llava/model/multimodal_encoder/eva_clip/eva_vit.py

"""
# Adapted from https://github.com/baaivision/EVA/tree/master/EVA-CLIP
"""

from math import pi
import torch
from torch import nn
from einops import rearrange, repeat
import logging
from llava.utils import rank0_print


def broadcat(tensors, dim=-1):
    num_tensors = len(tensors)
    shape_lens = set(list(map(lambda t: len(t.shape), tensors)))
    assert len(shape_lens) == 1, "tensors must all have the same number of dimensions"
    shape_len = list(shape_lens)[0]
    dim = (dim + shape_len) if dim < 0 else dim
    dims = list(zip(*map(lambda t: list(t.shape), tensors)))
    expandable_dims = [(i, val) for i, val in enumerate(dims) if i != dim]
    assert all([*map(lambda t: len(set(t[1])) <= 2, expandable_dims)]), "invalid dimensions for broadcastable concatentation"
    max_dims = list(map(lambda t: (t[0], max(t[1])), expandable_dims))
    expanded_dims = list(map(lambda t: (t[0], (t[1],) * num_tensors), max_dims))
    expanded_dims.insert(dim, (dim, dims[dim]))
    expandable_shapes = list(zip(*map(lambda t: t[1], expanded_dims)))
    tensors = list(map(lambda t: t[0].expand(*t[1]), zip(tensors, expandable_shapes)))
    return torch.cat(tensors, dim=dim)


def rotate_half(x):
    x = rearrange(x, "... (d r) -> ... d r", r=2)
    x1, x2 = x.unbind(dim=-1)
    x = torch.stack((-x2, x1), dim=-1)
    return rearrange(x, "... d r -> ... (d r)")


class VisionRotaryEmbeddingFast(nn.Module):
    def __init__(self, dim, pt_seq_len, ft_seq_len=None, custom_freqs=None, freqs_for="lang", theta=10000, max_freq=10, num_freqs=1, patch_dropout=0.0):
        super().__init__()
        if custom_freqs:
            freqs = custom_freqs
        elif freqs_for == "lang":
            freqs = 1.0 / (theta ** (torch.arange(0, dim, 2)[: (dim // 2)].float() / dim))
        elif freqs_for == "pixel":
            freqs = torch.linspace(1.0, max_freq / 2, dim // 2) * pi
        elif freqs_for == "constant":
            freqs = torch.ones(num_freqs).float()
        else:
            raise ValueError(f"unknown modality {freqs_for}")

        if ft_seq_len is None:
            ft_seq_len = pt_seq_len
        t = torch.arange(ft_seq_len) / ft_seq_len * pt_seq_len

        freqs = torch.einsum("..., f -> ... f", t, freqs)
        freqs = repeat(freqs, "... n -> ... (n r)", r=2)
        freqs = broadcat((freqs[:, None, :], freqs[None, :, :]), dim=-1)

        freqs_cos = freqs.cos().view(-1, freqs.shape[-1])
        freqs_sin = freqs.sin().view(-1, freqs.shape[-1])

        self.patch_dropout = patch_dropout

        self.register_buffer("freqs_cos", freqs_cos)
        self.register_buffer("freqs_sin", freqs_sin)

        logging.info(f"Shape of rope freq: {self.freqs_cos.shape}")

    def forward(self, t, patch_indices_keep=None):
        if patch_indices_keep is not None:
            batch = t.size()[0]
            batch_indices = torch.arange(batch)
            batch_indices = batch_indices[..., None]

            freqs_cos = repeat(self.freqs_cos, "i j -> n i m j", n=t.shape[0], m=t.shape[1])
            freqs_sin = repeat(self.freqs_sin, "i j -> n i m j", n=t.shape[0], m=t.shape[1])

            freqs_cos = freqs_cos[batch_indices, patch_indices_keep]
            freqs_cos = rearrange(freqs_cos, "n i m j -> n m i j")
            freqs_sin = freqs_sin[batch_indices, patch_indices_keep]
            freqs_sin = rearrange(freqs_sin, "n i m j -> n m i j")

            return t * freqs_cos + rotate_half(t) * freqs_sin

        return t * self.freqs_cos + rotate_half(t) * self.freqs_sin


class LayerNorm(nn.LayerNorm):
    """Subclass torch's LayerNorm (with cast back to input dtype)."""

    def forward(self, x: torch.Tensor):
        orig_type = x.dtype
        x = F.layer_norm(x, self.normalized_shape, self.weight, self.bias, self.eps)
        return x.to(orig_type)


class PatchDropout(nn.Module):
    """
    https://arxiv.org/abs/2212.00794
    """

    def __init__(self, prob, exclude_first_token=True):
        super().__init__()
        assert 0 <= prob < 1.0
        self.prob = prob
        self.exclude_first_token = exclude_first_token  # exclude CLS token
        logging.info(f"os.getenv('RoPE')={os.getenv('RoPE')}")

    def forward(self, x):
        if not self.training or self.prob == 0.0:
            return x

        if self.exclude_first_token:
            cls_tokens, x = x[:, :1], x[:, 1:]
        else:
            cls_tokens = torch.jit.annotate(torch.Tensor, x[:, :1])

        batch = x.size()[0]
        num_tokens = x.size()[1]

        batch_indices = torch.arange(batch)
        batch_indices = batch_indices[..., None]

        keep_prob = 1 - self.prob
        num_patches_keep = max(1, int(num_tokens * keep_prob))

        rand = torch.randn(batch, num_tokens)
        patch_indices_keep = rand.topk(num_patches_keep, dim=-1).indices

        x = x[batch_indices, patch_indices_keep]

        if self.exclude_first_token:
            x = torch.cat((cls_tokens, x), dim=1)

        if self.training and os.getenv("RoPE") == "1":
            return x, patch_indices_keep

        return x


# --------------------------------------------------------
# Adapted from  https://github.com/microsoft/unilm/tree/master/beit
# --------------------------------------------------------
import math
import os
import torch.nn as nn
import torch.nn.functional as F

try:
    from timm.models.layers import drop_path, to_2tuple, trunc_normal_
except:
    from timm.layers import drop_path, to_2tuple, trunc_normal_

if os.getenv("ENV_TYPE") == "deepspeed":
    try:
        from deepspeed.runtime.activation_checkpointing.checkpointing import checkpoint
    except:
        from torch.utils.checkpoint import checkpoint
else:
    from torch.utils.checkpoint import checkpoint

try:
    import xformers.ops as xops
except ImportError:
    xops = None
    # print("Please 'pip install xformers'")


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,
        norm_layer=nn.LayerNorm,
        drop=0.0,
        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.drop(x)
        # commit this for the orignal BERT implement
        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=0.0, 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=0.0, proj_drop=0.0, window_size=None, attn_head_dim=None, xattn=False, rope=None, subln=False, norm_layer=nn.LayerNorm):
        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.subln = subln
        if 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

        if window_size:
            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)
        else:
            self.window_size = None
            self.relative_position_bias_table = None
            self.relative_position_index = None

        self.attn_drop = nn.Dropout(attn_drop)
        self.inner_attn_ln = norm_layer(all_head_dim) if subln else nn.Identity()
        # self.proj = nn.Linear(all_head_dim, all_head_dim)
        self.proj = nn.Linear(all_head_dim, dim)
        self.proj_drop = nn.Dropout(proj_drop)
        self.xattn = xattn
        self.xattn_drop = attn_drop

        self.rope = rope

    def forward(self, x, rel_pos_bias=None, attn_mask=None):
        B, N, C = x.shape
        if 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:
            # slightly fast impl
            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 and xops is not None:
            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,
                p=self.xattn_drop,
                scale=self.scale,
            )
            x = x.reshape(B, N, -1)
            x = self.inner_attn_ln(x)
            x = self.proj(x)
            x = self.proj_drop(x)
        else:
            q = q * self.scale
            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 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)
            attn = self.attn_drop(attn)

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


class Block(nn.Module):

    def __init__(
        self,
        dim,
        num_heads,
        mlp_ratio=4.0,
        qkv_bias=False,
        qk_scale=None,
        drop=0.0,
        attn_drop=0.0,
        drop_path=0.0,
        init_values=None,
        act_layer=nn.GELU,
        norm_layer=nn.LayerNorm,
        window_size=None,
        attn_head_dim=None,
        xattn=False,
        rope=None,
        postnorm=False,
        subln=False,
        naiveswiglu=False,
    ):
        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, attn_head_dim=attn_head_dim, xattn=xattn, rope=rope, subln=subln, norm_layer=norm_layer
        )
        # NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
        self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
        self.norm2 = norm_layer(dim)
        mlp_hidden_dim = int(dim * mlp_ratio)

        if 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, act_layer=act_layer, subln=subln, drop=drop)

        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.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)))
            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


class EVAVisionTransformer(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=4.0,
        qkv_bias=False,
        qk_scale=None,
        drop_rate=0.0,
        attn_drop_rate=0.0,
        drop_path_rate=0.0,
        norm_layer=nn.LayerNorm,
        init_values=None,
        patch_dropout=0.0,
        use_abs_pos_emb=True,
        use_rel_pos_bias=False,
        use_shared_rel_pos_bias=False,
        rope=False,
        use_mean_pooling=True,
        init_scale=0.001,
        grad_checkpointing=False,
        xattn=False,
        postnorm=False,
        pt_hw_seq_len=16,
        intp_freq=False,
        naiveswiglu=False,
        subln=False,
    ):
        super().__init__()
        self.image_size = img_size
        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))
        # self.mask_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

        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,
                # patch_dropout=patch_dropout
            )
        else:
            self.rope = None

        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=self.patch_embed.patch_shape if use_rel_pos_bias else None,
                    xattn=xattn,
                    rope=self.rope,
                    postnorm=postnorm,
                    subln=subln,
                    naiveswiglu=naiveswiglu,
                )
                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=0.02)

        trunc_normal_(self.cls_token, std=0.02)
        # trunc_normal_(self.mask_token, std=.02)

        self.apply(self._init_weights)
        self.fix_init_weight()

        if isinstance(self.head, nn.Linear):
            trunc_normal_(self.head.weight, std=0.02)
            self.head.weight.data.mul_(init_scale)
            self.head.bias.data.mul_(init_scale)

        # setting a patch_dropout of 0. would mean it is disabled and this function would be the identity fn
        self.patch_dropout = PatchDropout(patch_dropout) if patch_dropout > 0.0 else nn.Identity()

        self.grad_checkpointing = grad_checkpointing

    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.naiveswiglu:
                rescale(layer.mlp.w3.weight.data, layer_id + 1)
            else:
                rescale(layer.mlp.fc2.weight.data, layer_id + 1)

    def get_cast_dtype(self) -> torch.dtype:
        return self.blocks[0].mlp.fc2.weight.dtype

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

    def lock(self, unlocked_groups=0, freeze_bn_stats=False):
        assert unlocked_groups == 0, "partial locking not currently supported for this model"
        for param in self.parameters():
            param.requires_grad = False

    @torch.jit.ignore
    def set_grad_checkpointing(self, enable=True):
        self.grad_checkpointing = enable

    @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_all_features=False):

        x = self.patch_embed(x)
        batch_size, seq_len, _ = x.size()

        cls_tokens = self.cls_token.expand(batch_size, -1, -1)  # stole cls_tokens impl from Phil Wang, thanks
        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)

        # a patch_dropout of 0. would mean it is disabled and this function would do nothing but return what was passed in
        if os.getenv("RoPE") == "1":
            if self.training and not isinstance(self.patch_dropout, nn.Identity):
                x, patch_indices_keep = self.patch_dropout(x)
                # Directly pass patch_indices_keep to self.rope.forward
                x = self.rope.forward(x, patch_indices_keep=patch_indices_keep)
            else:
                # Pass None or omit the patch_indices_keep argument for default behavior
                x = self.rope.forward(x, patch_indices_keep=None)
                x = self.patch_dropout(x)
        else:
            x = self.patch_dropout(x)

        rel_pos_bias = self.rel_pos_bias() if self.rel_pos_bias is not None else None
        for i, blk in enumerate(self.blocks):
            if i == len(self.blocks) - 1:
                continue
            if self.grad_checkpointing:
                x = checkpoint(blk, x, (rel_pos_bias,))
            else:
                x = blk(x, rel_pos_bias=rel_pos_bias)

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

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


def load_state_dict(checkpoint_path: str, map_location: str = "cpu", model_key: str = "model|module|state_dict", is_openai: bool = False, skip_list: list = []):
    if is_openai:
        model = torch.jit.load(checkpoint_path, map_location="cpu").eval()
        state_dict = model.state_dict()
        for key in ["input_resolution", "context_length", "vocab_size"]:
            state_dict.pop(key, None)
    else:
        checkpoint = torch.load(checkpoint_path, map_location=map_location)
        for mk in model_key.split("|"):
            if isinstance(checkpoint, dict) and mk in checkpoint:
                state_dict = checkpoint[mk]
                break
            else:
                state_dict = checkpoint
        if next(iter(state_dict.items()))[0].startswith("module"):
            state_dict = {k[7:]: v for k, v in state_dict.items()}

    for k in skip_list:
        if k in list(state_dict.keys()):
            logging.info(f"Removing key {k} from pretrained checkpoint")
            del state_dict[k]

    if os.getenv("RoPE") == "1":
        for k in list(state_dict.keys()):
            if "freqs_cos" in k or "freqs_sin" in k:
                del state_dict[k]
    return state_dict


def load_clip_visual_state_dict(checkpoint_path: str, map_location: str = "cpu", is_openai: bool = False, skip_list: list = []):
    state_dict = load_state_dict(checkpoint_path, map_location=map_location, is_openai=is_openai, skip_list=skip_list)
    # for k in list(state_dict.keys()):
    #     if not k.startswith("visual."):
    #         del state_dict[k]
    # for k in list(state_dict.keys()):
    #     if k.startswith("visual."):
    #         new_k = k[7:]
    #         state_dict[new_k] = state_dict[k]
    #         del state_dict[k]
    return state_dict


from dataclasses import dataclass
from typing import Optional, Tuple, Union

try:
    from apex.normalization import FusedLayerNorm
except:
    FusedLayerNorm = LayerNorm
    # print("Please build and install Nvidia apex package with option '--cuda_ext' according to https://github.com/NVIDIA/apex#from-source .")


@dataclass
class CLIPVisionCfg:
    layers: Union[Tuple[int, int, int, int], int] = 12
    width: int = 768
    head_width: int = 64
    mlp_ratio: float = 4.0
    patch_size: int = 16
    image_size: Union[Tuple[int, int], int] = 224
    ls_init_value: Optional[float] = None  # layer scale initial value
    patch_dropout: float = 0.0  # what fraction of patches to dropout during training (0 would mean disabled and no patches dropped) - 0.5 to 0.75 recommended in the paper for optimal results
    global_average_pool: bool = False  # whether to global average pool the last embedding layer, instead of using CLS token (https://arxiv.org/abs/2205.01580)
    drop_path_rate: Optional[float] = None  # drop path rate
    timm_model_name: str = None  # a valid model name overrides layers, width, patch_size
    timm_model_pretrained: bool = False  # use (imagenet) pretrained weights for named model
    timm_pool: str = "avg"  # feature pooling for timm model ('abs_attn', 'rot_attn', 'avg', '')
    timm_proj: str = "linear"  # linear projection for timm model output ('linear', 'mlp', '')
    timm_proj_bias: bool = False  # enable bias final projection
    eva_model_name: str = None  # a valid eva model name overrides layers, width, patch_size
    qkv_bias: bool = True
    fusedLN: bool = False
    xattn: bool = False
    postnorm: bool = False
    rope: bool = False
    pt_hw_seq_len: int = 16  # 224/14
    intp_freq: bool = False
    naiveswiglu: bool = False
    subln: bool = False


def create_norm_layer_factory(use_fused_ln, eps=1e-6):
    # Otherwise, use the standard LayerNorm
    return lambda num_features: nn.LayerNorm(num_features, eps=eps)


def _build_vision_tower(vision_tower_path: str, embed_dim: int, vision_cfg: CLIPVisionCfg, **kwargs):
    if isinstance(vision_cfg, dict):
        vision_cfg = CLIPVisionCfg(**vision_cfg)

    if vision_cfg.eva_model_name:
        vision_heads = vision_cfg.width // vision_cfg.head_width
        # Determine the appropriate norm layer factory based on the configuration
        norm_layer_factory = create_norm_layer_factory(vision_cfg.fusedLN, eps=1e-6)

        visual = EVAVisionTransformer(
            img_size=vision_cfg.image_size,
            patch_size=vision_cfg.patch_size,
            num_classes=embed_dim,
            use_mean_pooling=vision_cfg.global_average_pool,  # False
            init_values=vision_cfg.ls_init_value,
            patch_dropout=vision_cfg.patch_dropout,
            embed_dim=vision_cfg.width,
            depth=vision_cfg.layers,
            num_heads=vision_heads,
            mlp_ratio=vision_cfg.mlp_ratio,
            qkv_bias=vision_cfg.qkv_bias,
            drop_path_rate=vision_cfg.drop_path_rate,
            norm_layer=norm_layer_factory,
            xattn=vision_cfg.xattn,
            rope=vision_cfg.rope,
            postnorm=vision_cfg.postnorm,
            pt_hw_seq_len=vision_cfg.pt_hw_seq_len,  # 224/14
            intp_freq=vision_cfg.intp_freq,
            naiveswiglu=vision_cfg.naiveswiglu,
            subln=vision_cfg.subln,
        )

        state_dict = load_clip_visual_state_dict(vision_tower_path)
        incompatible_keys = visual.load_state_dict(state_dict, strict=False)
        rank0_print("EVA-CLIP incompatible_keys:", incompatible_keys)

    return visual


class EVAEncoderWrapper(nn.Module):
    def __init__(self, vision_tower_pretrained, config):
        super(EVAEncoderWrapper, self).__init__()
        self.config = config
        self.config["vision_tower_path"] = vision_tower_pretrained
        self.model = _build_vision_tower(**self.config)

    def forward(self, image, **kwargs):
        encode = self.model(image, return_all_features=True)[:, 1:, :]  # remove the CLS token
        return encode

    @property
    def dtype(self):
        return list(self.parameters())[-1].dtype

    @property
    def device(self):
        return list(self.parameters())[-1].device