Today @marqo_ai open-weight (Apache 2.0) released the two best embedding models for ecommerce search and recommendations available anywhere. Marqo ecommerce models significantly outperform models from Amazon, Google, Cohere and Jina (see below). Fun fact: we had to create a significantly smaller and easier evaluation dataset just to accommodate some of the private models! + Up to 88% improvement on the best private model, Amazon-Titan-Multimodal (and better than Google Vertex, Cohere). + Up to 31% improvement on the best open source model, ViT-SO400M-14-SigLIP. + 5ms single text/image inference (A10g). + Up to 231% improvement over other bench-marked models (see blog below). + Evaluated on over 4M products across 10,000's of categories. + Detailed performance comparisons across three major tasks: Text2Image, Category2Image, and AmazonProducts-Text2Image. + Released 2 evaluation datasets: GoogleShopping-1m and AmazonProducts-3m. + Released evaluation code. + Apache 2.0 model weights available on @huggingface and to test out on Hugging Face Spaces.

@SeunghyunSEO7 We see this a lot in contrastive learning which also couples problem difficulty

Seems to be quite noticeable style differences between o1 and gpt-4o for python. o1 is technically correct but seems to completely ignore a lot of popular ways of doing things. I suspect it is the training data, generated vs scraped?

@AravSrinivas Bluetooth connectivity

a few more

@elder_plinius

@karpathy patches

@Spotify did you kill the DJ?

I just...asked it...

@indigoorton Cloudflare sales playing 4D chess

@LoubnaBenAllal1 😢

RT @marqo_ai: Personalised feeds, starter packs, and lists? This platform is next level for creative content sharing Check out our @bluesk…

Come say hi!

RT @elsleightholm: breaking down @marqo_ai's new, state-of-the art embedding models for #ecommerce 🔥 the best part, these models are open…

Love to see more data tooling!

RT @marqo_ai: What a week for @marqo_ai on @huggingface! 🤗 1. We released two of the best embedding models for #Ecommerce Search and Recom…

@leavittron @datologyai Very interesting! Why not use SigLIP instead of CLIP?