@Tim_Dettmers@Rghv_Bali - this works like a charm. Especially - the 33B model. Much better output when compared to the 7B model. Inference time is low aswell!
Join us at #ODSCAPAC to learn about interactive ground truth data annotation, model training, building lighter models for fast inference, and more with these incredible speakers. Register now: https://t.co/mnoYF7sb0O #DeepLearning@AV_Kishore
@MDJDPathology@PacktPub Thank you for the shout out @MDJDPathology Happy that the readers find it useful - makes all the effort in writing the book, worth it.
Next reading has just arrived in the mail today! "Modern Computer Vision with PyTorch" by Kishore Ayyadevara (@AV_Kishore) and Yeshwanth Reddy (@PacktPub)
Really excited to start reading this complete masterpiece!
@st3llasia@PacktPub Thank you for the shoutout @st3llasia - it is a comprehensive book from the basics to the very latest in Computer Vision. Looking forward to your review of it.
MediaPipe (https://t.co/aTZEc9jaS6) has updated its #OpenSource, on-device, real-time hand tracking solution so that it's 15% faster, 37% more accurate and achieves 69% higher recall. Check it out at https://t.co/qDszc1phkM to track fast gestures, 👊 and 👋 with ease!
Our pharmacists here @DevotedHealth just blew my mind with this factoid😱
The US makes up ~ 5% of the world’s population, yet utilizes 80% of all opioids & 99% of all hydrocodone. Opioid prescriptions are 7% of the 4.2 billion prescriptions dispensed in 2012
Congratulations to James Zou, et al, for today's publication of a primer on deep learning in genomics in Nature Genetics, with an accompanying interactive online tutorial. Access primer at: https://t.co/3yIWOC146J and tutorial at: https://t.co/3YYMAT92OX
This is stunning work from @dcpage3, who has smashed our DAWNBench CIFAR10 training record, and written a fascinating and detailed series explaining all the improvements he made.
All researchers interested in model speed or accuracy need to read this!
https://t.co/pd6oJhqfX9