Add ConvNeXt-V2 support (model additions and weights) (huggingface#1614)

* Add ConvNeXt-V2 support (model additions and weights)

* ConvNeXt-V2 weights on HF Hub, tweaking some tests

* Update README, fixing convnextv2 tests

Author: rwightman

README.md

@@ -28,6 +28,11 @@ For a few months now, `timm` has been part of the Hugging Face ecosystem. Yearly
 If you have a couple of minutes and want to participate in shaping the future of the ecosystem, please share your thoughts:
 [**hf.co/oss-survey**](https://hf.co/oss-survey) 🙏
 
+### Jan 5, 2023
+* ConvNeXt-V2 models and weights added to existing `convnext.py`
+  * Paper: [ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders](http://arxiv.org/abs/2301.00808)
+  * Reference impl: https://github.com/facebookresearch/ConvNeXt-V2 (NOTE: weights currently CC-BY-NC)
+
 ### Dec 23, 2022 🎄☃
 * Add FlexiViT models and weights from https://github.com/google-research/big_vision (check out paper at https://arxiv.org/abs/2212.08013)
   * NOTE currently resizing is static on model creation, on-the-fly dynamic / train patch size sampling is a WIP
@@ -396,6 +401,7 @@ A full version of the list below with source links can be found in the [document
 * CoaT (Co-Scale Conv-Attentional Image Transformers) - https://arxiv.org/abs/2104.06399
 * CoAtNet (Convolution and Attention) - https://arxiv.org/abs/2106.04803
 * ConvNeXt - https://arxiv.org/abs/2201.03545
+* ConvNeXt-V2 - http://arxiv.org/abs/2301.00808
 * ConViT (Soft Convolutional Inductive Biases Vision Transformers)- https://arxiv.org/abs/2103.10697
 * CspNet (Cross-Stage Partial Networks) - https://arxiv.org/abs/1911.11929
 * DeiT - https://arxiv.org/abs/2012.12877
@@ -418,6 +424,7 @@ A full version of the list below with source links can be found in the [document
     * Single-Path NAS - https://arxiv.org/abs/1904.02877
     * TinyNet - https://arxiv.org/abs/2010.14819
 * EVA - https://arxiv.org/abs/2211.07636
+* FlexiViT - https://arxiv.org/abs/2212.08013
 * GCViT (Global Context Vision Transformer) - https://arxiv.org/abs/2206.09959
 * GhostNet - https://arxiv.org/abs/1911.11907
 * gMLP - https://arxiv.org/abs/2105.08050

tests/test_models.py

@@ -38,7 +38,7 @@
         '*efficientnet_l2*', '*resnext101_32x48d', '*in21k', '*152x4_bitm', '*101x3_bitm', '*50x3_bitm',
         '*nfnet_f3*', '*nfnet_f4*', '*nfnet_f5*', '*nfnet_f6*', '*nfnet_f7*', '*efficientnetv2_xl*',
         '*resnetrs350*', '*resnetrs420*', 'xcit_large_24_p8*', 'vit_huge*', 'vit_gi*', 'swin*huge*',
-        'swin*giant*']
+        'swin*giant*', 'convnextv2_huge*']
     NON_STD_EXCLUDE_FILTERS = ['vit_huge*', 'vit_gi*', 'swin*giant*', 'eva_giant*']
 else:
     EXCLUDE_FILTERS = []
@@ -129,7 +129,7 @@ def test_model_backward(model_name, batch_size):
 
 
 @pytest.mark.timeout(300)
-@pytest.mark.parametrize('model_name', list_models(exclude_filters=NON_STD_FILTERS))
+@pytest.mark.parametrize('model_name', list_models(exclude_filters=NON_STD_FILTERS, include_tags=True))
 @pytest.mark.parametrize('batch_size', [1])
 def test_model_default_cfgs(model_name, batch_size):
     """Run a single forward pass with each model"""
@@ -191,7 +191,7 @@ def test_model_default_cfgs(model_name, batch_size):
 
 
 @pytest.mark.timeout(300)
-@pytest.mark.parametrize('model_name', list_models(filter=NON_STD_FILTERS, exclude_filters=NON_STD_EXCLUDE_FILTERS))
+@pytest.mark.parametrize('model_name', list_models(filter=NON_STD_FILTERS, exclude_filters=NON_STD_EXCLUDE_FILTERS, include_tags=True))
 @pytest.mark.parametrize('batch_size', [1])
 def test_model_default_cfgs_non_std(model_name, batch_size):
     """Run a single forward pass with each model"""
@@ -304,7 +304,7 @@ def test_model_forward_torchscript(model_name, batch_size):
 
 
 @pytest.mark.timeout(120)
-@pytest.mark.parametrize('model_name', list_models(exclude_filters=EXCLUDE_FILTERS + EXCLUDE_FEAT_FILTERS))
+@pytest.mark.parametrize('model_name', list_models(exclude_filters=EXCLUDE_FILTERS + EXCLUDE_FEAT_FILTERS, include_tags=True))
 @pytest.mark.parametrize('batch_size', [1])
 def test_model_forward_features(model_name, batch_size):
     """Run a single forward pass with each model in feature extraction mode"""

timm/layers/__init__.py

@@ -26,7 +26,7 @@
 from .inplace_abn import InplaceAbn
 from .linear import Linear
 from .mixed_conv2d import MixedConv2d
-from .mlp import Mlp, GluMlp, GatedMlp, ConvMlp
+from .mlp import Mlp, GluMlp, GatedMlp, ConvMlp, GlobalResponseNormMlp
 from .non_local_attn import NonLocalAttn, BatNonLocalAttn
 from .norm import GroupNorm, GroupNorm1, LayerNorm, LayerNorm2d
 from .norm_act import BatchNormAct2d, GroupNormAct, convert_sync_batchnorm

timm/layers/grn.py

@@ -0,0 +1,39 @@
+""" Global Response Normalization Module
+
+Based on the GRN layer presented in
+`ConvNeXt-V2 - Co-designing and Scaling ConvNets with Masked Autoencoders` - https://arxiv.org/abs/2301.00808
+
+This implementation
+* works for both NCHW and NHWC tensor layouts
+* uses affine param names matching existing torch norm layers
+* slightly improves eager mode performance via fused addcmul
+
+Hacked together by / Copyright 2023 Ross Wightman
+"""
+
+import torch
+from torch import nn as nn
+
+
+class GlobalResponseNorm(nn.Module):
+    """ Global Response Normalization layer
+    """
+    def __init__(self, dim, eps=1e-6, channels_last=True):
+        super().__init__()
+        self.eps = eps
+        if channels_last:
+            self.spatial_dim = (1, 2)
+            self.channel_dim = -1
+            self.wb_shape = (1, 1, 1, -1)
+        else:
+            self.spatial_dim = (2, 3)
+            self.channel_dim = 1
+            self.wb_shape = (1, -1, 1, 1)
+
+        self.weight = nn.Parameter(torch.zeros(dim))
+        self.bias = nn.Parameter(torch.zeros(dim))
+
+    def forward(self, x):
+        x_g = x.norm(p=2, dim=self.spatial_dim, keepdim=True)
+        x_n = x_g / (x_g.mean(dim=self.channel_dim, keepdim=True) + self.eps)
+        return x + torch.addcmul(self.bias.view(self.wb_shape), self.weight.view(self.wb_shape), x * x_n)

timm/layers/mlp.py

@@ -2,25 +2,38 @@
 
 Hacked together by / Copyright 2020 Ross Wightman
 """
+from functools import partial
+
 from torch import nn as nn
 
+from .grn import GlobalResponseNorm
 from .helpers import to_2tuple
 
 
 class Mlp(nn.Module):
     """ MLP as used in Vision Transformer, MLP-Mixer and related networks
     """
-    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, bias=True, drop=0.):
+    def __init__(
+            self,
+            in_features,
+            hidden_features=None,
+            out_features=None,
+            act_layer=nn.GELU,
+            bias=True,
+            drop=0.,
+            use_conv=False,
+    ):
         super().__init__()
         out_features = out_features or in_features
         hidden_features = hidden_features or in_features
         bias = to_2tuple(bias)
         drop_probs = to_2tuple(drop)
+        linear_layer = partial(nn.Conv2d, kernel_size=1) if use_conv else nn.Linear
 
-        self.fc1 = nn.Linear(in_features, hidden_features, bias=bias[0])
+        self.fc1 = linear_layer(in_features, hidden_features, bias=bias[0])
         self.act = act_layer()
         self.drop1 = nn.Dropout(drop_probs[0])
-        self.fc2 = nn.Linear(hidden_features, out_features, bias=bias[1])
+        self.fc2 = linear_layer(hidden_features, out_features, bias=bias[1])
         self.drop2 = nn.Dropout(drop_probs[1])
 
     def forward(self, x):
@@ -36,18 +49,29 @@ class GluMlp(nn.Module):
     """ MLP w/ GLU style gating
     See: https://arxiv.org/abs/1612.08083, https://arxiv.org/abs/2002.05202
     """
-    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.Sigmoid, bias=True, drop=0.):
+    def __init__(
+            self,
+            in_features,
+            hidden_features=None,
+            out_features=None,
+            act_layer=nn.Sigmoid,
+            bias=True,
+            drop=0.,
+            use_conv=False,
+    ):
         super().__init__()
         out_features = out_features or in_features
         hidden_features = hidden_features or in_features
         assert hidden_features % 2 == 0
         bias = to_2tuple(bias)
         drop_probs = to_2tuple(drop)
+        linear_layer = partial(nn.Conv2d, kernel_size=1) if use_conv else nn.Linear
+        self.chunk_dim = 1 if use_conv else -1
 
-        self.fc1 = nn.Linear(in_features, hidden_features, bias=bias[0])
+        self.fc1 = linear_layer(in_features, hidden_features, bias=bias[0])
         self.act = act_layer()
         self.drop1 = nn.Dropout(drop_probs[0])
-        self.fc2 = nn.Linear(hidden_features // 2, out_features, bias=bias[1])
+        self.fc2 = linear_layer(hidden_features // 2, out_features, bias=bias[1])
         self.drop2 = nn.Dropout(drop_probs[1])
 
     def init_weights(self):
@@ -58,7 +82,7 @@ def init_weights(self):
 
     def forward(self, x):
         x = self.fc1(x)
-        x, gates = x.chunk(2, dim=-1)
+        x, gates = x.chunk(2, dim=self.chunk_dim)
         x = x * self.act(gates)
         x = self.drop1(x)
         x = self.fc2(x)
@@ -70,8 +94,15 @@ class GatedMlp(nn.Module):
     """ MLP as used in gMLP
     """
     def __init__(
-            self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU,
-            gate_layer=None, bias=True, drop=0.):
+            self,
+            in_features,
+            hidden_features=None,
+            out_features=None,
+            act_layer=nn.GELU,
+            gate_layer=None,
+            bias=True,
+            drop=0.,
+    ):
         super().__init__()
         out_features = out_features or in_features
         hidden_features = hidden_features or in_features
@@ -104,8 +135,15 @@ class ConvMlp(nn.Module):
     """ MLP using 1x1 convs that keeps spatial dims
     """
     def __init__(
-            self, in_features, hidden_features=None, out_features=None, act_layer=nn.ReLU,
-            norm_layer=None, bias=True, drop=0.):
+            self,
+            in_features,
+            hidden_features=None,
+            out_features=None,
+            act_layer=nn.ReLU,
+            norm_layer=None,
+            bias=True,
+            drop=0.,
+    ):
         super().__init__()
         out_features = out_features or in_features
         hidden_features = hidden_features or in_features
@@ -124,3 +162,40 @@ def forward(self, x):
         x = self.drop(x)
         x = self.fc2(x)
         return x
+
+
+class GlobalResponseNormMlp(nn.Module):
+    """ MLP w/ Global Response Norm (see grn.py), nn.Linear or 1x1 Conv2d
+    """
+    def __init__(
+            self,
+            in_features,
+            hidden_features=None,
+            out_features=None,
+            act_layer=nn.GELU,
+            bias=True,
+            drop=0.,
+            use_conv=False,
+    ):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        bias = to_2tuple(bias)
+        drop_probs = to_2tuple(drop)
+        linear_layer = partial(nn.Conv2d, kernel_size=1) if use_conv else nn.Linear
+
+        self.fc1 = linear_layer(in_features, hidden_features, bias=bias[0])
+        self.act = act_layer()
+        self.drop1 = nn.Dropout(drop_probs[0])
+        self.grn = GlobalResponseNorm(hidden_features, channels_last=not use_conv)
+        self.fc2 = linear_layer(hidden_features, out_features, bias=bias[1])
+        self.drop2 = nn.Dropout(drop_probs[1])
+
+    def forward(self, x):
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.drop1(x)
+        x = self.grn(x)
+        x = self.fc2(x)
+        x = self.drop2(x)
+        return x