Happy New Year! In 2020, I trained 417 models over 1,851 hours, with 1,140 hours of GPU training time. Get insights on your model training with @wandb! #myyearwrapped
Introducing Jukebox, a neural net that generates music, including rudimentary singing, as raw audio in a variety of genres and artist styles. We're releasing a tool for everyone to explore the generated samples, as well as the model and code: https://t.co/EUq7hNZv62
VQVAE-2 finally out!
Powerful autoregressive models in a hierarchical compressed latent space. No modes were collapsed in the creation of these samples ;)
Arixv: https://t.co/nQH5dzQhib
With @catamorphist and @vinyals
More samples and details 👇 [thread]
Unsupervised pre-training now outperforms supervised learning on ImageNet for any data regime (see figure) and also for transfer learning to Pascal VOC object detection
https://t.co/MvRu3dqTxk
Unsupervised pre-training now outperforms supervised learning on ImageNet for any data regime (see figure) and also for transfer learning to Pascal VOC object detection
https://t.co/MvRu3dqTxk