--- title: lingbotworld2 ---

fastvideo.configs.models.dits.lingbotworld2.DiTArchConfig

DiTArchConfig(stacked_params_mapping: list[tuple[str, str, str]] = list(), _fsdp_shard_conditions: list = list(), _compile_conditions: list = list(), param_names_mapping: dict = dict(), reverse_param_names_mapping: dict = dict(), lora_param_names_mapping: dict = dict(), cast_prompt_embeds_to_dit_dtype: bool = False, _supported_attention_backends: tuple[AttentionBackendEnum, ...] = (AttentionBackendEnum.SAGE_ATTN, AttentionBackendEnum.FLASH_ATTN, AttentionBackendEnum.TORCH_SDPA, AttentionBackendEnum.VIDEO_SPARSE_ATTN, AttentionBackendEnum.VMOBA_ATTN, AttentionBackendEnum.SAGE_ATTN_THREE, AttentionBackendEnum.ATTN_QAT_INFER, AttentionBackendEnum.ATTN_QAT_TRAIN, AttentionBackendEnum.SLA_ATTN, AttentionBackendEnum.SAGE_SLA_ATTN), hidden_size: int = 0, num_attention_heads: int = 0, num_channels_latents: int = 0, in_channels: int = 0, out_channels: int = 0, exclude_lora_layers: list[str] = list(), boundary_ratio: float | None = None)

fastvideo.configs.models.dits.lingbotworld2.DiTConfig

DiTConfig(arch_config: DiTArchConfig = DiTArchConfig(), prefix: str = '', quant_config: QuantizationConfig | None = None, *, _resolved_attention_backend: AttentionBackendEnum | None = None)

fastvideo.configs.models.dits.lingbotworld2.DiTConfig.add_cli_args

add_cli_args(parser: Any, prefix: str = 'dit-config') -> Any

Add CLI arguments for DiTConfig fields

fastvideo.configs.models.dits.lingbotworld2.is_blocks

is_blocks(n: str, m) -> bool
Source
def is_blocks(n: str, m) -> bool:
    return "blocks" in n and str.isdigit(n.split(".")[-1])

fastvideo.configs.models.dits.lingbotworld2.LingBotWorld2CausalFastArchConfig

LingBotWorld2CausalFastArchConfig(stacked_params_mapping: list[tuple[str, str, str]] = list(), _fsdp_shard_conditions: list = (lambda: [is_blocks])(), _compile_conditions: list = list(), param_names_mapping: dict = dict(), reverse_param_names_mapping: dict = dict(), lora_param_names_mapping: dict = dict(), cast_prompt_embeds_to_dit_dtype: bool = False, _supported_attention_backends: tuple[AttentionBackendEnum, ...] = (AttentionBackendEnum.SAGE_ATTN, AttentionBackendEnum.FLASH_ATTN, AttentionBackendEnum.TORCH_SDPA, AttentionBackendEnum.VIDEO_SPARSE_ATTN, AttentionBackendEnum.VMOBA_ATTN, AttentionBackendEnum.SAGE_ATTN_THREE, AttentionBackendEnum.ATTN_QAT_INFER, AttentionBackendEnum.ATTN_QAT_TRAIN, AttentionBackendEnum.SLA_ATTN, AttentionBackendEnum.SAGE_SLA_ATTN), hidden_size: int = 0, num_attention_heads: int = 0, num_channels_latents: int = 0, in_channels: int = 0, out_channels: int = 0, exclude_lora_layers: list[str] = list(), boundary_ratio: float | None = None, model_type: str = 'i2v', patch_size: tuple[int, int, int] = (1, 2, 2), text_len: int = 512, in_dim: int = 36, dim: int = 5120, ffn_dim: int = 13824, freq_dim: int = 256, text_dim: int = 4096, out_dim: int = 16, num_heads: int = 40, num_layers: int = 40, qk_norm: bool = True, cross_attn_norm: bool = True, eps: float = 1e-06, local_attn_size: int = 18, sink_size: int = 6, chunk_size: int = 4, sample_shift: float = 10.0, num_train_timesteps: int = 1000, timesteps_index: tuple[int, int, int, int] = (0, 250, 500, 750), max_area: int = 480 * 832)
Source
@dataclass
class LingBotWorld2CausalFastArchConfig(DiTArchConfig):
    _fsdp_shard_conditions: list = field(default_factory=lambda: [is_blocks])
    param_names_mapping: dict = field(default_factory=dict)
    reverse_param_names_mapping: dict = field(default_factory=dict)
    lora_param_names_mapping: dict = field(default_factory=dict)

    model_type: str = "i2v"
    patch_size: tuple[int, int, int] = (1, 2, 2)
    text_len: int = 512
    in_dim: int = 36
    dim: int = 5120
    ffn_dim: int = 13824
    freq_dim: int = 256
    text_dim: int = 4096
    out_dim: int = 16
    num_heads: int = 40
    num_layers: int = 40
    qk_norm: bool = True
    cross_attn_norm: bool = True
    eps: float = 1e-6

    local_attn_size: int = 18
    sink_size: int = 6
    chunk_size: int = 4
    sample_shift: float = 10.0
    num_train_timesteps: int = 1000
    timesteps_index: tuple[int, int, int, int] = (0, 250, 500, 750)
    max_area: int = 480 * 832

    def __post_init__(self):
        super().__post_init__()
        self.hidden_size = self.dim
        self.num_attention_heads = self.num_heads
        self.attention_head_dim = self.dim // self.num_heads
        self.in_channels = self.in_dim
        self.out_channels = self.out_dim
        self.num_channels_latents = self.out_dim

fastvideo.configs.models.dits.lingbotworld2.LingBotWorld2CausalFastVideoConfig

LingBotWorld2CausalFastVideoConfig(arch_config: DiTArchConfig = LingBotWorld2CausalFastArchConfig(), prefix: str = 'Wan', quant_config: QuantizationConfig | None = None, *, _resolved_attention_backend: AttentionBackendEnum | None = None)
Source
@dataclass
class LingBotWorld2CausalFastVideoConfig(DiTConfig):
    arch_config: DiTArchConfig = field(default_factory=LingBotWorld2CausalFastArchConfig)

    prefix: str = "Wan"