Upload 8 files

config.json

@@ -0,0 +1,29 @@
+{
+  "architectures": [
+    "SatDINOModel"
+  ],
+  "attn_drop_rate": 0.0,
+  "auto_map": {
+    "AutoConfig": "configuration_satdino.SatDINOConfig",
+    "AutoModel": "modeling_satdino.SatDINOModel"
+  },
+  "depth": 12,
+  "drop_path_rate": 0.0,
+  "drop_rate": 0.0,
+  "embed_dim": 384,
+  "img_size": [
+    224
+  ],
+  "in_chans": 3,
+  "mlp_ratio": 4,
+  "model_type": "satdino",
+  "norm_layer": 1e-06,
+  "num_classes": 0,
+  "num_heads": 6,
+  "patch_size": 16,
+  "pos_encoding_method": "learnable",
+  "qk_scale": null,
+  "qkv_bias": true,
+  "transformers_version": "4.51.2",
+  "use_xformers": false
+}

configuration_satdino.py

@@ -0,0 +1,44 @@
+from transformers import PretrainedConfig
+
+
+class SatDINOConfig(PretrainedConfig):
+    model_type = "satdino"
+
+    def __init__(
+            self,
+            img_size=[224],
+            patch_size=16,
+            in_chans=3,
+            num_classes=0,
+            embed_dim=768,
+            depth=12,
+            num_heads=12,
+            mlp_ratio=4.,
+            qkv_bias=False,
+            qk_scale=None,
+            drop_rate=0.,
+            attn_drop_rate=0.,
+            drop_path_rate=0.,
+            norm_layer=1e-6,
+            use_xformers=False,
+            pos_encoding_method="learnable",
+            **kwargs
+    ):
+        self.img_size = img_size
+        self.patch_size = patch_size
+        self.in_chans = in_chans
+        self.num_classes = num_classes
+        self.embed_dim = embed_dim
+        self.depth = depth
+        self.num_heads = num_heads
+        self.mlp_ratio = mlp_ratio
+        self.qkv_bias = qkv_bias
+        self.qk_scale = qk_scale
+        self.drop_rate = drop_rate
+        self.attn_drop_rate = attn_drop_rate
+        self.drop_path_rate = drop_path_rate
+        self.norm_layer = norm_layer
+        self.use_xformers = use_xformers
+        self.pos_encoding_method = pos_encoding_method
+        super().__init__(**kwargs)
+

modeling_satdino.py

@@ -0,0 +1,307 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+# 
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+# 
+#     http://www.apache.org/licenses/LICENSE-2.0
+# 
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+Mostly copy-paste from timm library.
+https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
+"""
+import os
+import math
+from functools import partial
+
+import torch
+import torch.nn as nn
+
+from transformers import PreTrainedModel
+from .utils import trunc_normal_, get_1d_sincos_pos_embed
+from .configuration_satdino import SatDINOConfig
+
+try:    
+    from xformers.helpers.timm_sparse_attention import TimmSparseAttention
+except:
+    TimmSparseAttention = None
+
+
+def drop_path(x, drop_prob: float = 0., training: bool = False):
+    if drop_prob == 0. or not training:
+        return x
+    keep_prob = 1 - drop_prob
+    shape = (x.shape[0],) + (1,) * (x.ndim - 1)  # work with diff dim tensors, not just 2D ConvNets
+    random_tensor = keep_prob + torch.rand(shape, dtype=x.dtype, device=x.device)
+    random_tensor.floor_()  # binarize
+    output = x.div(keep_prob) * random_tensor
+    return output
+
+
+class DropPath(nn.Module):
+    """Drop paths (Stochastic Depth) per sample  (when applied in main path of residual blocks).
+    """
+
+    def __init__(self, drop_prob=None):
+        super(DropPath, self).__init__()
+        self.drop_prob = drop_prob
+
+    def forward(self, x):
+        return drop_path(x, self.drop_prob, self.training)
+
+
+class Mlp(nn.Module):
+    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.fc1 = nn.Linear(in_features, hidden_features)
+        self.act = act_layer()
+        self.fc2 = nn.Linear(hidden_features, out_features)
+        self.drop = nn.Dropout(drop)
+
+    def forward(self, x):
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.drop(x)
+        x = self.fc2(x)
+        x = self.drop(x)
+        return x
+
+
+class Attention(nn.Module):
+    def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0.):
+        super().__init__()
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+        self.scale = qk_scale or head_dim ** -0.5
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.attn_drop = nn.Dropout(attn_drop)
+        self.proj = nn.Linear(dim, dim)
+        self.proj_drop = nn.Dropout(proj_drop)
+
+    def forward(self, x):
+        B, N, C = x.shape
+        qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
+        q, k, v = qkv[0], qkv[1], qkv[2]
+
+        attn = (q @ k.transpose(-2, -1)) * self.scale
+        attn = attn.softmax(dim=-1)
+        attn = self.attn_drop(attn)
+
+        x = (attn @ v).transpose(1, 2).reshape(B, N, C)
+        x = self.proj(x)
+        x = self.proj_drop(x)
+        return x, attn
+
+
+class Block(nn.Module):
+    def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0.,
+                 drop_path=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm, use_xformers=False):
+        super().__init__()
+        self.norm1 = norm_layer(dim)
+
+        if TimmSparseAttention is not None and use_xformers:
+            # print("Using xFormers attention.")
+            self.attn = TimmSparseAttention(dim, num_heads=num_heads, qkv_bias=qkv_bias, attn_drop=attn_drop,
+                                            proj_drop=drop)
+        else:
+            # print("Using timm attention.")
+            self.attn = Attention(dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop,
+                                  proj_drop=drop)
+
+        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
+        self.norm2 = norm_layer(dim)
+        mlp_hidden_dim = int(dim * mlp_ratio)
+        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)
+
+    def forward(self, x, return_attention=False):
+        attn_res = self.attn(self.norm1(x))
+        if not isinstance(attn_res, tuple):
+            attn_res = (attn_res, None)
+        y, attn = attn_res
+
+        if return_attention:
+            return attn
+        x = x + self.drop_path(y)
+        x = x + self.drop_path(self.mlp(self.norm2(x)))
+        return x
+
+
+class PatchEmbed(nn.Module):
+    """ Image to Patch Embedding
+    """
+
+    def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768):
+        super().__init__()
+        num_patches = (img_size // patch_size) * (img_size // patch_size)
+        self.img_size = img_size
+        self.patch_size = patch_size
+        self.num_patches = num_patches
+
+        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)
+
+    def forward(self, x):
+        B, C, H, W = x.shape
+        x = self.proj(x).flatten(2).transpose(1, 2)
+        return x
+
+
+class SatDINOModel(PreTrainedModel):
+    """ Vision Transformer """
+    config_class = SatDINOConfig
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.num_features = self.embed_dim = config.embed_dim
+        self.pos_encoding_method = config.pos_encoding_method
+
+        self.patch_embed = PatchEmbed(
+            img_size=config.img_size[0],
+            patch_size=config.patch_size,
+            in_chans=config.in_chans,
+            embed_dim=config.embed_dim
+        )
+        num_patches = self.patch_embed.num_patches
+        self.num_patches = num_patches
+
+        # cls token
+        self.cls_token = nn.Parameter(torch.zeros(1, 1, config.embed_dim))
+        trunc_normal_(self.cls_token, std=.02)
+        self.gsd_register = nn.Parameter(torch.zeros(1, 1, config.embed_dim))
+        trunc_normal_(self.gsd_register, std=.02)
+
+        # positional encoding
+        if config.pos_encoding_method == "learnable":
+            self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + 2, config.embed_dim))
+            trunc_normal_(self.pos_embed, std=.02)
+        elif config.pos_encoding_method == "sin_cos":
+            positions = torch.arange(num_patches + 2)
+            self.pos_embed = get_1d_sincos_pos_embed(config.embed_dim, positions).unsqueeze(0).cuda()
+
+        # define blocks
+        norm_layer = partial(nn.LayerNorm, eps=config.norm_layer)
+        dpr = [x.item() for x in torch.linspace(0, config.drop_path_rate, config.depth)]  # stochastic depth decay rule
+        block_kwargs = {
+            "dim": config.embed_dim,
+            "num_heads": config.num_heads,
+            "mlp_ratio": config.mlp_ratio,
+            "qkv_bias": config.qkv_bias,
+            "qk_scale": config.qk_scale,
+            "drop": config.drop_rate,
+            "attn_drop": config.attn_drop_rate,
+            "norm_layer": norm_layer,
+            "use_xformers": config.use_xformers
+        }
+        self.blocks = nn.ModuleList([Block(drop_path=dpr[i], **block_kwargs) for i in range(config.depth)])
+
+        self.pos_drop = nn.Dropout(p=config.drop_rate)
+        self.norm = norm_layer(config.embed_dim)
+
+        # Classifier head
+        self.head = nn.Linear(config.embed_dim, config.num_classes) if config.num_classes > 0 else None
+
+        self.apply(self._init_weights)
+
+    def _init_weights(self, m):
+        if isinstance(m, nn.Linear):
+            trunc_normal_(m.weight, std=.02)
+            if isinstance(m, nn.Linear) and m.bias is not None:
+                nn.init.constant_(m.bias, 0)
+        elif isinstance(m, nn.LayerNorm):
+            nn.init.constant_(m.bias, 0)
+            nn.init.constant_(m.weight, 1.0)
+
+    def interpolate_pos_encoding(self, x, w, h):
+        npatch = x.shape[1] - 1
+        N = self.pos_embed.shape[1] - 1
+        if npatch == N and w == h:
+            return self.pos_embed
+        class_pos_embed = self.pos_embed[:, 0]
+        patch_pos_embed = self.pos_embed[:, 1:-1]
+        register_pos_embed = self.pos_embed[:, -1]
+        dim = x.shape[-1]
+        w0 = w // self.patch_embed.patch_size
+        h0 = h // self.patch_embed.patch_size
+        # we add a small number to avoid floating point error in the interpolation
+        # see discussion at https://github.com/facebookresearch/dino/issues/8
+        w0, h0 = w0 + 0.1, h0 + 0.1
+        patch_pos_embed = nn.functional.interpolate(
+            patch_pos_embed.reshape(1, int(math.sqrt(N)), int(math.sqrt(N)), dim).permute(0, 3, 1, 2),
+            scale_factor=(w0 / math.sqrt(N), h0 / math.sqrt(N)),
+            mode='bicubic',
+        )
+        assert int(w0) == patch_pos_embed.shape[-2] and int(h0) == patch_pos_embed.shape[-1]
+        patch_pos_embed = patch_pos_embed.permute(0, 2, 3, 1).view(1, -1, dim)
+        return torch.cat((class_pos_embed.unsqueeze(0), patch_pos_embed, register_pos_embed.unsqueeze(0)), dim=1)
+
+    def prepare_tokens(self, x):
+        B, nc, w, h = x.shape
+        x = self.patch_embed(x)  # patch linear embedding
+
+        # add the [CLS] token to the embed patch tokens
+        cls_tokens = self.cls_token.expand(B, -1, -1)
+        gsd_register = self.gsd_register.expand(B, -1, -1)
+        x = torch.cat((cls_tokens, x, gsd_register), dim=1)
+
+        # add positional encoding to each token
+        x = x + self.interpolate_pos_encoding(x, w, h)
+
+        return self.pos_drop(x)
+
+    def forward(self, x, return_all=False, return_registers=False):
+        x = self.prepare_tokens(x)
+        for blk in self.blocks:
+            x = blk(x)
+        x = self.norm(x)
+
+        if return_all:
+            return x
+
+        if return_registers:
+            return x[:, 0], x[:, -1]
+
+        return x[:, 0]
+    
+    def forward_intermediate_layers(self, x, return_all=False):
+        output = []
+        x = self.prepare_tokens(x)
+        for blk in self.blocks:
+            x = blk(x)
+            if return_all:
+                output.append(self.norm(x[:, :-1]))
+            else:
+                output.append(x[:, 0])
+
+        return output
+        
+    def get_last_selfattention(self, x):
+        x = self.prepare_tokens(x)
+        for i, blk in enumerate(self.blocks):
+            if i < len(self.blocks) - 1:
+                x = blk(x)
+            else:
+                # return attention of the last block
+                return blk(x, return_attention=True)
+
+    def get_intermediate_layers(self, x, n=1):
+        x = self.prepare_tokens(x)
+        # we return the output tokens from the `n` last blocks
+        output = []
+        for i, blk in enumerate(self.blocks):
+            x = blk(x)
+            if len(self.blocks) - i <= n:
+                output.append(self.norm(x))
+        return output
+
+
+
+
+
+

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec2602f5658d02b2efad23640ccacd0929f1e03247762aa0842bdd0e99f9499f
+size 86713830

satdino-vit_small-16-finetune.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3403cd5799df59819910d6cc25ac0a956411238fe319091c0238a9a8db9ad022
+size 86822014

satdino-vit_small-16.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57421cac575f5b6359ad2c294e27835f7272184c223f2215c3caa5fe35079a4c
+size 86813497

utils.py

@@ -0,0 +1,85 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+# 
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+# 
+#     http://www.apache.org/licenses/LICENSE-2.0
+# 
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+Misc functions.
+
+Mostly copy-paste from torchvision references or other public repos like DETR:
+https://github.com/facebookresearch/detr/blob/master/util/misc.py
+"""
+import torch
+import math
+import warnings
+
+
+def get_1d_sincos_pos_embed(embed_dim, pos, gsd=1, ref_gsd=1):
+    """
+    embed_dim: output dimension for each position
+    pos: a list of positions to be encoded: size (M,)
+    out: (M, D)
+    """
+    assert embed_dim % 2 == 0
+    omega = torch.arange(embed_dim // 2, dtype=torch.float, device=pos.device)
+    omega /= embed_dim / 2.
+    omega = 1. / 10000**omega  # (D/2,)
+
+    pos = pos.reshape(-1)  # (M,)
+    out = torch.einsum('m,d->md', pos, omega)  # (M, D/2), outer product
+    
+    emb_sin = torch.sin(gsd/ref_gsd * out) # (M, D/2)
+    emb_cos = torch.cos(gsd/ref_gsd * out) # (M, D/2)
+
+    emb = torch.zeros([len(pos), embed_dim]) # (M, D)
+    emb[:, 0::2] = emb_sin
+    emb[:, 1::2] = emb_cos
+    
+    return emb.float()
+
+def _no_grad_trunc_normal_(tensor, mean, std, a, b):
+    # Cut & paste from PyTorch official master until it's in a few official releases - RW
+    # Method based on https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf
+    def norm_cdf(x):
+        # Computes standard normal cumulative distribution function
+        return (1. + math.erf(x / math.sqrt(2.))) / 2.
+
+    if (mean < a - 2 * std) or (mean > b + 2 * std):
+        warnings.warn("mean is more than 2 std from [a, b] in nn.init.trunc_normal_. "
+                      "The distribution of values may be incorrect.",
+                      stacklevel=2)
+
+    with torch.no_grad():
+        # Values are generated by using a truncated uniform distribution and
+        # then using the inverse CDF for the normal distribution.
+        # Get upper and lower cdf values
+        l = norm_cdf((a - mean) / std)
+        u = norm_cdf((b - mean) / std)
+
+        # Uniformly fill tensor with values from [l, u], then translate to
+        # [2l-1, 2u-1].
+        tensor.uniform_(2 * l - 1, 2 * u - 1)
+
+        # Use inverse cdf transform for normal distribution to get truncated
+        # standard normal
+        tensor.erfinv_()
+
+        # Transform to proper mean, std
+        tensor.mul_(std * math.sqrt(2.))
+        tensor.add_(mean)
+
+        # Clamp to ensure it's in the proper range
+        tensor.clamp_(min=a, max=b)
+        return tensor
+
+def trunc_normal_(tensor, mean=0., std=1., a=-2., b=2.):
+    # type: (Tensor, float, float, float, float) -> Tensor
+    return _no_grad_trunc_normal_(tensor, mean, std, a, b)