LAION-5B: 5 billion image-text-pairs dataset (with the authors)

May 02, 2022

|

35 views

Yannic Kilcher

Details

#laion #clip #dalle LAION-5B is an open, free dataset consisting of over 5 billion image-text-pairs. Today's video is an interview with three of its creators. We dive into the mechanics and challenges of operating at such large scale, how to keep cost low, what new possibilities are enabled with open datasets like this, and how to best handle safety and legal concerns. OUTLINE: 0:00 - Intro 1:30 - Start of Interview 2:30 - What is LAION? 11:10 - What are the effects of CLIP filtering? 16:40 - How big is this dataset? 19:05 - Does the text always come from the alt-property? 22:45 - What does it take to work at scale? 25:50 -When will we replicate DALL-E? 31:30 - The surprisingly efficient pipeline 35:20 - How do you cover the S3 costs? 40:30 - Addressing safety & legal concerns 55:15 - Where can people get started? References: LAION website: https://laion.ai/ LAION Discord: https://discord.com/invite/mVcgxMPD7e LAION-5B: https://laion.ai/laion-5b-a-new-era-of-open-large-scale-multi-modal-datasets/ img2dataset tool: https://github.com/rom1504/img2dataset LAION-400M: https://paperswithcode.com/dataset/laion-400m Links: TabNine Code Completion (Referral): http://bit.ly/tabnine-yannick YouTube: https://www.youtube.com/c/yannickilcher Twitter: https://twitter.com/ykilcher Discord: https://discord.gg/4H8xxDF BitChute: https://www.bitchute.com/channel/yannic-kilcher LinkedIn: https://www.linkedin.com/in/ykilcher BiliBili: https://space.bilibili.com/2017636191 If you want to support me, the best thing to do is to share out the content :) If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this): SubscribeStar: https://www.subscribestar.com/yannickilcher Patreon: https://www.patreon.com/yannickilcher Bitcoin (BTC): bc1q49lsw3q325tr58ygf8sudx2dqfguclvngvy2cq Ethereum (ETH): 0x7ad3513E3B8f66799f507Aa7874b1B0eBC7F85e2 Litecoin (LTC): LQW2TRyKYetVC8WjFkhpPhtpbDM4Vw7r9m Monero (XMR): 4ACL8AGrEo5hAir8A9CeVrW8pEauWvnp1WnSDZxW7tziCDLhZAGsgzhRQABDnFy8yuM9fWJDviJPHKRjV4FWt19CJZN9D4n

0:00 - Intro 1:30 - Start of Interview 2:30 - What is LAION? 11:10 - What are the effects of CLIP filtering? 16:40 - How big is this dataset? 19:05 - Does the text always come from the alt-property? 22:45 - What does it take to work at scale? 25:50 -When will we replicate DALL-E? 31:30 - The surprisingly efficient pipeline 35:20 - How do you cover the S3 costs? 40:30 - Addressing safety & legal concerns 55:15 - Where can people get started?

Category: Misc.

Category: Research Paper

Category: Research Talk

Comments

loading...

Reactions (0) | Note

📝 No reactions yet

Be the first one to share your thoughts!

Reactions(0)

Note

loading...

Recommended

SREcon15 - From Zero to Hero: Recommended Practices for Training your Ever-Evolving SRE Teams

USENIX | Apr 13, 2015

SREcon15 - Architecting and Launching the Halo 4 Services

USENIX | Apr 13, 2015

SREcon15 - Being Afraid—How Paranoia at Dropbox Protects Your Data

USENIX | Apr 13, 2015

ICAPS 2014: Jendrik Seipp on "Diverse and Additive Cartesian Abstraction Heuristics"

ICAPS | Apr 14, 2015

ICAPS 2014: Bram Ridder on "Heuristic Evaluation Based on Lifted Relaxed Planning Graphs"

ICAPS | Apr 14, 2015