ru-clip/ruclip/model.py

# -*- coding: utf-8 -*-
import os
import json
from collections import OrderedDict

import torch
import numpy as np
from torch import nn


class LayerNorm(nn.LayerNorm):
    """Subclass torch's LayerNorm to handle fp16."""

    def forward(self, x: torch.Tensor):
        orig_type = x.dtype
        ret = super().forward(x.type(torch.float32))
        return ret.type(orig_type)


class QuickGELU(nn.Module):
    def forward(self, x: torch.Tensor):
        return x * torch.sigmoid(1.702 * x)


class ResidualAttentionBlock(nn.Module):
    def __init__(self, d_model: int, n_head: int, attn_mask: torch.Tensor = None):
        super().__init__()

        self.attn = nn.MultiheadAttention(d_model, n_head)
        self.ln_1 = LayerNorm(d_model)
        self.mlp = nn.Sequential(OrderedDict([
            ('c_fc', nn.Linear(d_model, d_model * 4)),
            ('gelu', QuickGELU()),
            ('c_proj', nn.Linear(d_model * 4, d_model))
        ]))
        self.ln_2 = LayerNorm(d_model)
        self.attn_mask = attn_mask

    def attention(self, x: torch.Tensor):
        self.attn_mask = self.attn_mask.to(dtype=x.dtype, device=x.device) if self.attn_mask is not None else None
        return self.attn(x, x, x, need_weights=False, attn_mask=self.attn_mask)[0]

    def forward(self, x: torch.Tensor):
        x = x + self.attention(self.ln_1(x))
        x = x + self.mlp(self.ln_2(x))
        return x


class Transformer(nn.Module):
    def __init__(self, width: int, layers: int, heads: int, attn_mask: torch.Tensor = None):
        super().__init__()
        self.width = width
        self.layers = layers
        self.resblocks = nn.Sequential(*[ResidualAttentionBlock(width, heads, attn_mask) for _ in range(layers)])

    def forward(self, x: torch.Tensor):
        return self.resblocks(x)


class VisionTransformer(nn.Module):
    def __init__(self, input_resolution: int, patch_size: int, width: int, layers: int, heads: int, output_dim: int):
        super().__init__()
        self.input_resolution = input_resolution
        self.output_dim = output_dim
        self.conv1 = nn.Conv2d(in_channels=3, out_channels=width, kernel_size=patch_size, stride=patch_size, bias=False)

        scale = width ** -0.5
        self.class_embedding = nn.Parameter(scale * torch.randn(width))
        self.positional_embedding = nn.Parameter(scale * torch.randn((input_resolution // patch_size) ** 2 + 1, width))
        self.ln_pre = LayerNorm(width)

        self.transformer = Transformer(width, layers, heads)

        self.ln_post = LayerNorm(width)
        self.proj = nn.Parameter(scale * torch.randn(width, output_dim))

    def forward(self, x: torch.Tensor):
        x = self.conv1(x)  # shape = [*, width, grid, grid]
        x = x.reshape(x.shape[0], x.shape[1], -1)  # shape = [*, width, grid ** 2]
        x = x.permute(0, 2, 1)  # shape = [*, grid ** 2, width]
        x = torch.cat([
            self.class_embedding.to(x.dtype) +
            torch.zeros(x.shape[0], 1, x.shape[-1], dtype=x.dtype, device=x.device), x
        ], dim=1)  # shape = [*, grid ** 2 + 1, width]
        x = x + self.positional_embedding.to(x.dtype)
        x = self.ln_pre(x)

        x = x.permute(1, 0, 2)  # NLD -> LND
        x = self.transformer(x)
        x = x.permute(1, 0, 2)  # LND -> NLD

        x = self.ln_post(x[:, 0, :])

        if self.proj is not None:
            x = x @ self.proj

        return x


class CLIP(nn.Module):
    def __init__(
            self,
            embed_dim,
            image_resolution,
            vision_layers,
            vision_width,
            vision_patch_size,
            context_length,
            vocab_size,
            transformer_width,
            transformer_heads,
            transformer_layers,
            eos_id=3,
    ):
        super().__init__()

        self.eos_id = eos_id
        self.context_length = context_length

        vision_heads = vision_width // 64
        self.visual = VisionTransformer(
            input_resolution=image_resolution,
            patch_size=vision_patch_size,
            width=vision_width,
            layers=vision_layers,
            heads=vision_heads,
            output_dim=embed_dim,
        )

        self.transformer = Transformer(
            width=transformer_width,
            layers=transformer_layers,
            heads=transformer_heads,
            attn_mask=self.build_attention_mask(),
        )

        self.vocab_size = vocab_size
        self.token_embedding = nn.Embedding(vocab_size, transformer_width)
        self.positional_embedding = nn.Parameter(torch.empty(self.context_length, transformer_width))
        self.ln_final = LayerNorm(transformer_width)

        self.text_projection = nn.Parameter(torch.empty(transformer_width, embed_dim))
        self.logit_scale = nn.Parameter(torch.ones([]) * np.log(1 / 0.07))

        self.initialize_parameters()

    def initialize_parameters(self):
        nn.init.normal_(self.token_embedding.weight, std=0.02)
        nn.init.normal_(self.positional_embedding, std=0.01)
        proj_std = (self.transformer.width ** -0.5) * ((2 * self.transformer.layers) ** -0.5)
        attn_std = self.transformer.width ** -0.5
        fc_std = (2 * self.transformer.width) ** -0.5
        for block in self.transformer.resblocks:
            nn.init.normal_(block.attn.in_proj_weight, std=attn_std)
            nn.init.normal_(block.attn.out_proj.weight, std=proj_std)
            nn.init.normal_(block.mlp.c_fc.weight, std=fc_std)
            nn.init.normal_(block.mlp.c_proj.weight, std=proj_std)

        if self.text_projection is not None:
            nn.init.normal_(self.text_projection, std=self.transformer.width ** -0.5)

    def build_attention_mask(self):
        mask = torch.empty(self.context_length, self.context_length)
        mask.fill_(float('-inf'))
        mask.triu_(1)
        return mask

    @property
    def dtype(self):
        return self.visual.conv1.weight.dtype

    def encode_image(self, pixel_values):
        """Encode images
        Parameters
        ----------
        pixel_values: torch.Tensor
            Processed images from RuCLIPProcessor class
        Returns
        -------
        image_latents : torch.Tensor
            Image embeddings
        """
        return self.visual(pixel_values.type(self.dtype))

    def encode_text(self, input_ids):
        """Encode texts
        Parameters
        ----------
        input_ids: torch.Tensor
            Tokenized texts from RuCLIPProcessor class
        Returns
        -------
        text_latents : torch.Tensor
            Text embeddings
        """
        x = self.token_embedding(input_ids).type(self.dtype)  # [batch_size, n_ctx, d_model]
        x = x + self.positional_embedding.type(self.dtype)
        x = x.permute(1, 0, 2)  # NLD -> LND
        x = self.transformer(x)
        x = x.permute(1, 0, 2)  # LND -> NLD
        x = self.ln_final(x).type(self.dtype)
        # x.shape = [batch_size, n_ctx, transformer.width]
        x = x[torch.arange(x.shape[0]), torch.where(input_ids == self.eos_id)[1]] @ self.text_projection
        return x

    def forward(self, input_ids, pixel_values):
        image_features = self.encode_image(pixel_values)
        text_features = self.encode_text(input_ids)

        # normalize features
        image_features = image_features / image_features.norm(dim=-1, keepdim=True)
        text_features = text_features / text_features.norm(dim=-1, keepdim=True)

        # cosine similarity as logits
        logit_scale = self.logit_scale.exp()
        logits_per_image = logit_scale * image_features @ text_features.t()
        logits_per_text = logits_per_image.t()

        return logits_per_image, logits_per_text

    @classmethod
    def from_pretrained(cls, folder):
        """Load model from folder"""
        config = json.load(open(os.path.join(folder, 'config.json')))
        model = cls(
            embed_dim=config['embed_dim'],
            image_resolution=config['image_resolution'],
            vision_layers=config['vision_layers'],
            vision_width=config['vision_width'],
            vision_patch_size=config['vision_patch_size'],
            context_length=config['context_length'],
            vocab_size=config['vocab_size'],
            transformer_width=config['transformer_width'],
            transformer_heads=config['transformer_heads'],
            transformer_layers=config['transformer_layers'],
        )
        checkpoint = torch.load(os.path.join(folder, 'pytorch_model.bin'), map_location='cpu')
        model.load_state_dict(checkpoint)
        return model