAI #DataAISummit
account_circle
Databricks(@databricks) 's Twitter Profile Photo

We’re over the moon to announce we’ve agreed to acquire ., a leading generative AI platform. Together, we’ll make generative AI accessible for every organization, so you can build, own & secure models with your data. databricks.com/mosaic-news

account_circle
Databricks(@databricks) 's Twitter Profile Photo

🚨SOLD OUT 🚨

We’ve sold out in-person tickets for ! But you can still catch all the keynotes, select technical sessions, and live-streamed community events when you join us virtually! And it’s entirely free. Hurry & register: bit.ly/3ICPryg

🚨SOLD OUT 🚨 We’ve sold out in-person tickets for #DataAISummit! But you can still catch all the keynotes, select technical sessions, and live-streamed community events when you join us virtually! And it’s entirely free. Hurry & register: bit.ly/3ICPryg
account_circle
Ali Ghodsi(@alighodsi) 's Twitter Profile Photo

Big news: we've agreed to acquire ., a leading generative AI platform. I couldn’t be more excited to join forces once the deal closes. databricks.com/mosaic-news

account_circle
Jonathan Frankle(@jefrankle) 's Twitter Profile Photo

MPT-30B is here! Same MPT architecture, 30B parameters, > 1T tokens, 8k context window, trained on H100s, great perf (esp on coding), single-GPU inference, commercially usable, and massively upgraded instruct and chat datasets. Take it for a spin! huggingface.co/spaces/mosaicm…

account_circle
Kent Marten(@datakmart) 's Twitter Profile Photo

Learn about Mosaic, a Databricks Labs project, for geospatial processing. Tuesday, June 28 @ 4:45pm. databricks.com/dataaisummit/s…

account_circle