SYMBOL INDEX (259 symbols across 29 files)

FILE: hallo/animate/face_animate.py
  class FaceAnimatePipelineOutput (line 46) | class FaceAnimatePipelineOutput(BaseOutput):
  class FaceAnimatePipeline (line 58) | class FaceAnimatePipeline(DiffusionPipeline):
    method __init__ (line 90) | def __init__(
    method _execution_device (line 124) | def _execution_device(self):
    method prepare_latents (line 136) | def prepare_latents(
    method prepare_extra_step_kwargs (line 190) | def prepare_extra_step_kwargs(self, generator, eta):
    method decode_latents (line 222) | def decode_latents(self, latents):
    method __call__ (line 250) | def __call__(

FILE: hallo/animate/face_animate_static.py
  class StaticPipelineOutput (line 65) | class StaticPipelineOutput(BaseOutput):
  class StaticPipeline (line 76) | class StaticPipeline(DiffusionPipeline):
    method __init__ (line 86) | def __init__(
    method enable_vae_slicing (line 124) | def enable_vae_slicing(self):
    method disable_vae_slicing (line 132) | def disable_vae_slicing(self):
    method enable_sequential_cpu_offload (line 143) | def enable_sequential_cpu_offload(self, gpu_id=0):
    method _execution_device (line 157) | def _execution_device(self):
    method decode_latents (line 169) | def decode_latents(self, latents):
    method prepare_extra_step_kwargs (line 194) | def prepare_extra_step_kwargs(self, generator, eta):
    method prepare_latents (line 226) | def prepare_latents(
    method prepare_condition (line 278) | def prepare_condition(
    method __call__ (line 313) | def __call__(

FILE: hallo/datasets/audio_processor.py
  class AudioProcessor (line 22) | class AudioProcessor:
    method __init__ (line 37) | def __init__(
    method preprocess (line 76) | def preprocess(self, wav_file: str, clip_length: int=-1):
    method get_embedding (line 131) | def get_embedding(self, wav_file: str):
    method close (line 167) | def close(self):
    method __enter__ (line 173) | def __enter__(self):
    method __exit__ (line 176) | def __exit__(self, _exc_type, _exc_val, _exc_tb):

FILE: hallo/datasets/image_processor.py
  class ImageProcessor (line 25) | class ImageProcessor:
    method __init__ (line 53) | def __init__(self, img_size, face_analysis_model_path) -> None:
    method preprocess (line 107) | def preprocess(self, source_image_path: str, cache_dir: str, face_regi...
    method close (line 184) | def close(self):
    method _augmentation (line 198) | def _augmentation(self, images, transform, state=None):
    method __enter__ (line 208) | def __enter__(self):
    method __exit__ (line 211) | def __exit__(self, _exc_type, _exc_val, _exc_tb):
  class ImageProcessorForDataProcessing (line 215) | class ImageProcessorForDataProcessing():
    method __init__ (line 243) | def __init__(self, face_analysis_model_path, landmark_model_path, step...
    method preprocess (line 265) | def preprocess(self, source_image_path: str):
    method close (line 318) | def close(self):
    method _augmentation (line 332) | def _augmentation(self, images, transform, state=None):
    method __enter__ (line 342) | def __enter__(self):
    method __exit__ (line 345) | def __exit__(self, _exc_type, _exc_val, _exc_tb):

FILE: hallo/datasets/mask_image.py
  class FaceMaskDataset (line 21) | class FaceMaskDataset(Dataset):
    method __init__ (line 40) | def __init__(
    method augmentation (line 78) | def augmentation(self, image, transform, state=None):
    method __getitem__ (line 94) | def __getitem__(self, index):
    method __len__ (line 143) | def __len__(self):

FILE: hallo/datasets/talk_video.py
  class TalkingVideoDataset (line 83) | class TalkingVideoDataset(Dataset):
    method __init__ (line 106) | def __init__(
    method augmentation (line 175) | def augmentation(self, images, transform, state=None):
    method __getitem__ (line 201) | def __getitem__(self, index):
    method __len__ (line 315) | def __len__(self):

FILE: hallo/models/attention.py
  class GatedSelfAttentionDense (line 29) | class GatedSelfAttentionDense(nn.Module):
    method __init__ (line 40) | def __init__(self, query_dim: int, context_dim: int, n_heads: int, d_h...
    method forward (line 57) | def forward(self, x: torch.Tensor, objs: torch.Tensor) -> torch.Tensor:
  class BasicTransformerBlock (line 79) | class BasicTransformerBlock(nn.Module):
    method __init__ (line 114) | def __init__(
    method set_chunk_feed_forward (line 242) | def set_chunk_feed_forward(self, chunk_size: Optional[int], dim: int =...
    method forward (line 257) | def forward(
  class TemporalBasicTransformerBlock (line 410) | class TemporalBasicTransformerBlock(nn.Module):
    method __init__ (line 429) | def __init__(
    method forward (line 542) | def forward(
  class AudioTemporalBasicTransformerBlock (line 621) | class AudioTemporalBasicTransformerBlock(nn.Module):
    method __init__ (line 643) | def __init__(
    method forward (line 784) | def forward(
  function zero_module (line 909) | def zero_module(module):

FILE: hallo/models/audio_proj.py
  class AudioProjModel (line 40) | class AudioProjModel(ModelMixin):
    method __init__ (line 68) | def __init__(
    method forward (line 96) | def forward(self, audio_embeds):

FILE: hallo/models/face_locator.py
  class FaceLocator (line 34) | class FaceLocator(ModelMixin):
    method __init__ (line 60) | def __init__(
    method forward (line 94) | def forward(self, conditioning):

FILE: hallo/models/image_proj.py
  class ImageProjModel (line 23) | class ImageProjModel(ModelMixin):
    method __init__ (line 40) | def __init__(
    method forward (line 56) | def forward(self, image_embeds):

FILE: hallo/models/motion_module.py
  function zero_module (line 68) | def zero_module(module):
  class TemporalTransformer3DModelOutput (line 83) | class TemporalTransformer3DModelOutput(BaseOutput):
    method get_sample_shape (line 92) | def get_sample_shape(self):
  function get_motion_module (line 102) | def get_motion_module(in_channels, motion_module_type: str, motion_modul...
  class VanillaTemporalModule (line 126) | class VanillaTemporalModule(nn.Module):
    method __init__ (line 142) | def __init__(
    method forward (line 174) | def forward(
  class TemporalTransformer3DModel (line 200) | class TemporalTransformer3DModel(nn.Module):
    method __init__ (line 220) | def __init__(
    method forward (line 270) | def forward(self, hidden_states, encoder_hidden_states=None):
  class TemporalTransformerBlock (line 319) | class TemporalTransformerBlock(nn.Module):
    method __init__ (line 337) | def __init__(
    method forward (line 387) | def forward(
  class PositionalEncoding (line 426) | class PositionalEncoding(nn.Module):
    method __init__ (line 435) | def __init__(self, d_model, dropout=0.0, max_len=24):
    method forward (line 447) | def forward(self, x):
  class VersatileAttention (line 464) | class VersatileAttention(Attention):
    method __init__ (line 473) | def __init__(
    method extra_repr (line 498) | def extra_repr(self):
    method set_use_memory_efficient_attention_xformers (line 507) | def set_use_memory_efficient_attention_xformers(
    method forward (line 553) | def forward(

FILE: hallo/models/mutual_self_attention.py
  function torch_dfs (line 19) | def torch_dfs(model: torch.nn.Module):
  class ReferenceAttentionControl (line 39) | class ReferenceAttentionControl:
    method __init__ (line 64) | def __init__(
    method register_reference_hooks (line 115) | def register_reference_hooks(
    method update (line 404) | def update(self, writer, dtype=torch.float16):
    method clear (line 456) | def clear(self):

FILE: hallo/models/resnet.py
  class InflatedConv3d (line 30) | class InflatedConv3d(nn.Conv2d):
    method forward (line 50) | def forward(self, x):
  class InflatedGroupNorm (line 69) | class InflatedGroupNorm(nn.GroupNorm):
    method forward (line 88) | def forward(self, x):
  class Upsample3D (line 104) | class Upsample3D(nn.Module):
    method __init__ (line 115) | def __init__(
    method forward (line 135) | def forward(self, hidden_states, output_size=None):
  class Downsample3D (line 188) | class Downsample3D(nn.Module):
    method __init__ (line 204) | def __init__(
    method forward (line 232) | def forward(self, hidden_states):
  class ResnetBlock3D (line 255) | class ResnetBlock3D(nn.Module):
    method __init__ (line 279) | def __init__(
    method forward (line 372) | def forward(self, input_tensor, temb):
  class Mish (line 415) | class Mish(torch.nn.Module):
    method forward (line 425) | def forward(self, hidden_states):

FILE: hallo/models/transformer_2d.py
  class Transformer2DModelOutput (line 51) | class Transformer2DModelOutput(BaseOutput):
  class Transformer2DModel (line 66) | class Transformer2DModel(ModelMixin, ConfigMixin):
    method __init__ (line 97) | def __init__(
    method _set_gradient_checkpointing (line 241) | def _set_gradient_checkpointing(self, module, value=False):
    method forward (line 245) | def forward(

FILE: hallo/models/transformer_3d.py
  class Transformer3DModelOutput (line 26) | class Transformer3DModelOutput(BaseOutput):
  class Transformer3DModel (line 38) | class Transformer3DModel(ModelMixin, ConfigMixin):
    method __init__ (line 47) | def __init__(
    method _set_gradient_checkpointing (line 143) | def _set_gradient_checkpointing(self, module, value=False):
    method forward (line 147) | def forward(

FILE: hallo/models/unet_2d_blocks.py
  function get_down_block (line 35) | def get_down_block(
  function get_up_block (line 132) | def get_up_block(
  class AutoencoderTinyBlock (line 215) | class AutoencoderTinyBlock(nn.Module):
    method __init__ (line 231) | def __init__(self, in_channels: int, out_channels: int, act_fn: str):
    method forward (line 248) | def forward(self, x: torch.FloatTensor) -> torch.FloatTensor:
  class UNetMidBlock2D (line 261) | class UNetMidBlock2D(nn.Module):
    method __init__ (line 292) | def __init__(
    method forward (line 384) | def forward(
  class UNetMidBlock2DCrossAttn (line 407) | class UNetMidBlock2DCrossAttn(nn.Module):
    method __init__ (line 428) | def __init__(
    method forward (line 523) | def forward(
  class CrossAttnDownBlock2D (line 595) | class CrossAttnDownBlock2D(nn.Module):
    method __init__ (line 627) | def __init__(
    method forward (line 722) | def forward(
  class DownBlock2D (line 812) | class DownBlock2D(nn.Module):
    method __init__ (line 842) | def __init__(
    method forward (line 897) | def forward(
  class CrossAttnUpBlock2D (line 950) | class CrossAttnUpBlock2D(nn.Module):
    method __init__ (line 987) | def __init__(
    method forward (line 1079) | def forward(
  class UpBlock2D (line 1186) | class UpBlock2D(nn.Module):
    method __init__ (line 1217) | def __init__(
    method forward (line 1268) | def forward(

FILE: hallo/models/unet_2d_condition.py
  class UNet2DConditionOutput (line 80) | class UNet2DConditionOutput(BaseOutput):
  class UNet2DConditionModel (line 93) | class UNet2DConditionModel(ModelMixin, ConfigMixin, UNet2DConditionLoade...
    method __init__ (line 191) | def __init__(
    method attn_processors (line 703) | def attn_processors(self) -> Dict[str, AttentionProcessor]:
    method set_attn_processor (line 733) | def set_attn_processor(
    method set_default_attn_processor (line 774) | def set_default_attn_processor(self):
    method set_attention_slice (line 795) | def set_attention_slice(self, slice_size):
    method _set_gradient_checkpointing (line 866) | def _set_gradient_checkpointing(self, module, value=False):
    method enable_freeu (line 870) | def enable_freeu(self, s1, s2, b1, b2):
    method disable_freeu (line 894) | def disable_freeu(self):
    method forward (line 905) | def forward(
    method load_change_cross_attention_dim (line 1361) | def load_change_cross_attention_dim(

FILE: hallo/models/unet_3d.py
  class UNet3DConditionOutput (line 47) | class UNet3DConditionOutput(BaseOutput):
  class UNet3DConditionModel (line 59) | class UNet3DConditionModel(ModelMixin, ConfigMixin):
    method __init__ (line 121) | def __init__(
    method attn_processors (line 365) | def attn_processors(self) -> Dict[str, AttentionProcessor]:
    method set_attention_slice (line 395) | def set_attention_slice(self, slice_size):
    method _set_gradient_checkpointing (line 466) | def _set_gradient_checkpointing(self, module, value=False):
    method set_attn_processor (line 471) | def set_attn_processor(
    method forward (line 510) | def forward(
    method from_pretrained_2d (line 718) | def from_pretrained_2d(

FILE: hallo/models/unet_3d_blocks.py
  function get_down_block (line 26) | def get_down_block(
  function get_up_block (line 137) | def get_up_block(
  class UNetMidBlock3DCrossAttn (line 247) | class UNetMidBlock3DCrossAttn(nn.Module):
    method __init__ (line 283) | def __init__(
    method forward (line 407) | def forward(
  class CrossAttnDownBlock3D (line 497) | class CrossAttnDownBlock3D(nn.Module):
    method __init__ (line 509) | def __init__(
    method forward (line 638) | def forward(
  class DownBlock3D (line 783) | class DownBlock3D(nn.Module):
    method __init__ (line 812) | def __init__(
    method forward (line 884) | def forward(
  class CrossAttnUpBlock3D (line 940) | class CrossAttnUpBlock3D(nn.Module):
    method __init__ (line 969) | def __init__(
    method forward (line 1092) | def forward(
  class UpBlock3D (line 1238) | class UpBlock3D(nn.Module):
    method __init__ (line 1281) | def __init__(
    method forward (line 1347) | def forward(

FILE: hallo/models/wav2vec.py
  class Wav2VecModel (line 21) | class Wav2VecModel(Wav2Vec2Model):
    method forward (line 42) | def forward(
    method feature_extract (line 112) | def feature_extract(
    method encode (line 133) | def encode(
  function linear_interpolation (line 196) | def linear_interpolation(features, seq_len):

FILE: hallo/utils/config.py
  function filter_non_none (line 8) | def filter_non_none(dict_obj: Dict):

FILE: hallo/utils/util.py
  function seed_everything (line 84) | def seed_everything(seed):
  function import_filename (line 97) | def import_filename(filename):
  function delete_additional_ckpt (line 120) | def delete_additional_ckpt(base_path, num_keep):
  function save_videos_from_pil (line 154) | def save_videos_from_pil(pil_images, path, fps=8):
  function save_videos_grid (line 206) | def save_videos_grid(videos: torch.Tensor, path: str, rescale=False, n_r...
  function read_frames (line 244) | def read_frames(video_path):
  function get_fps (line 280) | def get_fps(video_path):
  function tensor_to_video (line 297) | def tensor_to_video(tensor, output_video_file, audio_source, fps=25):
  function compute_face_landmarks (line 332) | def compute_face_landmarks(detection_result, h, w):
  function get_landmark (line 351) | def get_landmark(file):
  function get_landmark_overframes (line 382) | def get_landmark_overframes(landmark_model, frames_path):
  function get_lip_mask (line 407) | def get_lip_mask(landmarks, height, width, out_path=None, expand_ratio=2...
  function get_union_lip_mask (line 433) | def get_union_lip_mask(landmarks, height, width, expand_ratio=1):
  function get_face_mask (line 451) | def get_face_mask(landmarks, height, width, out_path=None, expand_ratio=...
  function get_union_face_mask (line 479) | def get_union_face_mask(landmarks, height, width, expand_ratio=1):
  function get_mask (line 497) | def get_mask(file, cache_dir, face_expand_raio):
  function expand_region (line 529) | def expand_region(region, image_w, image_h, expand_ratio=1.0):
  function get_blur_mask (line 567) | def get_blur_mask(file_path, output_file_path, resize_dim=(64, 64), kern...
  function blur_mask (line 589) | def blur_mask(mask, resize_dim=(64, 64), kernel_size=(51, 51)):
  function get_background_mask (line 611) | def get_background_mask(file_path, output_file_path):
  function get_sep_face_mask (line 638) | def get_sep_face_mask(file_path1, file_path2, output_file_path):
  function resample_audio (line 668) | def resample_audio(input_audio_file: str, output_audio_file: str, sample...
  function get_face_region (line 676) | def get_face_region(image_path: str, detector):
  function save_checkpoint (line 707) | def save_checkpoint(model: torch.nn.Module, save_dir: str, prefix: str, ...
  function init_output_dir (line 771) | def init_output_dir(dir_list: List[str]):
  function load_checkpoint (line 784) | def load_checkpoint(cfg, save_dir, accelerator):
  function compute_snr (line 822) | def compute_snr(noise_scheduler, timesteps):
  function extract_audio_from_videos (line 854) | def extract_audio_from_videos(video_path: Path, audio_output_path: Path)...
  function convert_video_to_images (line 889) | def convert_video_to_images(video_path: Path, output_dir: Path) -> Path:
  function get_union_mask (line 923) | def get_union_mask(masks):
  function move_final_checkpoint (line 960) | def move_final_checkpoint(save_dir, module_dir, prefix):

FILE: scripts/app.py
  function predict (line 18) | def predict(image, audio, pose_weight, face_weight, lip_weight, face_exp...

FILE: scripts/data_preprocess.py
  function setup_directories (line 33) | def setup_directories(video_path: Path) -> dict:
  function process_single_video (line 59) | def process_single_video(video_path: Path,
  function process_all_videos (line 116) | def process_all_videos(input_video_list: List[Path], output_dir: Path, s...
  function get_video_paths (line 148) | def get_video_paths(source_dir: Path, parallelism: int, rank: int) -> Li...

FILE: scripts/extract_meta_info_stage1.py
  function collect_video_folder_paths (line 27) | def collect_video_folder_paths(root_path: Path) -> list:
  function construct_meta_info (line 40) | def construct_meta_info(frames_dir_path: Path) -> dict:
  function main (line 68) | def main():

FILE: scripts/extract_meta_info_stage2.py
  function get_video_paths (line 37) | def get_video_paths(root_path: Path, extensions: list) -> list:
  function file_exists (line 51) | def file_exists(file_path: str) -> bool:
  function construct_paths (line 64) | def construct_paths(video_path: str, base_dir: str, new_dir: str, new_ex...
  function extract_meta_info (line 80) | def extract_meta_info(video_path: str) -> dict:
  function main (line 155) | def main():

FILE: scripts/inference.py
  class Net (line 51) | class Net(nn.Module):
    method __init__ (line 62) | def __init__(
    method forward (line 77) | def forward(self,):
    method get_modules (line 82) | def get_modules(self):
  function process_audio_emb (line 95) | def process_audio_emb(audio_emb):
  function inference_process (line 118) | def inference_process(args: argparse.Namespace):

FILE: scripts/train_stage1.py
  class Net (line 68) | class Net(nn.Module):
    method __init__ (line 93) | def __init__(
    method forward (line 110) | def forward(
  function get_noise_scheduler (line 157) | def get_noise_scheduler(cfg: argparse.Namespace):
  function log_validation (line 181) | def log_validation(
  function train_stage1_process (line 289) | def train_stage1_process(cfg: argparse.Namespace) -> None:
  function load_config (line 765) | def load_config(config_path: str) -> dict:

FILE: scripts/train_stage2.py
  class Net (line 74) | class Net(nn.Module):
    method __init__ (line 104) | def __init__(
    method forward (line 123) | def forward(
  function get_attention_mask (line 182) | def get_attention_mask(mask: torch.Tensor, weight_dtype: torch.dtype) ->...
  function get_noise_scheduler (line 203) | def get_noise_scheduler(cfg: argparse.Namespace) -> Tuple[DDIMScheduler,...
  function process_audio_emb (line 228) | def process_audio_emb(audio_emb: torch.Tensor) -> torch.Tensor:
  function log_validation (line 250) | def log_validation(
  function train_stage2_process (line 421) | def train_stage2_process(cfg: argparse.Namespace) -> None:
  function load_config (line 962) | def load_config(config_path: str) -> dict: