# Copyright (c) 2022 Graphcore Ltd. All rights reserved.
# --------------------------------------------------------
# Swin Transformer
# This file has been modified by Graphcore Ltd.
# Copyright (c) 2021 Microsoft
# Licensed under The MIT License
# The LICENSE referenced above is reproduced below:
# MIT License
#
#     Copyright (c) Microsoft Corporation.
#
#     Permission is hereby granted, free of charge, to any person obtaining a copy
#     of this software and associated documentation files (the "Software"), to deal
#     in the Software without restriction, including without limitation the rights
#     to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
#     copies of the Software, and to permit persons to whom the Software is
#     furnished to do so, subject to the following conditions:
#
#     The above copyright notice and this permission notice shall be included in all
#     copies or substantial portions of the Software.
#
#     THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
#     IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
#     FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
#     AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
#     LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
#     OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
#     SOFTWARE
# Written by Ze Liu
# --------------------------------------------------------

from numpy import dtype
import torch
import torch.nn as nn
import torch.utils.checkpoint as checkpoint
from timm.models.layers import DropPath, to_2tuple, trunc_normal_
import poptorch


def remap_tensor(
    x,
    fwd_grain_size=8,
    bwd_grain_size=0,
    fwd_clone_layout=False,
    bwd_clone_layout=False,
    fwd_after_matmul=False,
    bwd_after_matmul=False,
    debug_str="",
):
    if 0 == bwd_grain_size:
        bwd_grain_size = fwd_grain_size
    return poptorch.custom_op(
        [x],
        "RemapCE",
        "ai.graphcore",
        1,
        example_outputs=[x],
        attributes={
            "fwd_grain_size": fwd_grain_size,
            "bwd_grain_size": bwd_grain_size,
            "fwd_clone_layout": 1 if fwd_clone_layout else 0,
            "bwd_clone_layout": 1 if bwd_clone_layout else 0,
            "fwd_after_matmul": 1 if fwd_after_matmul else 0,
            "bwd_after_matmul": 1 if bwd_after_matmul else 0,
            "debug_str": debug_str,
        },
    )[0]


def padding_aligned(x, dim, remap_grain_size=8, need_remap=True, bwd_after_matmul=False):
    src_shape = x.shape
    rank = len(src_shape)
    padding_size = [0 for i in range(2 * rank)]
    cur_size = int(src_shape[dim])
    padding_idx = 0
    if dim >= 0:
        padding_idx = 2 * (rank - 1 - dim) + 1
    else:
        padding_idx = 2 * (-(dim + 1)) + 1
    if 0 != (cur_size & 7):
        expected_size = ((cur_size + 7) >> 3) << 3
        padding_size[padding_idx] = expected_size - cur_size
        if -1 == dim or (rank - 1) == dim:
            x = torch.nn.functional.pad(x, padding_size, "constant", value=0.0)
            x = remap_tensor(x, expected_size)
        else:
            x = torch.nn.functional.pad(x, padding_size, "constant", value=0.0)
            if need_remap:
                x = remap_tensor(x, remap_grain_size, bwd_after_matmul=bwd_after_matmul)
    else:
        if need_remap:
            x = remap_tensor(x, remap_grain_size)
    return x


class Mlp(nn.Module):
    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.0):
        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


def window_partition(x, window_size, remap):
    """
    Args:
        x: (B, H, W, C)
        window_size (int): window size

    Returns:
        windows: (num_windows*B, window_size, window_size, C)
    """
    B, H, W, C = x.shape
    x = x.view(B, H // window_size, window_size, W // window_size, window_size, C)

    windows = x.permute(0, 1, 3, 2, 4, 5)
    if remap:
        windows = remap_tensor(windows, 16)
    else:
        windows = windows.contiguous()
    windows = windows.reshape(-1, window_size, window_size, C)
    return windows


def window_reverse(windows, window_size, H, W, remap):
    """
    Args:
        windows: (num_windows*B, window_size, window_size, C)
        window_size (int): Window size
        H (int): Height of image
        W (int): Width of image

    Returns:
        x: (B, H, W, C)
    """
    B = int(windows.shape[0] / (H * W / window_size / window_size))
    x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1)
    x = x.permute(0, 1, 3, 2, 4, 5)
    if remap:
        x = remap_tensor(x, 16)
    else:
        x = x.contiguous()
    x = x.reshape(B, H, W, -1)
    return x


class WindowAttention(nn.Module):
    r"""Window based multi-head self attention (W-MSA) module with relative position bias.
    It supports both of shifted and non-shifted window.

    Args:
        dim (int): Number of input channels.
        window_size (tuple[int]): The height and width of the window.
        num_heads (int): Number of attention heads.
        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True
        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set
        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0
        proj_drop (float, optional): Dropout ratio of output. Default: 0.0
    """

    def __init__(
        self,
        dim,
        window_size,
        num_heads,
        qkv_bias=True,
        qk_scale=None,
        attn_drop=0.0,
        proj_drop=0.0,
        device="cpu",
        use_half=True,
    ):

        super().__init__()
        self.model_type = torch.float16 if use_half else torch.float32
        self.device = device
        self.dim = dim
        self.window_size = window_size  # Wh, Ww
        self.num_heads = num_heads
        head_dim = dim // num_heads
        self.scale = qk_scale or head_dim**-0.5

        # define a parameter table of relative position bias
        self.relative_position_bias_table = nn.Parameter(
            torch.zeros((2 * window_size[0] - 1) * (2 * window_size[1] - 1), num_heads)
        )  # 2*Wh-1 * 2*Ww-1, nH

        # get pair-wise relative position index for each token inside the
        # window
        coords_h = torch.arange(self.window_size[0])
        coords_w = torch.arange(self.window_size[1])
        coords = torch.stack(torch.meshgrid([coords_h, coords_w]))  # 2, Wh, Ww
        coords_flatten = torch.flatten(coords, 1)  # 2, Wh*Ww
        relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :]  # 2, Wh*Ww, Wh*Ww
        relative_coords = relative_coords.permute(1, 2, 0).contiguous()  # Wh*Ww, Wh*Ww, 2
        relative_coords[:, :, 0] += self.window_size[0] - 1  # shift to start from 0
        relative_coords[:, :, 1] += self.window_size[1] - 1
        relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1
        relative_position_index = relative_coords.sum(-1)  # Wh*Ww, Wh*Ww
        self.register_buffer("relative_position_index", relative_position_index)

        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)

        if self.window_size[0] == 7:
            qkv_mask = torch.ones((49, dim * 3))
            qkv_v_mask1 = torch.zeros((7, dim * 3))
            self.qkv_mask = torch.cat([qkv_mask, qkv_v_mask1], dim=0).type(self.model_type)

            mask0 = torch.zeros((49, 49))
            h_mask0 = torch.zeros((49, 7))
            v_mask0 = torch.zeros((7, 56))
            h_mask0 = h_mask0.fill_(-65450.0)
            v_mask0 = v_mask0.fill_(0.0)
            mask0 = torch.cat([mask0, h_mask0], dim=-1)
            self.mask0 = torch.cat([mask0, v_mask0], dim=-2).type(self.model_type)
            mask1 = torch.ones((49, 49))
            h_mask1 = torch.zeros((49, 7))
            v_mask1 = torch.zeros((7, 56))
            mask1 = torch.cat([mask1, h_mask1], dim=-1)
            self.mask1 = torch.cat([mask1, v_mask1], dim=-2).type(self.model_type)
        else:
            self.qkv_mask = torch.ones(1).type(self.model_type)
            self.mask0 = torch.zeros(1).type(self.model_type)
            self.mask1 = torch.ones(1).type(self.model_type)

        trunc_normal_(self.relative_position_bias_table, std=0.02)
        self.softmax = nn.Softmax(dim=-1)

    def forward(self, x, mask=None):
        """
        Args:
            x: input features with shape of (num_windows*B, N, C)
            mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None
        """
        B_, N, C = x.shape
        x = padding_aligned(x, -2, 16, bwd_after_matmul=True)
        qkv = self.qkv(x)
        if self.qkv.bias is not None:
            qkv = remap_tensor(qkv, 16, fwd_after_matmul=True)
            self.qkv_mask = self.qkv_mask.to(device=x.device)
            qkv = qkv * self.qkv_mask.to(device=x.device)
            qkv = remap_tensor(qkv, 16, 16, True, False, bwd_after_matmul=True)
        qkv = qkv.reshape(B_, qkv.shape[1], 3, self.num_heads, C // self.num_heads)
        qkv = qkv.permute(2, 0, 3, 1, 4)
        qkv = remap_tensor(qkv, 16, 16, False, True)
        # make torchscript happy (cannot use tensor as tuple)
        q, k, v = qkv[0], qkv[1], qkv[2]
        q = q * torch.tensor(self.scale, dtype=self.model_type)
        q = remap_tensor(q, 16, fwd_clone_layout=True, bwd_after_matmul=True)
        attn = q @ k.transpose(-2, -1)
        attn = remap_tensor(attn, int(attn.shape[-1]), fwd_after_matmul=True)
        self.relative_position_bias_table = self.relative_position_bias_table.to(device=x.device)
        relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
            self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1
        )  # Wh*Ww,Wh*Ww,nH
        relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww
        relative_position_bias = padding_aligned(relative_position_bias, -1)
        relative_position_bias = padding_aligned(relative_position_bias, -2, int(relative_position_bias.shape[-1]))
        attn = attn + relative_position_bias.unsqueeze(0)
        self.mask0 = self.mask0.to(device=x.device)
        if mask is not None:

            mask = padding_aligned(mask, -1)
            mask = padding_aligned(mask, -2, int(mask.shape[-1]))
            nW = mask.shape[0]
            attn = attn.view(B_ // nW, nW, self.num_heads, mask.shape[-2], mask.shape[-1]) + mask.unsqueeze(
                1
            ).unsqueeze(0)
            attn = attn.view(-1, self.num_heads, mask.shape[-2], mask.shape[-1])
            attn = attn + self.mask0.to(device=x.device)
            attn = self.softmax(attn)
        else:

            attn = attn + self.mask0.to(device=x.device)
            attn = self.softmax(attn)
        attn = attn * self.mask1.to(device=x.device)
        attn = remap_tensor(attn, int(attn.shape[-1]), int(attn.shape[-1]), True, False, bwd_after_matmul=True)
        attn = self.attn_drop(attn)
        x = attn @ v
        x = x.transpose(1, 2).reshape(B_, -1, C)
        x = self.proj(x)
        x = remap_tensor(x, 16, fwd_after_matmul=True)
        x = x[:, :N, :]
        x = self.proj_drop(x)
        return x

    def extra_repr(self) -> str:
        return f"dim={self.dim}, window_size={self.window_size}, num_heads={self.num_heads}"

    def flops(self, N):
        # calculate flops for 1 window with token length of N
        flops = 0
        # qkv = self.qkv(x)
        flops += N * self.dim * 3 * self.dim
        # attn = (q @ k.transpose(-2, -1))
        flops += self.num_heads * N * (self.dim // self.num_heads) * N
        #  x = (attn @ v)
        flops += self.num_heads * N * N * (self.dim // self.num_heads)
        # x = self.proj(x)
        flops += N * self.dim * self.dim
        return flops


class SwinTransformerBlock(nn.Module):
    r"""Swin Transformer Block.

    Args:
        dim (int): Number of input channels.
        input_resolution (tuple[int]): Input resolution.
        num_heads (int): Number of attention heads.
        window_size (int): Window size.
        shift_size (int): Shift size for SW-MSA.
        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
        drop (float, optional): Dropout rate. Default: 0.0
        attn_drop (float, optional): Attention dropout rate. Default: 0.0
        drop_path (float, optional): Stochastic depth rate. Default: 0.0
        act_layer (nn.Module, optional): Activation layer. Default: nn.GELU
        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
    """

    def __init__(
        self,
        dim,
        input_resolution,
        num_heads,
        window_size=7,
        shift_size=0,
        mlp_ratio=4.0,
        qkv_bias=True,
        qk_scale=None,
        drop=0.0,
        attn_drop=0.0,
        drop_path=0.0,
        act_layer=nn.GELU,
        norm_layer=nn.LayerNorm,
        norm_before_mlp="ln",
        device="gpu",
        use_half=True,
    ):
        super().__init__()
        self.device = device
        self.dim = dim
        self.input_resolution = input_resolution
        self.num_heads = num_heads
        self.window_size = window_size
        self.shift_size = shift_size
        self.mlp_ratio = mlp_ratio
        self.norm_before_mlp = norm_before_mlp
        if min(self.input_resolution) <= self.window_size:
            # if window size is larger than input resolution, we don't
            # partition windows
            self.shift_size = 0
            self.window_size = min(self.input_resolution)
        assert 0 <= self.shift_size < self.window_size, "shift_size must in 0-window_size"

        self.norm1 = norm_layer(dim)
        self.attn = WindowAttention(
            dim,
            window_size=to_2tuple(self.window_size),
            num_heads=num_heads,
            qkv_bias=qkv_bias,
            qk_scale=qk_scale,
            attn_drop=attn_drop,
            proj_drop=drop,
            device=device,
            use_half=use_half,
        )

        self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
        if self.norm_before_mlp == "ln":
            self.norm2 = nn.LayerNorm(dim)
        elif self.norm_before_mlp == "bn":
            self.norm2 = lambda x: nn.BatchNorm1d(dim)(x.transpose(1, 2)).transpose(1, 2)
        else:
            raise NotImplementedError
        mlp_hidden_dim = int(dim * mlp_ratio)
        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)

        if self.shift_size > 0:
            # calculate attention mask for SW-MSA
            H, W = self.input_resolution
            img_mask = torch.zeros((1, H, W, 1))  # 1 H W 1
            h_slices = (
                slice(0, -self.window_size),
                slice(-self.window_size, -self.shift_size),
                slice(-self.shift_size, None),
            )
            w_slices = (
                slice(0, -self.window_size),
                slice(-self.window_size, -self.shift_size),
                slice(-self.shift_size, None),
            )
            cnt = 0
            for h in h_slices:
                for w in w_slices:
                    img_mask[:, h, w, :] = cnt
                    cnt += 1

            mask_windows = window_partition(img_mask, self.window_size, False)  # nW, window_size, window_size, 1
            mask_windows = mask_windows.view(-1, self.window_size * self.window_size)
            attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
            attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))
        else:
            attn_mask = None

        self.register_buffer("attn_mask", attn_mask)

    def forward(self, x):
        H, W = self.input_resolution
        B, L, C = x.shape
        assert L == H * W, "input feature has wrong size"

        shortcut = x
        x = self.norm1(x)
        x = x.view(B, H, W, C)

        # cyclic shift
        if self.shift_size > 0:
            # workaround for roll
            if self.device == "ipu":
                tmp1 = x[:, self.shift_size :, :, :]
                tmp2 = x[:, : self.shift_size, :, :]
                tmp = torch.cat([tmp1, tmp2], dim=1)
                tmp1 = tmp[:, :, self.shift_size :, :]
                tmp2 = tmp[:, :, : self.shift_size, :]
                shifted_x = torch.cat([tmp1, tmp2], dim=2)
                shifted_x = remap_tensor(shifted_x, 16)
            elif self.device == "cpu" or self.device == "gpu":
                shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
        else:
            shifted_x = x

        # partition windows
        # nW*B, window_size, window_size, C
        x_windows = window_partition(shifted_x, self.window_size, True)
        x_windows = x_windows.view(-1, self.window_size * self.window_size, C)  # nW*B, window_size*window_size, C

        # W-MSA/SW-MSA
        # nW*B, window_size*window_size, C
        attn_windows = self.attn(x_windows, mask=self.attn_mask)

        # merge windows
        attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C)
        shifted_x = window_reverse(attn_windows, self.window_size, H, W, True)  # B H' W' C

        # reverse cyclic shift
        if self.shift_size > 0:
            # workaround for roll
            if self.device == "ipu":
                index = shifted_x.size()[1] - self.shift_size
                tmp1 = shifted_x[:, index:, :, :]
                tmp2 = shifted_x[:, :index, :, :]
                tmp = torch.cat([tmp1, tmp2], dim=1)
                tmp1 = tmp[:, :, index:, :]
                tmp2 = tmp[:, :, :index, :]
                x = torch.cat([tmp1, tmp2], dim=2)
            elif self.device == "cpu" or self.device == "gpu":
                x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
        else:
            x = shifted_x

        x = remap_tensor(x, 16)
        x = x.view(B, H * W, C)

        # FFN
        x = shortcut + self.drop_path(x)
        x_mlp = self.norm2(x)
        x_mlp = remap_tensor(x_mlp, 16, 16, True, False, bwd_after_matmul=True)
        x_mlp = self.mlp(x_mlp)
        x_mlp = remap_tensor(x_mlp, 16, fwd_after_matmul=True)
        x = x + self.drop_path(x_mlp)
        return x

    def extra_repr(self) -> str:
        return (
            f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, "
            f"window_size={self.window_size}, shift_size={self.shift_size}, mlp_ratio={self.mlp_ratio}"
        )

    def flops(self):
        flops = 0
        H, W = self.input_resolution
        # norm1
        flops += self.dim * H * W
        # W-MSA/SW-MSA
        nW = H * W / self.window_size / self.window_size
        flops += nW * self.attn.flops(self.window_size * self.window_size)
        # mlp
        flops += 2 * H * W * self.dim * self.dim * self.mlp_ratio
        # norm2
        flops += self.dim * H * W
        return flops


class PatchMerging(nn.Module):
    r"""Patch Merging Layer.

    Args:
        input_resolution (tuple[int]): Resolution of input feature.
        dim (int): Number of input channels.
        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
    """

    def __init__(self, input_resolution, dim, norm_layer=nn.LayerNorm):
        super().__init__()
        self.input_resolution = input_resolution
        self.dim = dim
        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
        self.norm = norm_layer(4 * dim)

    def forward(self, x):
        """
        x: B, H*W, C
        """
        H, W = self.input_resolution
        B, L, C = x.shape
        assert L == H * W, "input feature has wrong size"
        assert H % 2 == 0 and W % 2 == 0, f"x size ({H}*{W}) are not even."

        x = x.view(B, H, W, C)
        x = x.view(B, H // 2, 2, W // 2, 2, C).transpose(2, 3).transpose(3, 4)
        x = x.reshape(B, -1, 4 * C)
        x = remap_tensor(x, 16)

        x = self.norm(x)
        x = self.reduction(x)

        return x

    def extra_repr(self) -> str:
        return f"input_resolution={self.input_resolution}, dim={self.dim}"

    def flops(self):
        H, W = self.input_resolution
        flops = H * W * self.dim
        flops += (H // 2) * (W // 2) * 4 * self.dim * 2 * self.dim
        return flops


class BasicLayer(nn.Module):
    """A basic Swin Transformer layer for one stage.

    Args:
        dim (int): Number of input channels.
        input_resolution (tuple[int]): Input resolution.
        depth (int): Number of blocks.
        num_heads (int): Number of attention heads.
        window_size (int): Local window size.
        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
        drop (float, optional): Dropout rate. Default: 0.0
        attn_drop (float, optional): Attention dropout rate. Default: 0.0
        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.
    """

    def __init__(
        self,
        dim,
        input_resolution,
        depth,
        num_heads,
        window_size,
        mlp_ratio=4.0,
        qkv_bias=True,
        qk_scale=None,
        drop=0.0,
        attn_drop=0.0,
        drop_path=0.0,
        norm_layer=nn.LayerNorm,
        downsample=None,
        use_checkpoint=False,
        norm_before_mlp="ln",
        device="gpu",
        use_half=True,
    ):

        super().__init__()
        self.dim = dim
        self.input_resolution = input_resolution
        self.depth = depth
        self.use_checkpoint = use_checkpoint

        # build blocks
        self.blocks = nn.ModuleList(
            [
                SwinTransformerBlock(
                    dim=dim,
                    input_resolution=input_resolution,
                    num_heads=num_heads,
                    window_size=window_size,
                    shift_size=0 if (i % 2 == 0) else window_size // 2,
                    mlp_ratio=mlp_ratio,
                    qkv_bias=qkv_bias,
                    qk_scale=qk_scale,
                    drop=drop,
                    attn_drop=attn_drop,
                    drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,
                    norm_layer=norm_layer,
                    norm_before_mlp=norm_before_mlp,
                    device=device,
                    use_half=use_half,
                )
                for i in range(depth)
            ]
        )

        # patch merging layer
        if downsample is not None:
            self.downsample = downsample(input_resolution, dim=dim, norm_layer=norm_layer)
        else:
            self.downsample = None

    def forward(self, x):
        for blk in self.blocks:
            x = remap_tensor(x, 16)
            if self.use_checkpoint:
                x = checkpoint.checkpoint(blk, x)
            else:
                x = blk(x)
        if self.downsample is not None:
            x = self.downsample(x)
        return x

    def extra_repr(self) -> str:
        return f"dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}"

    def flops(self):
        flops = 0
        for blk in self.blocks:
            flops += blk.flops()
        if self.downsample is not None:
            flops += self.downsample.flops()
        return flops


class PatchEmbed(nn.Module):
    r"""Image to Patch Embedding

    Args:
        img_size (int): Image size.  Default: 224.
        patch_size (int): Patch token size. Default: 4.
        in_chans (int): Number of input image channels. Default: 3.
        embed_dim (int): Number of linear projection output channels. Default: 96.
        norm_layer (nn.Module, optional): Normalization layer. Default: None
    """

    def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None):
        super().__init__()
        img_size = to_2tuple(img_size)
        patch_size = to_2tuple(patch_size)
        patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]]
        self.img_size = img_size
        self.patch_size = patch_size
        self.patches_resolution = patches_resolution
        self.num_patches = patches_resolution[0] * patches_resolution[1]

        self.in_chans = in_chans
        self.embed_dim = embed_dim

        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)
        if norm_layer is not None:
            self.norm = norm_layer(embed_dim)
        else:
            self.norm = None

    def forward(self, x):
        B, C, H, W = x.shape
        # FIXME look at relaxing size constraints
        assert (
            H == self.img_size[0] and W == self.img_size[1]
        ), f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})."
        x = self.proj(x)
        x = x.flatten(2).transpose(1, 2)  # B Ph*Pw C
        x = remap_tensor(x, 16)
        if self.norm is not None:
            x = self.norm(x)
        return x

    def flops(self):
        Ho, Wo = self.patches_resolution
        flops = Ho * Wo * self.embed_dim * self.in_chans * (self.patch_size[0] * self.patch_size[1])
        if self.norm is not None:
            flops += Ho * Wo * self.embed_dim
        return flops


class SwinTransformer(nn.Module):
    r"""Swin Transformer
        A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows`  -
          https://arxiv.org/pdf/2103.14030

    Args:
        img_size (int | tuple(int)): Input image size. Default 224
        patch_size (int | tuple(int)): Patch size. Default: 4
        in_chans (int): Number of input image channels. Default: 3
        num_classes (int): Number of classes for classification head. Default: 1000
        embed_dim (int): Patch embedding dimension. Default: 96
        depths (tuple(int)): Depth of each Swin Transformer layer.
        num_heads (tuple(int)): Number of attention heads in different layers.
        window_size (int): Window size. Default: 7
        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4
        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: True
        qk_scale (float): Override default qk scale of head_dim ** -0.5 if set. Default: None
        drop_rate (float): Dropout rate. Default: 0
        attn_drop_rate (float): Attention dropout rate. Default: 0
        drop_path_rate (float): Stochastic depth rate. Default: 0.1
        norm_layer (nn.Module): Normalization layer. Default: nn.LayerNorm.
        ape (bool): If True, add absolute position embedding to the patch embedding. Default: False
        patch_norm (bool): If True, add normalization after patch embedding. Default: True
        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False
    """

    def __init__(
        self,
        img_size=224,
        patch_size=4,
        in_chans=3,
        num_classes=1000,
        embed_dim=96,
        depths=[2, 2, 6, 2],
        num_heads=[3, 6, 12, 24],
        window_size=7,
        mlp_ratio=4.0,
        qkv_bias=True,
        qk_scale=None,
        drop_rate=0.0,
        attn_drop_rate=0.0,
        drop_path_rate=0.1,
        norm_layer=nn.LayerNorm,
        ape=False,
        patch_norm=True,
        use_checkpoint=False,
        norm_before_mlp="ln",
        device="gpu",
        train_loss_fn=None,
        use_half=True,
        **kwargs,
    ):
        super().__init__()

        self.num_classes = num_classes
        self.num_layers = len(depths)
        self.embed_dim = embed_dim
        self.ape = ape
        self.patch_norm = patch_norm
        self.num_features = int(embed_dim * 2 ** (self.num_layers - 1))
        self.mlp_ratio = mlp_ratio

        # split image into non-overlapping patches
        self.patch_embed = PatchEmbed(
            img_size=img_size,
            patch_size=patch_size,
            in_chans=in_chans,
            embed_dim=embed_dim,
            norm_layer=norm_layer if self.patch_norm else None,
        )
        num_patches = self.patch_embed.num_patches
        patches_resolution = self.patch_embed.patches_resolution
        self.patches_resolution = patches_resolution

        # absolute position embedding
        if self.ape:
            self.absolute_pos_embed = nn.Parameter(torch.zeros(1, num_patches, embed_dim))
            trunc_normal_(self.absolute_pos_embed, std=0.02)

        self.pos_drop = nn.Dropout(p=drop_rate)

        # stochastic depth
        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))]  # stochastic depth decay rule

        # build layers
        self.layers = nn.ModuleList()
        for i_layer in range(self.num_layers):
            layer = BasicLayer(
                dim=int(embed_dim * 2**i_layer),
                input_resolution=(patches_resolution[0] // (2**i_layer), patches_resolution[1] // (2**i_layer)),
                depth=depths[i_layer],
                num_heads=num_heads[i_layer],
                window_size=window_size,
                mlp_ratio=self.mlp_ratio,
                qkv_bias=qkv_bias,
                qk_scale=qk_scale,
                drop=drop_rate,
                attn_drop=attn_drop_rate,
                drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])],
                norm_layer=norm_layer,
                downsample=PatchMerging if (i_layer < self.num_layers - 1) else None,
                use_checkpoint=use_checkpoint,
                norm_before_mlp=norm_before_mlp,
                device=device,
                use_half=use_half,
            )
            self.layers.append(layer)

        self.norm = norm_layer(self.num_features)
        self.avgpool = nn.AdaptiveAvgPool1d(1)
        self.head = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()

        self.apply(self._init_weights)
        if num_classes > 0:
            if train_loss_fn:
                self.celoss = train_loss_fn
            else:
                self.celoss = nn.CrossEntropyLoss()

    def _init_weights(self, m):
        if isinstance(m, nn.Linear):
            trunc_normal_(m.weight, std=0.02)
            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)

    @torch.jit.ignore
    def no_weight_decay(self):
        return {"absolute_pos_embed"}

    @torch.jit.ignore
    def no_weight_decay_keywords(self):
        return {"relative_position_bias_table"}

    def forward_features(self, x):
        x = self.patch_embed(x)
        if self.ape:
            x = x + self.absolute_pos_embed
        x = self.pos_drop(x)

        for layer in self.layers:
            x = layer(x)

        x = self.norm(x)  # B L C
        x = self.avgpool(x.transpose(1, 2))  # B C 1
        x = torch.flatten(x, 1)
        return x

    def forward(self, x, target=None):
        x = self.forward_features(x)
        x = self.head(x)

        if target is None:
            return x
        else:
            return x, poptorch.identity_loss(self.celoss(x.type(torch.float32), target.type(torch.float32)), "none")

    def flops(self):
        flops = 0
        flops += self.patch_embed.flops()
        for i, layer in enumerate(self.layers):
            flops += layer.flops()
        flops += self.num_features * self.patches_resolution[0] * self.patches_resolution[1] // (2**self.num_layers)
        flops += self.num_features * self.num_classes
        return flops