docs/finetrainers-src-codebase/finetrainers/models/cogvideox/base_specification.py

import functools
import os
from typing import Any, Dict, List, Optional, Tuple

import torch
from accelerate import init_empty_weights
from diffusers import (
    AutoencoderKLCogVideoX,
    CogVideoXDDIMScheduler,
    CogVideoXImageToVideoPipeline,
    CogVideoXPipeline,
    CogVideoXTransformer3DModel,
)
from PIL.Image import Image
from transformers import AutoModel, AutoTokenizer, T5EncoderModel, T5Tokenizer

from finetrainers.data import VideoArtifact
from finetrainers.logging import get_logger
from finetrainers.models.modeling_utils import ModelSpecification
from finetrainers.models.utils import DiagonalGaussianDistribution
from finetrainers.processors import ProcessorMixin, T5Processor
from finetrainers.typing import ArtifactType, SchedulerType
from finetrainers.utils import _enable_vae_memory_optimizations, get_non_null_items, safetensors_torch_save_function

from .utils import prepare_rotary_positional_embeddings


logger = get_logger()


class CogVideoXLatentEncodeProcessor(ProcessorMixin):
    r"""
    Processor to encode image/video into latents using the CogVideoX VAE.

    Args:
        output_names (`List[str]`):
            The names of the outputs that the processor returns. The outputs are in the following order:
            - latents: The latents of the input image/video.
    """

    def __init__(self, output_names: List[str]):
        super().__init__()
        self.output_names = output_names
        assert len(self.output_names) == 1

    def forward(
        self,
        vae: AutoencoderKLCogVideoX,
        image: Optional[torch.Tensor] = None,
        video: Optional[torch.Tensor] = None,
        generator: Optional[torch.Generator] = None,
        compute_posterior: bool = True,
    ) -> Dict[str, torch.Tensor]:
        device = vae.device
        dtype = vae.dtype

        if image is not None:
            video = image.unsqueeze(1)

        assert video.ndim == 5, f"Expected 5D tensor, got {video.ndim}D tensor"
        video = video.to(device=device, dtype=vae.dtype)
        video = video.permute(0, 2, 1, 3, 4).contiguous()  # [B, F, C, H, W] -> [B, C, F, H, W]

        if compute_posterior:
            latents = vae.encode(video).latent_dist.sample(generator=generator)
            latents = latents.to(dtype=dtype)
        else:
            if vae.use_slicing and video.shape[0] > 1:
                encoded_slices = [vae._encode(x_slice) for x_slice in video.split(1)]
                moments = torch.cat(encoded_slices)
            else:
                moments = vae._encode(video)
            latents = moments.to(dtype=dtype)

        latents = latents.permute(0, 2, 1, 3, 4)  # [B, C, F, H, W] -> [B, F, C, H, W]
        return {self.output_names[0]: latents}


class CogVideoXModelSpecification(ModelSpecification):
    def __init__(
        self,
        pretrained_model_name_or_path: str = "THUDM/CogVideoX-5b",
        tokenizer_id: Optional[str] = None,
        text_encoder_id: Optional[str] = None,
        transformer_id: Optional[str] = None,
        vae_id: Optional[str] = None,
        text_encoder_dtype: torch.dtype = torch.bfloat16,
        transformer_dtype: torch.dtype = torch.bfloat16,
        vae_dtype: torch.dtype = torch.bfloat16,
        revision: Optional[str] = None,
        cache_dir: Optional[str] = None,
        condition_model_processors: List[ProcessorMixin] = None,
        latent_model_processors: List[ProcessorMixin] = None,
        **kwargs,
    ) -> None:
        super().__init__(
            pretrained_model_name_or_path=pretrained_model_name_or_path,
            tokenizer_id=tokenizer_id,
            text_encoder_id=text_encoder_id,
            transformer_id=transformer_id,
            vae_id=vae_id,
            text_encoder_dtype=text_encoder_dtype,
            transformer_dtype=transformer_dtype,
            vae_dtype=vae_dtype,
            revision=revision,
            cache_dir=cache_dir,
        )

        if condition_model_processors is None:
            condition_model_processors = [T5Processor(["encoder_hidden_states", "prompt_attention_mask"])]
        if latent_model_processors is None:
            latent_model_processors = [CogVideoXLatentEncodeProcessor(["latents"])]

        self.condition_model_processors = condition_model_processors
        self.latent_model_processors = latent_model_processors

    @property
    def _resolution_dim_keys(self):
        return {"latents": (1, 3, 4)}

    def load_condition_models(self) -> Dict[str, torch.nn.Module]:
        common_kwargs = {"revision": self.revision, "cache_dir": self.cache_dir}

        if self.tokenizer_id is not None:
            tokenizer = AutoTokenizer.from_pretrained(self.tokenizer_id, **common_kwargs)
        else:
            tokenizer = T5Tokenizer.from_pretrained(
                self.pretrained_model_name_or_path, subfolder="tokenizer", **common_kwargs
            )

        if self.text_encoder_id is not None:
            text_encoder = AutoModel.from_pretrained(
                self.text_encoder_id, torch_dtype=self.text_encoder_dtype, **common_kwargs
            )
        else:
            text_encoder = T5EncoderModel.from_pretrained(
                self.pretrained_model_name_or_path,
                subfolder="text_encoder",
                torch_dtype=self.text_encoder_dtype,
                **common_kwargs,
            )

        return {"tokenizer": tokenizer, "text_encoder": text_encoder}

    def load_latent_models(self) -> Dict[str, torch.nn.Module]:
        common_kwargs = {"revision": self.revision, "cache_dir": self.cache_dir}

        if self.vae_id is not None:
            vae = AutoencoderKLCogVideoX.from_pretrained(self.vae_id, torch_dtype=self.vae_dtype, **common_kwargs)
        else:
            vae = AutoencoderKLCogVideoX.from_pretrained(
                self.pretrained_model_name_or_path, subfolder="vae", torch_dtype=self.vae_dtype, **common_kwargs
            )

        return {"vae": vae}

    def load_diffusion_models(self) -> Dict[str, torch.nn.Module]:
        common_kwargs = {"revision": self.revision, "cache_dir": self.cache_dir}

        if self.transformer_id is not None:
            transformer = CogVideoXTransformer3DModel.from_pretrained(
                self.transformer_id, torch_dtype=self.transformer_dtype, **common_kwargs
            )
        else:
            transformer = CogVideoXTransformer3DModel.from_pretrained(
                self.pretrained_model_name_or_path,
                subfolder="transformer",
                torch_dtype=self.transformer_dtype,
                **common_kwargs,
            )

        scheduler = CogVideoXDDIMScheduler.from_pretrained(
            self.pretrained_model_name_or_path, subfolder="scheduler", **common_kwargs
        )

        return {"transformer": transformer, "scheduler": scheduler}

    def load_pipeline(
        self,
        tokenizer: Optional[T5Tokenizer] = None,
        text_encoder: Optional[T5EncoderModel] = None,
        transformer: Optional[CogVideoXTransformer3DModel] = None,
        vae: Optional[AutoencoderKLCogVideoX] = None,
        scheduler: Optional[CogVideoXDDIMScheduler] = None,
        enable_slicing: bool = False,
        enable_tiling: bool = False,
        enable_model_cpu_offload: bool = False,
        training: bool = False,
        **kwargs,
    ) -> CogVideoXPipeline:
        components = {
            "tokenizer": tokenizer,
            "text_encoder": text_encoder,
            "transformer": transformer,
            "vae": vae,
            "scheduler": scheduler,
        }
        components = get_non_null_items(components)

        pipe = CogVideoXPipeline.from_pretrained(
            self.pretrained_model_name_or_path, **components, revision=self.revision, cache_dir=self.cache_dir
        )
        pipe.text_encoder.to(self.text_encoder_dtype)
        pipe.vae.to(self.vae_dtype)

        _enable_vae_memory_optimizations(pipe.vae, enable_slicing, enable_tiling)
        if not training:
            pipe.transformer.to(self.transformer_dtype)
        if enable_model_cpu_offload:
            pipe.enable_model_cpu_offload()
        return pipe

    @torch.no_grad()
    def prepare_conditions(
        self,
        tokenizer: T5Tokenizer,
        text_encoder: T5EncoderModel,
        caption: str,
        max_sequence_length: int = 226,
        **kwargs,
    ) -> Dict[str, Any]:
        conditions = {
            "tokenizer": tokenizer,
            "text_encoder": text_encoder,
            "caption": caption,
            "max_sequence_length": max_sequence_length,
            **kwargs,
        }
        input_keys = set(conditions.keys())
        conditions = super().prepare_conditions(**conditions)
        conditions = {k: v for k, v in conditions.items() if k not in input_keys}
        conditions.pop("prompt_attention_mask", None)
        return conditions

    @torch.no_grad()
    def prepare_latents(
        self,
        vae: AutoencoderKLCogVideoX,
        image: Optional[torch.Tensor] = None,
        video: Optional[torch.Tensor] = None,
        generator: Optional[torch.Generator] = None,
        compute_posterior: bool = True,
        **kwargs,
    ) -> Dict[str, torch.Tensor]:
        conditions = {
            "vae": vae,
            "image": image,
            "video": video,
            "generator": generator,
            "compute_posterior": compute_posterior,
            **kwargs,
        }
        input_keys = set(conditions.keys())
        conditions = super().prepare_latents(**conditions)
        conditions = {k: v for k, v in conditions.items() if k not in input_keys}
        return conditions

    def forward(
        self,
        transformer: CogVideoXTransformer3DModel,
        scheduler: CogVideoXDDIMScheduler,
        condition_model_conditions: Dict[str, torch.Tensor],
        latent_model_conditions: Dict[str, torch.Tensor],
        sigmas: torch.Tensor,
        generator: Optional[torch.Generator] = None,
        compute_posterior: bool = True,
        **kwargs,
    ) -> Tuple[torch.Tensor, ...]:
        # Just hardcode for now. In Diffusers, we will refactor such that RoPE would be handled within the model itself.
        VAE_SPATIAL_SCALE_FACTOR = 8
        rope_base_height = self.transformer_config.sample_height * VAE_SPATIAL_SCALE_FACTOR
        rope_base_width = self.transformer_config.sample_width * VAE_SPATIAL_SCALE_FACTOR
        patch_size = self.transformer_config.patch_size
        patch_size_t = getattr(self.transformer_config, "patch_size_t", None)

        if compute_posterior:
            latents = latent_model_conditions.pop("latents")
        else:
            posterior = DiagonalGaussianDistribution(latent_model_conditions.pop("latents"), _dim=2)
            latents = posterior.sample(generator=generator)
            del posterior

        if not getattr(self.vae_config, "invert_scale_latents", False):
            latents = latents * self.vae_config.scaling_factor

        if patch_size_t is not None:
            latents = self._pad_frames(latents, patch_size_t)

        timesteps = (sigmas.flatten() * 1000.0).long()

        noise = torch.zeros_like(latents).normal_(generator=generator)
        noisy_latents = scheduler.add_noise(latents, noise, timesteps)

        batch_size, num_frames, num_channels, height, width = latents.shape
        ofs_emb = (
            None
            if getattr(self.transformer_config, "ofs_embed_dim", None) is None
            else latents.new_full((batch_size,), fill_value=2.0)
        )

        image_rotary_emb = (
            prepare_rotary_positional_embeddings(
                height=height * VAE_SPATIAL_SCALE_FACTOR,
                width=width * VAE_SPATIAL_SCALE_FACTOR,
                num_frames=num_frames,
                vae_scale_factor_spatial=VAE_SPATIAL_SCALE_FACTOR,
                patch_size=patch_size,
                patch_size_t=patch_size_t,
                attention_head_dim=self.transformer_config.attention_head_dim,
                device=transformer.device,
                base_height=rope_base_height,
                base_width=rope_base_width,
            )
            if self.transformer_config.use_rotary_positional_embeddings
            else None
        )

        latent_model_conditions["hidden_states"] = noisy_latents.to(latents)
        latent_model_conditions["image_rotary_emb"] = image_rotary_emb
        latent_model_conditions["ofs"] = ofs_emb

        velocity = transformer(
            **latent_model_conditions,
            **condition_model_conditions,
            timestep=timesteps,
            return_dict=False,
        )[0]
        # For CogVideoX, the transformer predicts the velocity. The denoised output is calculated by applying the same
        # code paths as scheduler.get_velocity(), which can be confusing to understand.
        pred = scheduler.get_velocity(velocity, noisy_latents, timesteps)
        target = latents

        return pred, target, sigmas

    def validation(
        self,
        pipeline: CogVideoXPipeline,
        prompt: str,
        image: Optional[Image] = None,
        height: Optional[int] = None,
        width: Optional[int] = None,
        num_frames: Optional[int] = None,
        num_inference_steps: int = 50,
        generator: Optional[torch.Generator] = None,
        **kwargs,
    ) -> List[ArtifactType]:
        # TODO(aryan): add support for more parameters
        if image is not None:
            pipeline = CogVideoXImageToVideoPipeline.from_pipe(pipeline)

        generation_kwargs = {
            "prompt": prompt,
            "image": image,
            "height": height,
            "width": width,
            "num_frames": num_frames,
            "num_inference_steps": num_inference_steps,
            "generator": generator,
            "return_dict": True,
            "output_type": "pil",
        }
        generation_kwargs = get_non_null_items(generation_kwargs)
        video = pipeline(**generation_kwargs).frames[0]
        return [VideoArtifact(value=video)]

    def _save_lora_weights(
        self,
        directory: str,
        transformer_state_dict: Optional[Dict[str, torch.Tensor]] = None,
        scheduler: Optional[SchedulerType] = None,
        metadata: Optional[Dict[str, str]] = None,
        *args,
        **kwargs,
    ) -> None:
        # TODO(aryan): this needs refactoring
        if transformer_state_dict is not None:
            CogVideoXPipeline.save_lora_weights(
                directory,
                transformer_state_dict,
                save_function=functools.partial(safetensors_torch_save_function, metadata=metadata),
                safe_serialization=True,
            )
        if scheduler is not None:
            scheduler.save_pretrained(os.path.join(directory, "scheduler"))

    def _save_model(
        self,
        directory: str,
        transformer: CogVideoXTransformer3DModel,
        transformer_state_dict: Optional[Dict[str, torch.Tensor]] = None,
        scheduler: Optional[SchedulerType] = None,
    ) -> None:
        # TODO(aryan): this needs refactoring
        if transformer_state_dict is not None:
            with init_empty_weights():
                transformer_copy = CogVideoXTransformer3DModel.from_config(transformer.config)
            transformer_copy.load_state_dict(transformer_state_dict, strict=True, assign=True)
            transformer_copy.save_pretrained(os.path.join(directory, "transformer"))
        if scheduler is not None:
            scheduler.save_pretrained(os.path.join(directory, "scheduler"))

    @staticmethod
    def _pad_frames(latents: torch.Tensor, patch_size_t: int) -> torch.Tensor:
        num_frames = latents.size(1)
        additional_frames = patch_size_t - (num_frames % patch_size_t)
        if additional_frames > 0:
            last_frame = latents[:, -1:]
            padding_frames = last_frame.expand(-1, additional_frames, -1, -1, -1)
            latents = torch.cat([latents, padding_frames], dim=1)
        return latents