# perceiver-pytorch

**Repository Path**: lsb829/perceiver-pytorch

## Basic Information

- **Project Name**: perceiver-pytorch
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2026-05-13
- **Last Updated**: 2026-05-13

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

<img src="./perceiver.png" width="600px"></img>

## Perceiver - Pytorch

Implementation of <a href="https://arxiv.org/abs/2103.03206">Perceiver</a>, General Perception with Iterative Attention, in Pytorch

<a href="https://www.youtube.com/watch?v=P_xeshTnPZg">Yannic Kilcher explanation!</a>

## Install

```bash
$ pip install perceiver-pytorch
```

## Usage

```python
import torch
from perceiver_pytorch import Perceiver

model = Perceiver(
    input_channels = 3,          # number of channels for each token of the input
    input_axis = 2,              # number of axis for input data (2 for images, 3 for video)
    num_freq_bands = 6,          # number of freq bands, with original value (2 * K + 1)
    max_freq = 10.,              # maximum frequency, hyperparameter depending on how fine the data is
    depth = 6,                   # depth of net
    num_latents = 256,           # number of latents, or induced set points, or centroids. different papers giving it different names
    cross_dim = 512,             # cross attention dimension
    latent_dim = 512,            # latent dimension
    cross_heads = 1,             # number of heads for cross attention. paper said 1
    latent_heads = 8,            # number of heads for latent self attention, 8
    cross_dim_head = 64,
    latent_dim_head = 64,
    num_classes = 1000,          # output number of classes
    attn_dropout = 0.,
    ff_dropout = 0.,
    weight_tie_layers = False    # whether to weight tie layers (optional, as indicated in the diagram)
)

img = torch.randn(1, 224, 224, 3) # 1 imagenet image, pixelized

model(img) # (1, 1000)
```

## Experimental

I have also included a version of Perceiver that includes bottom-up (in addition to top-down) attention, using the same scheme as presented in the original <a href="https://arxiv.org/abs/1810.00825">Set Transformers</a> paper as the <a href="https://github.com/lucidrains/isab-pytorch">Induced Set Attention Block</a>.

You simply have to change the above import to

```python
from perceiver_pytorch.experimental import Perceiver
```

## Citations

```bibtex
@misc{jaegle2021perceiver,
    title   = {Perceiver: General Perception with Iterative Attention},
    author  = {Andrew Jaegle and Felix Gimeno and Andrew Brock and Andrew Zisserman and Oriol Vinyals and Joao Carreira},
    year    = {2021},
    eprint  = {2103.03206},
    archivePrefix = {arXiv},
    primaryClass = {cs.CV}
}
```