# FocalNet

**Category:** Hybrid · **Code:**
[`timm/models/focalnet.py`](https://github.com/huggingface/pytorch-image-models/blob/main/timm/models/focalnet.py) ·
**Architectures:** 12 · **Pretrained weights:** 12 (6 with ImageNet-1k results)

Focal Modulation Networks (FocalNet) replace self-attention with focal modulation, which aggregates context at several
granularities with depthwise convolutions and gates it into each query token.

## Papers

* [Focal Modulation Networks](https://arxiv.org/abs/2203.11926)

## Usage

See the [Quickstart](../quickstart) for loading pretrained weights, image preprocessing and fine-tuning, and [Feature
Extraction](../feature_extraction) for using the models as backbones.

```py
>>> import timm
>>> model = timm.create_model('focalnet_tiny_srf.ms_in1k', pretrained=True)
>>> data_config = timm.data.resolve_model_data_config(model)
```

## Model API support

| Feature | `FocalNet` |
|---|---|
| [`features_only=True` feature maps](../model_api#intermediate-features) | ✓ (strides 4, 8, 16, 32) |
| [`forward_intermediates()`](../model_api#intermediate-features) | ✓ |
| [`prune_intermediate_layers()`](../model_api#intermediate-features) | ✓ |
| [Any input size, no changes needed](../model_api#input-size) | ✓ |
| [`set_input_size()`](../model_api#input-size) | – |
| [`dynamic_img_size=True`](../model_api#input-size) | – |
| [`set_grad_checkpointing()`](../model_api#training-hooks) | ✓ |
| [Layer-wise LR decay (`group_matcher()`)](../model_api#training-hooks) | ✓ |
| [`reset_classifier()`](../model_api#classifier) | ✓ |

✓ supported, ✓ (square) works for square sizes only, ✗ not supported, – not needed (the model already accepts any input
size). Checked by running each feature on a randomly initialized model (`FocalNet` with `focalnet_tiny_srf` at 224),
other architectures of the same class are expected to behave the same. See the [Model API](../model_api) page for
details.

## Architectures

| Architecture | Params (M) | GMACs | Input size | Pretrained weights |
|---|---:|---:|---:|---:|
| `focalnet_base_lrf` | 88.75 | 15.43 | 224 | 1 |
| `focalnet_base_srf` | 88.15 | 15.28 | 224 | 1 |
| `focalnet_huge_fl3` | 745.28 | 118.26 | 224 | 1 |
| `focalnet_huge_fl4` | 686.46 | 118.90 | 224 | 1 |
| `focalnet_large_fl3` | 239.13 | 105.06 | 384 | 1 |
| `focalnet_large_fl4` | 239.32 | 105.20 | 384 | 1 |
| `focalnet_small_lrf` | 50.34 | 8.74 | 224 | 1 |
| `focalnet_small_srf` | 49.89 | 8.62 | 224 | 1 |
| `focalnet_tiny_lrf` | 28.65 | 4.49 | 224 | 1 |
| `focalnet_tiny_srf` | 28.43 | 4.42 | 224 | 1 |
| `focalnet_xlarge_fl3` | 408.79 | 185.61 | 384 | 1 |
| `focalnet_xlarge_fl4` | 409.03 | 185.79 | 384 | 1 |

GMACs and parameter counts are for the default input size and classifier of each architecture, from the [benchmark
results](https://github.com/huggingface/pytorch-image-models/tree/main/results).

## Pretrained weights

Top-1 is ImageNet-1k validation accuracy from
[`results-imagenet.csv`](https://github.com/huggingface/pytorch-image-models/blob/main/results/results-imagenet.csv) at
the listed eval size, - if the weight has no ImageNet-1k result (e.g. pretraining or image-text weights). Classes is the
size of the pretrained classifier, none for weights without one.

| Weight | Input size | Top-1 (%) | Eval size | Classes | License | Hub |
|---|---:|---:|---:|---:|---|---|
| `focalnet_base_lrf.ms_in1k` | 224 | 83.828 | 224 | 1000 | mit | [timm](https://huggingface.co/timm/focalnet_base_lrf.ms_in1k) |
| `focalnet_base_srf.ms_in1k` | 224 | 83.808 | 224 | 1000 | mit | [timm](https://huggingface.co/timm/focalnet_base_srf.ms_in1k) |
| `focalnet_huge_fl3.ms_in22k` | 224 | - | - | 21842 | mit | [timm](https://huggingface.co/timm/focalnet_huge_fl3.ms_in22k) |
| `focalnet_huge_fl4.ms_in22k` | 224 | - | - | none | mit | [timm](https://huggingface.co/timm/focalnet_huge_fl4.ms_in22k) |
| `focalnet_large_fl3.ms_in22k` | 384 | - | - | 21842 | mit | [timm](https://huggingface.co/timm/focalnet_large_fl3.ms_in22k) |
| `focalnet_large_fl4.ms_in22k` | 384 | - | - | 21842 | mit | [timm](https://huggingface.co/timm/focalnet_large_fl4.ms_in22k) |
| `focalnet_small_lrf.ms_in1k` | 224 | 83.508 | 224 | 1000 | mit | [timm](https://huggingface.co/timm/focalnet_small_lrf.ms_in1k) |
| `focalnet_small_srf.ms_in1k` | 224 | 83.428 | 224 | 1000 | mit | [timm](https://huggingface.co/timm/focalnet_small_srf.ms_in1k) |
| `focalnet_tiny_lrf.ms_in1k` | 224 | 82.174 | 224 | 1000 | mit | [timm](https://huggingface.co/timm/focalnet_tiny_lrf.ms_in1k) |
| `focalnet_tiny_srf.ms_in1k` | 224 | 82.146 | 224 | 1000 | mit | [timm](https://huggingface.co/timm/focalnet_tiny_srf.ms_in1k) |
| `focalnet_xlarge_fl3.ms_in22k` | 384 | - | - | 21842 | mit | [timm](https://huggingface.co/timm/focalnet_xlarge_fl3.ms_in22k) |
| `focalnet_xlarge_fl4.ms_in22k` | 384 | - | - | 21842 | mit | [timm](https://huggingface.co/timm/focalnet_xlarge_fl4.ms_in22k) |