modeling_finetune.py

# --------------------------------------------------------
# EVA-02: A Visual Representation for Neon Genesis
# Github source: https://github.com/baaivision/EVA/EVA02
# Copyright (c) 2023 Beijing Academy of Artificial Intelligence (BAAI)
# Licensed under The MIT License [see LICENSE for details]
# By Yuxin Fang
#
# Based on EVA: Exploring the Limits of Masked Visual Representation Learning at Scale (https://arxiv.org/abs/2211.07636)
# https://github.com/baaivision/EVA/tree/master/EVA-01
# --------------------------------------------------------'


import math
from functools import partial

import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils.checkpoint as checkpoint
from torch.nn.parameter import Parameter
from timm.models.layers import drop_path, to_2tuple, trunc_normal_
from timm.models.registry import register_model

from torch import Tensor, Size
from typing import Union, List
import numbers

import xformers.ops as xops
from apex.normalization import FusedLayerNorm

from rope import *



def _cfg(url='', **kwargs):
    return {
        'url': url,
        'num_classes': 1000, 'input_size': (3, 224, 224), 'pool_size': None,
        'crop_pct': .9, 'interpolation': 'bicubic',
        'mean': (0.5, 0.5, 0.5), 'std': (0.5, 0.5, 0.5),
        **kwargs
    }


_shape_t = Union[int, List[int], Size]



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)
    
    def extra_repr(self) -> str:
        return 'p={}'.format(self.drop_prob)


class Mlp(nn.Module):
    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0., 
                norm_layer=nn.LayerNorm, 
                subln=False
            ):
        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.ffn_ln = norm_layer(hidden_features) if subln else nn.Identity()

        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.ffn_ln(x)

        x = self.fc2(x)
        x = self.drop(x)
        return x


class SwiGLU(nn.Module):
    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.SiLU, drop=0., 
                norm_layer=nn.LayerNorm, subln=False
            ):
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features

        self.w1 = nn.Linear(in_features, hidden_features)
        self.w2 = nn.Linear(in_features, hidden_features)

        self.act = act_layer()
        self.ffn_ln = norm_layer(hidden_features) if subln else nn.Identity()
        self.w3 = nn.Linear(hidden_features, out_features)
        
        self.drop = nn.Dropout(drop)

    def forward(self, x):
        x1 = self.w1(x)
        x2 = self.w2(x)
        hidden = self.act(x1) * x2
        x = self.ffn_ln(hidden)
        x = self.w3(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., window_size=None, attn_head_dim=None, use_decoupled_rel_pos_bias=False,
            deepnorm=False,
            subln=False,
            norm_layer=nn.LayerNorm,
            xattn=False,
            rope=None,
        ):
        super().__init__()
        self.num_heads = num_heads
        head_dim = dim // num_heads
        if attn_head_dim is not None:
            head_dim = attn_head_dim
        all_head_dim = head_dim * self.num_heads
        self.scale = qk_scale or head_dim ** -0.5

        self.deepnorm = deepnorm
        self.subln = subln
        if self.deepnorm or self.subln:
            self.q_proj = nn.Linear(dim, all_head_dim, bias=False)
            self.k_proj = nn.Linear(dim, all_head_dim, bias=False)
            self.v_proj = nn.Linear(dim, all_head_dim, bias=False)
        else:
            self.qkv = nn.Linear(dim, all_head_dim * 3, bias=False)

        if qkv_bias:
            self.q_bias = nn.Parameter(torch.zeros(all_head_dim))
            self.v_bias = nn.Parameter(torch.zeros(all_head_dim))
        else:
            self.q_bias = None
            self.v_bias = None

        self.rel_pos_bias = None
        self.qk_float = True

        self.window_size = None
        self.relative_position_bias_table = None

        if window_size:
            if use_decoupled_rel_pos_bias:
                self.rel_pos_bias = DecoupledRelativePositionBias(window_size=window_size, num_heads=num_heads)
            else:
                self.window_size = window_size
                self.num_relative_distance = (2 * window_size[0] - 1) * (2 * window_size[1] - 1) + 3    # (2*14-1) * (2*14-1) + 3
                self.relative_position_bias_table = nn.Parameter(
                    torch.zeros(self.num_relative_distance, num_heads))  # 2*Wh-1 * 2*Ww-1, nH
                # cls to token & token 2 cls & cls to cls

                # get pair-wise relative position index for each token inside the window
                coords_h = torch.arange(window_size[0])
                coords_w = torch.arange(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] += window_size[0] - 1  # shift to start from 0
                relative_coords[:, :, 1] += window_size[1] - 1
                relative_coords[:, :, 0] *= 2 * window_size[1] - 1
                relative_position_index = \
                    torch.zeros(size=(window_size[0] * window_size[1] + 1, ) * 2, dtype=relative_coords.dtype)
                relative_position_index[1:, 1:] = relative_coords.sum(-1)  # Wh*Ww, Wh*Ww
                relative_position_index[0, 0:] = self.num_relative_distance - 3
                relative_position_index[0:, 0] = self.num_relative_distance - 2
                relative_position_index[0, 0] = self.num_relative_distance - 1

                self.register_buffer("relative_position_index", relative_position_index)

        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(all_head_dim, dim)
        self.proj_drop = nn.Dropout(proj_drop)

        self.xattn = xattn
        self.rope = rope

    def forward(self, x, rel_pos_bias=None, attn_mask=None):
        B, N, C = x.shape
        
        if self.deepnorm or self.subln: 
            q = F.linear(input=x, weight=self.q_proj.weight, bias=self.q_bias)
            k = F.linear(input=x, weight=self.k_proj.weight, bias=None)
            v = F.linear(input=x, weight=self.v_proj.weight, bias=self.v_bias)

            q = q.reshape(B, N, self.num_heads, -1).permute(0, 2, 1, 3)     # B, num_heads, N, C
            k = k.reshape(B, N, self.num_heads, -1).permute(0, 2, 1, 3)  
            v = v.reshape(B, N, self.num_heads, -1).permute(0, 2, 1, 3) 
        else: 
            qkv_bias = None
            if self.q_bias is not None:
                qkv_bias = torch.cat((self.q_bias, torch.zeros_like(self.v_bias, requires_grad=False), self.v_bias))
            qkv = F.linear(input=x, weight=self.qkv.weight, bias=qkv_bias)
            qkv = qkv.reshape(B, N, 3, self.num_heads, -1).permute(2, 0, 3, 1, 4)   # 3, B, num_heads, N, C
            q, k, v = qkv[0], qkv[1], qkv[2]   

        if self.rope:
            q_t = q[:, :, 1:, :]
            ro_q_t = self.rope(q_t)
            q = torch.cat((q[:, :, :1, :], ro_q_t), -2).type_as(v)

            k_t = k[:, :, 1:, :]
            ro_k_t = self.rope(k_t)
            k = torch.cat((k[:, :, :1, :], ro_k_t), -2).type_as(v)

        if self.xattn:
            q = q.permute(0, 2, 1, 3)   # B, num_heads, N, C -> B, N, num_heads, C
            k = k.permute(0, 2, 1, 3)
            v = v.permute(0, 2, 1, 3)

            x = xops.memory_efficient_attention(q, k, v)
            x = x.reshape(B, N, -1)
            x = self.proj(x)
            x = self.proj_drop(x)
        else:
            q = q * self.scale
            if self.qk_float:
                attn = (q.float() @ k.float().transpose(-2, -1))
            else:
                attn = (q @ k.transpose(-2, -1))

            if self.relative_position_bias_table is not None:
                relative_position_bias = \
                    self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
                        self.window_size[0] * self.window_size[1] + 1,
                        self.window_size[0] * self.window_size[1] + 1, -1)  # Wh*Ww,Wh*Ww,nH
                relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww
                attn = attn + relative_position_bias.unsqueeze(0).type_as(attn)

            if self.rel_pos_bias is not None:
                attn = attn + self.rel_pos_bias().type_as(attn)

            if rel_pos_bias is not None:
                attn = attn + rel_pos_bias.type_as(attn)
            if attn_mask is not None:
                attn_mask = attn_mask.bool()
                attn = attn.masked_fill(~attn_mask[:, None, None, :], float("-inf"))
            attn = attn.softmax(dim=-1).type_as(x)
            attn = self.attn_drop(attn)

            x = (attn @ v).transpose(1, 2).reshape(B, N, -1)
            x = self.proj(x)
            x = self.proj_drop(x)

        return x


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., init_values=None, act_layer=nn.GELU, norm_layer=nn.LayerNorm,
                 window_size=None, attn_head_dim=None, use_decoupled_rel_pos_bias=False,
                 depth=None,
                 postnorm=False, 
                 deepnorm=False,
                 subln=False,
                 xattn=False,
                 swiglu=False,
                 naiveswiglu=False,
                 rope=None,
                ):
        super().__init__()
        self.norm1 = norm_layer(dim)
        self.attn = Attention(
            dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale,
            attn_drop=attn_drop, proj_drop=drop, window_size=window_size, 
            use_decoupled_rel_pos_bias=use_decoupled_rel_pos_bias, attn_head_dim=attn_head_dim,
            deepnorm=deepnorm,
            subln=subln,
            norm_layer=norm_layer,
            xattn=xattn,
            rope=rope,
        )

        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)
        if swiglu:
            self.mlp = xops.SwiGLU(
                in_features=dim, 
                hidden_features=mlp_hidden_dim
            ) # hidden_features: 2/3
        elif naiveswiglu:
            self.mlp = SwiGLU(
                in_features=dim, 
                hidden_features=mlp_hidden_dim, 
                subln=subln,
                norm_layer=norm_layer,
            )
        else:
            self.mlp = Mlp(
                in_features=dim, 
                hidden_features=mlp_hidden_dim, 
                subln=subln,
                norm_layer=norm_layer
            ) 

        if init_values is not None and init_values > 0:
            self.gamma_1 = nn.Parameter(init_values * torch.ones((dim)),requires_grad=True)
            self.gamma_2 = nn.Parameter(init_values * torch.ones((dim)),requires_grad=True)
        else:
            self.gamma_1, self.gamma_2 = None, None

        self.deepnorm = deepnorm
        if self.deepnorm: self.alpha = math.pow(2.0 * depth, 0.25)
        
        self.postnorm = postnorm

    def forward(self, x, rel_pos_bias=None, attn_mask=None):
        if self.gamma_1 is None:
            if self.postnorm:
                x = x + self.drop_path(
                    self.norm1(self.attn(x, rel_pos_bias=rel_pos_bias, attn_mask=attn_mask)))
                x = x + self.drop_path(self.norm2(self.mlp(x)))
            elif self.deepnorm:
                residual = x
                x = self.attn(x, rel_pos_bias=rel_pos_bias, attn_mask=attn_mask)
                x = self.drop_path(x)
                x = residual * self.alpha + x
                x = self.norm1(x)

                residual = x
                x = self.mlp(x)
                x = self.drop_path(x)
                x = residual * self.alpha + x
                x = self.norm2(x)
            else:
                x = x + self.drop_path(
                    self.attn(self.norm1(x), rel_pos_bias=rel_pos_bias, attn_mask=attn_mask))
                x = x + self.drop_path(self.mlp(self.norm2(x)))
        else:
            if self.postnorm:
                x = x + self.drop_path(
                    self.gamma_1 * self.norm1(self.attn(x, rel_pos_bias=rel_pos_bias, attn_mask=attn_mask)))
                x = x + self.drop_path(self.gamma_2 * self.norm2(self.mlp(x)))
            else:
                x = x + self.drop_path(
                    self.gamma_1 * self.attn(self.norm1(x), rel_pos_bias=rel_pos_bias, attn_mask=attn_mask))
                x = x + self.drop_path(self.gamma_2 * 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__()
        img_size = to_2tuple(img_size)
        patch_size = to_2tuple(patch_size)
        num_patches = (img_size[1] // patch_size[1]) * (img_size[0] // patch_size[0])
        self.patch_shape = (img_size[0] // patch_size[0], img_size[1] // patch_size[1])
        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, **kwargs):
        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).flatten(2).transpose(1, 2)
        return x


class RelativePositionBias(nn.Module):

    def __init__(self, window_size, num_heads):
        super().__init__()
        self.window_size = window_size
        self.num_relative_distance = (2 * window_size[0] - 1) * (2 * window_size[1] - 1) + 3
        self.relative_position_bias_table = nn.Parameter(
            torch.zeros(self.num_relative_distance, num_heads))  # 2*Wh-1 * 2*Ww-1, nH
        # cls to token & token 2 cls & cls to cls

        # get pair-wise relative position index for each token inside the window
        coords_h = torch.arange(window_size[0])
        coords_w = torch.arange(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] += window_size[0] - 1  # shift to start from 0
        relative_coords[:, :, 1] += window_size[1] - 1
        relative_coords[:, :, 0] *= 2 * window_size[1] - 1
        relative_position_index = \
            torch.zeros(size=(window_size[0] * window_size[1] + 1,) * 2, dtype=relative_coords.dtype)
        relative_position_index[1:, 1:] = relative_coords.sum(-1)  # Wh*Ww, Wh*Ww
        relative_position_index[0, 0:] = self.num_relative_distance - 3
        relative_position_index[0:, 0] = self.num_relative_distance - 2
        relative_position_index[0, 0] = self.num_relative_distance - 1

        self.register_buffer("relative_position_index", relative_position_index)

    def forward(self):
        relative_position_bias = \
            self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
                self.window_size[0] * self.window_size[1] + 1,
                self.window_size[0] * self.window_size[1] + 1, -1)  # Wh*Ww,Wh*Ww,nH
        return relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww


def _mask_1d_rel_pos_index(seq_len):
    index = torch.arange(seq_len)
    return index.view(1, seq_len) - index.view(seq_len, 1) + seq_len - 1


def _add_cls_to_index_matrix(index, num_tokens, offset):
    index = index.contiguous().view(num_tokens, num_tokens)
    new_index = torch.zeros(size=(num_tokens + 1, num_tokens + 1), dtype=index.dtype)
    new_index[1:, 1:] = index
    new_index[0, 0:] = offset
    new_index[0:, 0] = offset + 1
    new_index[0, 0] = offset + 2
    return new_index



class DecoupledRelativePositionBias(nn.Module):

    def __init__(self, window_size, num_heads):
        super().__init__()
        self.window_size = window_size
        self.num_relative_distance = (2 * window_size[0] + 2, 2 * window_size[1] + 2)

        num_tokens = window_size[0] * window_size[1]

        self.relative_position_bias_for_high = nn.Parameter(torch.zeros(self.num_relative_distance[0], num_heads))
        self.relative_position_bias_for_width = nn.Parameter(torch.zeros(self.num_relative_distance[1], num_heads))
        # cls to token & token 2 cls & cls to cls

        h_index = _mask_1d_rel_pos_index(window_size[0]).view(
            window_size[0], 1, window_size[0], 1).expand(-1, window_size[1], -1, window_size[1])
        h_index = _add_cls_to_index_matrix(h_index, num_tokens, 2 * window_size[0] - 1)
        self.register_buffer("relative_position_high_index", h_index)

        w_index = _mask_1d_rel_pos_index(window_size[1]).view(
            1, window_size[1], 1, window_size[1]).expand(window_size[0], -1, window_size[0], -1)
        w_index = _add_cls_to_index_matrix(w_index, num_tokens, 2 * window_size[1] - 1)

        self.register_buffer("relative_position_width_index", w_index)

    def forward(self):
        relative_position_bias = \
            F.embedding(input=self.relative_position_high_index, weight=self.relative_position_bias_for_high) + \
            F.embedding(input=self.relative_position_width_index, weight=self.relative_position_bias_for_width)
        return relative_position_bias.permute(2, 0, 1).contiguous()



class VisionTransformer(nn.Module):
    """ Vision Transformer with support for patch or hybrid CNN input stage
    """
    def __init__(self, img_size=224, patch_size=16, in_chans=3, num_classes=1000, 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=nn.LayerNorm, init_values=None, use_abs_pos_emb=True,
                 use_rel_pos_bias=False, use_shared_rel_pos_bias=False, use_decoupled_rel_pos_bias=False,
                 use_mean_pooling=True, init_scale=0.001, use_checkpoint=True, stop_grad_conv1=True,
                 postnorm=False,
                 subln=False,
                 xattn=False,
                 swiglu=False,
                 naiveswiglu=False,
                 rope=False,
                 pt_hw_seq_len=16,
                 intp_freq=False,
            ):
        super().__init__()
        self.num_classes = num_classes
        self.num_features = self.embed_dim = embed_dim  # num_features for consistency with other models

        self.patch_embed = PatchEmbed(
            img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim)
        num_patches = self.patch_embed.num_patches

        self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
        if use_abs_pos_emb:
            self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + 1, embed_dim))
        else:
            self.pos_embed = None
        self.pos_drop = nn.Dropout(p=drop_rate)

        if use_shared_rel_pos_bias:
            self.rel_pos_bias = RelativePositionBias(window_size=self.patch_embed.patch_shape, num_heads=num_heads)
        else:
            self.rel_pos_bias = None

        self.use_decoupled_rel_pos_bias = use_decoupled_rel_pos_bias
        self.use_checkpoint = use_checkpoint
        self.stop_grad_conv1 = stop_grad_conv1

        if use_decoupled_rel_pos_bias or use_rel_pos_bias:
            window_size = self.patch_embed.patch_shape
        else:
            window_size = None

        if rope:
            half_head_dim = embed_dim // num_heads // 2
            hw_seq_len = img_size // patch_size
            self.rope = VisionRotaryEmbeddingFast(
                dim=half_head_dim,
                pt_seq_len=pt_hw_seq_len,
                ft_seq_len=hw_seq_len if intp_freq else None,
            )
        else: self.rope = None

        self.swiglu = swiglu
        self.naiveswiglu = naiveswiglu

        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)]  # stochastic depth decay rule
        self.use_rel_pos_bias = use_rel_pos_bias
        self.blocks = nn.ModuleList([
            Block(
                dim=embed_dim, num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
                drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer,
                init_values=init_values, window_size=window_size, use_decoupled_rel_pos_bias=use_decoupled_rel_pos_bias,
                postnorm=postnorm,
                subln=subln,
                xattn=xattn,
                swiglu=swiglu,
                naiveswiglu=naiveswiglu,
                rope=self.rope,
            )
            for i in range(depth)])
        self.norm = nn.Identity() if use_mean_pooling else norm_layer(embed_dim)
        self.fc_norm = norm_layer(embed_dim) if use_mean_pooling else None
        self.head = nn.Linear(embed_dim, num_classes) if num_classes > 0 else nn.Identity()

        if self.pos_embed is not None:
            trunc_normal_(self.pos_embed, std=.02)
        trunc_normal_(self.cls_token, std=.02)
        # trunc_normal_(self.mask_token, std=.02)
        if isinstance(self.head, nn.Linear):
            trunc_normal_(self.head.weight, std=.02)
        self.apply(self._init_weights)
        self.fix_init_weight()

        if isinstance(self.head, nn.Linear):
            self.head.weight.data.mul_(init_scale)
            self.head.bias.data.mul_(init_scale)

    def fix_init_weight(self):
        def rescale(param, layer_id):
            param.div_(math.sqrt(2.0 * layer_id))

        for layer_id, layer in enumerate(self.blocks):
            rescale(layer.attn.proj.weight.data, layer_id + 1)
            if self.swiglu or self.naiveswiglu:
                rescale(layer.mlp.w3.weight.data, layer_id + 1)
            else:
                rescale(layer.mlp.fc2.weight.data, layer_id + 1)

    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 get_num_layers(self):
        return len(self.blocks)

    @torch.jit.ignore
    def no_weight_decay(self):
        return {'pos_embed', 'cls_token'}

    def get_classifier(self):
        return self.head

    def reset_classifier(self, num_classes, global_pool=''):
        self.num_classes = num_classes
        self.head = nn.Linear(self.embed_dim, num_classes) if num_classes > 0 else nn.Identity()

    def forward_features(self, x, return_patch_tokens=False):
        x = self.patch_embed(x)

        if self.stop_grad_conv1:
            x = x.detach()
        
        batch_size, seq_len, _ = x.size()

        cls_tokens = self.cls_token.expand(batch_size, -1, -1)
        x = torch.cat((cls_tokens, x), dim=1)
        if self.pos_embed is not None:
            x = x + self.pos_embed
        x = self.pos_drop(x)

        rel_pos_bias = self.rel_pos_bias() if self.rel_pos_bias is not None else None
        for blk in self.blocks:
            if self.use_checkpoint:
                x = checkpoint.checkpoint(blk, x, rel_pos_bias)
            else:
                x = blk(x, rel_pos_bias)

        x = self.norm(x)
        if self.fc_norm is not None:
            t = x[:, 1:, :]
            if return_patch_tokens:
                return self.fc_norm(t)
            else:
                return self.fc_norm(t.mean(1))
        else:
            if return_patch_tokens:
                return x[:, 1:]
            else:
                return x[:, 0]

    def forward(self, x, return_patch_tokens=False):
        x = self.forward_features(x, return_patch_tokens=return_patch_tokens)
        x = self.head(x)
        return x




    


@register_model
def eva02_tiny_patch14_xattn_fusedLN_SwiGLU_preln_RoPE(pretrained=False, **kwargs):
    model = VisionTransformer(
        patch_size=14, 
        embed_dim=192, 
        depth=12, 
        num_heads=3, 
        mlp_ratio=4*2/3, 
        qkv_bias=True,
        norm_layer=partial(FusedLayerNorm, eps=1e-6),
        xattn=True,
        swiglu=True,
        rope=True,
        pt_hw_seq_len=16,   # 224/14
        intp_freq=True,
        **kwargs)
    model.default_cfg = _cfg()
    if pretrained:
        checkpoint = torch.load(
            kwargs["init_ckpt"], map_location="cpu"
        )
        model.load_state_dict(checkpoint["model"])
    return model
    



@register_model
def eva02_small_patch14_xattn_fusedLN_SwiGLU_preln_RoPE(pretrained=False, **kwargs):
    model = VisionTransformer(
        patch_size=14, 
        embed_dim=384, 
        depth=12, 
        num_heads=6, 
        mlp_ratio=4*2/3, 
        qkv_bias=True,
        norm_layer=partial(FusedLayerNorm, eps=1e-6),
        xattn=True,
        swiglu=True,
        rope=True,
        pt_hw_seq_len=16,   # 224/14
        intp_freq=True,
        **kwargs)
    model.default_cfg = _cfg()
    if pretrained:
        checkpoint = torch.load(
            kwargs["init_ckpt"], map_location="cpu"
        )
        model.load_state_dict(checkpoint["model"])
    return model




@register_model
def eva02_base_patch14_xattn_fusedLN_NaiveSwiGLU_subln_RoPE(pretrained=False, **kwargs):
    model = VisionTransformer(
        patch_size=14, 
        embed_dim=768, 
        depth=12, 
        num_heads=12, 
        mlp_ratio=4*2/3, 
        qkv_bias=True,
        norm_layer=partial(FusedLayerNorm, eps=1e-6), 
        subln=True,
        xattn=True,
        naiveswiglu=True,
        rope=True, 
        pt_hw_seq_len=16,   # 224/14
        intp_freq=True,
        **kwargs)
    model.default_cfg = _cfg()
    if pretrained:
        checkpoint = torch.load(
            kwargs["init_ckpt"], map_location="cpu"
        )
        model.load_state_dict(checkpoint["model"])
    return model




@register_model
def eva02_large_patch14_xattn_fusedLN_NaiveSwiGLU_subln_RoPE(pretrained=False, **kwargs):
    model = VisionTransformer(
        patch_size=14, 
        embed_dim=1024, 
        depth=24, 
        num_heads=16, 
        mlp_ratio=4*2/3, 
        qkv_bias=True,
        norm_layer=partial(FusedLayerNorm, eps=1e-6), 
        subln=True,
        xattn=True,
        naiveswiglu=True,
        rope=True, 
        pt_hw_seq_len=16,   # 224/14
        intp_freq=True,
        **kwargs)
    model.default_cfg = _cfg()
    if pretrained:
        checkpoint = torch.load(
            kwargs["init_ckpt"], map_location="cpu"
        )
        model.load_state_dict(checkpoint["model"])
    return model