fastvideo.v1.fastvideo_args

`fastvideo.v1.fastvideo_args`#

The arguments of FastVideo Inference.

Module Contents#

Classes#

`FastVideoArgs`
`TrainingArgs`	Training arguments. Inherits from FastVideoArgs and adds training-specific arguments. If there are any conflicts, the training arguments will take precedence.

Functions#

`get_current_fastvideo_args`
`postprocess_text`
`prepare_fastvideo_args`	Prepare the inference arguments from the command line arguments.
`preprocess_text`
`set_current_fastvideo_args`	Temporarily set the current fastvideo config. Used during model initialization. We save the current fastvideo config in a global variable, so that all modules can access it, e.g. custom ops can access the fastvideo config to determine how to dispatch.

Data#

logger

API#

class fastvideo.v1.fastvideo_args.FastVideoArgs[source]#

DEFAULT_TEXT_ENCODER_PRECISIONS[source]#: (‘fp16’, ‘fp16’)

static add_cli_args(parser: fastvideo.v1.utils.FlexibleArgumentParser) → fastvideo.v1.utils.FlexibleArgumentParser[source]#

cache_strategy: str[source]#: ‘none’

check_fastvideo_args() → None[source]#: Validate inference arguments for consistency

device[source]#: None

device_str: Optional[str][source]#: None

disable_autocast: bool[source]#: False

dist_timeout: Optional[int][source]#: None

distributed_executor_backend: str[source]#: ‘mp’

dit_config: fastvideo.v1.configs.models.DiTConfig[source]#: ‘field(…)’

embedded_cfg_scale: float[source]#: 6.0

enable_torch_compile: bool[source]#: False

flow_shift: Optional[float][source]#: None

classmethod from_cli_args(args: argparse.Namespace) → fastvideo.v1.fastvideo_args.FastVideoArgs[source]#

image_encoder_config: fastvideo.v1.configs.models.EncoderConfig[source]#: ‘field(…)’

image_encoder_precision: str[source]#: ‘fp32’

inference_mode: bool[source]#: True

log_level: str[source]#: ‘info’

mask_strategy_file_path: Optional[str][source]#: None

model_path: str[source]#: None

neg_magic: Optional[str][source]#: None

num_gpus: int[source]#: 1

output_type: str[source]#: ‘pil’

pos_magic: Optional[str][source]#: None

postprocess_text_funcs: Tuple[Callable[[Any], Any], ...][source]#: ‘field(…)’

precision: str[source]#: ‘bf16’

preprocess_text_funcs: Tuple[Callable[[str], str], ...][source]#: ‘field(…)’

revision: Optional[str][source]#: None

sp_size: Optional[int][source]#: None

text_encoder_configs: Tuple[fastvideo.v1.configs.models.EncoderConfig, ...][source]#: ‘field(…)’

text_encoder_precisions: Tuple[str, ...][source]#: ‘field(…)’

timesteps_scale: Optional[bool][source]#: None

tp_size: Optional[int][source]#: None

property training_mode: bool[source]#

trust_remote_code: bool[source]#: False

use_cpu_offload: bool[source]#: False

vae_config: fastvideo.v1.configs.models.VAEConfig[source]#: ‘field(…)’

vae_precision: str[source]#: ‘fp16’

vae_sp: bool[source]#: False

vae_tiling: bool[source]#: True

class fastvideo.v1.fastvideo_args.TrainingArgs[source]#

Bases: fastvideo.v1.fastvideo_args.FastVideoArgs

Training arguments. Inherits from FastVideoArgs and adds training-specific arguments. If there are any conflicts, the training arguments will take precedence.

static add_cli_args(parser: fastvideo.v1.utils.FlexibleArgumentParser) → fastvideo.v1.utils.FlexibleArgumentParser[source]#

allow_tf32: bool[source]#: False

cache_dir: str = <Multiline-String>[source]#

cfg: float[source]#: 0.0

checkpointing_steps: int[source]#: 0

checkpoints_total_limit: int[source]#: 0

data_path: str = <Multiline-String>[source]#

dataloader_num_workers: int[source]#: 0

distill_cfg: float[source]#: 0.0

dit_model_name_or_path: str = <Multiline-String>[source]#

ema_decay: float[source]#: 0.0

ema_start_step: int[source]#: 0

classmethod from_cli_args(args: argparse.Namespace) → fastvideo.v1.fastvideo_args.TrainingArgs[source]#

fsdp_sharding_startegy: str = <Multiline-String>[source]#

gradient_accumulation_steps: int[source]#: 0

gradient_checkpointing: bool[source]#: False

group_frame: bool[source]#: False

group_resolution: bool[source]#: False

hunyuan_teacher_disable_cfg: bool[source]#: False

learning_rate: float[source]#: 0.0

linear_quadratic_threshold: float[source]#: 0.0

linear_range: float[source]#: 0.0

log_validation: bool[source]#: False

logging_dir: str = <Multiline-String>[source]#

logit_mean: float[source]#: 0.0

logit_std: float[source]#: 1.0

lr_num_cycles: int[source]#: 0

lr_power: float[source]#: 0.0

lr_scheduler: str = <Multiline-String>[source]#

lr_warmup_steps: int[source]#: 0

master_weight_type: str = <Multiline-String>[source]#

max_grad_norm: float[source]#: 0.0

max_train_steps: int[source]#: 0

mixed_precision: str = <Multiline-String>[source]#

mode_scale: float[source]#: 0.0

multi_phased_distill_schedule: str = <Multiline-String>[source]#

not_apply_cfg_solver: bool[source]#: False

num_euler_timesteps: int[source]#: 0

num_frames: int[source]#: 0

num_height: int[source]#: 0

num_latent_t: int[source]#: 0

num_train_epochs: int[source]#: 0

num_width: int[source]#: 0

output_dir: str = <Multiline-String>[source]#

precondition_outputs: bool[source]#: False

pred_decay_type: str = <Multiline-String>[source]#

pred_decay_weight: float[source]#: 0.0

pretrained_model_name_or_path: str = <Multiline-String>[source]#

resume_from_checkpoint: bool[source]#: False

scale_lr: bool[source]#: False

scheduler_type: str = <Multiline-String>[source]#

seed: Optional[int][source]#: None

selective_checkpointing: float[source]#: 0.0

tracker_project_name: str = <Multiline-String>[source]#

train_batch_size: int[source]#: 0

train_sp_batch_size: int[source]#: 0

use_ema: bool[source]#: False

validation_guidance_scale: str = <Multiline-String>[source]#

validation_prompt_dir: str = <Multiline-String>[source]#

validation_sampling_steps: str = <Multiline-String>[source]#

validation_steps: float[source]#: 0.0

weight_decay: float[source]#: 0.0

weighting_scheme: str = <Multiline-String>[source]#

fastvideo.v1.fastvideo_args.get_current_fastvideo_args() → fastvideo.v1.fastvideo_args.FastVideoArgs[source]#

fastvideo.v1.fastvideo_args.logger[source]#: ‘init_logger(…)’

fastvideo.v1.fastvideo_args.postprocess_text(output: Any) → Any[source]#

fastvideo.v1.fastvideo_args.prepare_fastvideo_args(argv: List[str]) → fastvideo.v1.fastvideo_args.FastVideoArgs[source]#

Prepare the inference arguments from the command line arguments.

Parameters:: argv – The command line arguments. Typically, it should be sys.argv[1:] to ensure compatibility with parse_args when no arguments are passed.
Returns:: The inference arguments.

fastvideo.v1.fastvideo_args.preprocess_text(prompt: str) → str[source]#

fastvideo.v1.fastvideo_args.set_current_fastvideo_args(fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs)[source]#: Temporarily set the current fastvideo config. Used during model initialization. We save the current fastvideo config in a global variable, so that all modules can access it, e.g. custom ops can access the fastvideo config to determine how to dispatch.