Perceiver AR: general-purpose, long-context autoregressive era

Over the previous few years, autoregressive Transformers have introduced a gentle stream of breakthroughs in generative modeling. These fashions generate every factor of a pattern – the pixels of a picture, the characters of a textual content (usually in “token” chunks), the samples of an audio waveform, and so forth – by predicting one factor after the opposite. When predicting the subsequent factor, the mannequin can look again at people who have been created earlier.

Nonetheless, every of a Transformer’s layers grows costlier as extra parts are used as enter, and practitioners can solely afford to coach deep Transformers on sequences not more than about 2,048 parts in size. And so, most Transformer-based fashions ignore all parts past the latest previous (round 1,500 phrases or 1/6 of a small picture) when making a prediction.

In distinction, our not too long ago developed Perceiver fashions give wonderful outcomes on quite a lot of real-world duties with as much as round 100,000 parts. Perceivers use cross-attention to encode inputs right into a latent house, decoupling the enter’s compute necessities from mannequin depth. Perceivers additionally spend a set price, no matter enter dimension, at practically each layer.

Whereas latent-space encoding handles all parts in a single cross, autoregressive era assumes processing occurs one factor at a time. To deal with this drawback, Perceiver AR proposes a easy answer: align the latents one after the other with the ultimate parts of the enter, and thoroughly masks the enter so latents see solely earlier parts.

Perceiver AR maps an enter sequence (P erceiver AR) to a small latent house by cross-attention to provide one latent for every goal token (3 latents proven, one for the targets AR , for End Of Sequence). These latents are then processed by a deep stack of self-attention layers. Perceiver AR could be skilled for end-to-end autoregressive era, all whereas making use of very lengthy enter sequences.

The result’s an structure (proven above) that attends to as a lot as 50x longer inputs as commonplace Transformers, whereas deploying as broadly (and primarily as simply) as commonplace decoder-only Transformers.

As context size or mannequin dimension will increase, the quantity of compute wanted to coach a mannequin grows. We are able to quantify the compute price range for various fashions by measuring their pace on actual {hardware} (steps per second on TPUv3), because the enter context size and mannequin dimension improve. In contrast to different generative fashions like Transformer or Transformer-XL, Perceiver AR decouples enter context size from mannequin depth, permitting us to simply deploy the deep fashions wanted to mannequin lengthy sequences on current-generation TPUs or GPUs.

Perceiver AR scales significantly higher with dimension than each commonplace Transformers and Transformer-XL fashions at a variety of sequence lengths in actual phrases. This property permits us to construct very efficient long-context fashions. For instance, we discover {that a} 60-layer Perceiver AR with context size 8192 outperforms a 42-layer Transformer-XL on a book-length era activity, whereas working quicker in actual wall-clock phrases.

On commonplace, long-context picture (ImageNet 64×64), language (PG-19), and music (MAESTRO) era benchmarks, Perceiver AR produces state-of-the-art outcomes. Rising enter context by decoupling enter dimension from compute price range results in a number of intriguing outcomes:

  • Compute price range could be tailored at eval time, permitting us to spend much less and easily degrade high quality or to spend extra for improved era.
  • A bigger context permits Perceiver AR to outperform Transformer-XL, even when spending the identical on compute. We discover that higher context results in improved mannequin efficiency even at inexpensive scale (~ 1B parameters).
  • Perceiver AR’s pattern high quality reveals a lot much less sensitivity to the order wherein it generates parts. This makes Perceiver AR straightforward to use to settings that don’t have a pure left-to-right ordering, comparable to information like photographs, with construction that spans multiple dimension.

Utilizing a dataset of piano music, we skilled Perceiver AR to generate new items of music from scratch. As a result of every new be aware is predicted primarily based on the complete sequence of notes that got here earlier than, Perceiver AR is ready to produce items with a excessive degree of melodic, harmonic, and rhythmic coherence:


Study extra about utilizing Perceiver AR:

  • Obtain the JAX code for coaching Perceiver AR on Github
  • Learn our paper on arXiv
  • Try our highlight presentation at ICML 2022

See the Google Magenta weblog put up with extra music!

Leave a Comment

%d bloggers like this: