Core I/O and media

Logical paths, local and remote URIs, serialization, image/video normalization, and tiled media processing.

On this page

Core I/O gives model integrations one contract for where data lives and how it is represented. Path helpers resolve logical WorldFoundry locations; storage helpers operate on local paths and supported URI schemes; serialization chooses a format explicitly or from a suffix; media helpers normalize common image/video inputs before model-specific preprocessing begins.

Resolve logical paths, do not hard-code hosts

worldfoundry_path_tokens computes roots for checkpoints, datasets, models, artifacts, caches, source repositories, and Conda environments. Explicit environment values win; otherwise the resolver uses predictable WorldFoundry cache or repository-adjacent defaults.

from worldfoundry.core.io.paths import (
    checkpoint_root_path,
    resolve_worldfoundry_path,
    worldfoundry_path_tokens,
)

env = {
    "WORLDFOUNDRY_HOME": "/srv/wf",
    "WORLDFOUNDRY_CKPT_DIR": "/models/checkpoints",
}

tokens = worldfoundry_path_tokens(env)
assert tokens["WORLDFOUNDRY_CKPT_DIR"] == "/models/checkpoints"
assert checkpoint_root_path("matrix-game-2", env=env) == \
    resolve_worldfoundry_path("${WORLDFOUNDRY_CKPT_DIR}/matrix-game-2", env)

Passing an env mapping makes path resolution testable without mutating the process environment. resolve_data_path is different: it points inside package-owned static data and should not be used for downloaded datasets.

Serialization round trip

from pathlib import Path
from tempfile import TemporaryDirectory

from worldfoundry.core import dump_serialized, load_serialized

with TemporaryDirectory() as directory:
    path = Path(directory) / "request.yaml"
    dump_serialized({"seed": 42, "actions": ["forward", "left"]}, path)
    payload = load_serialized(path)
    assert payload["seed"] == 42

When file_format is omitted, the suffix selects JSON, YAML, JSONL, pickle/gzip, NumPy, Torch, image, video, CSV/Pandas, or tar handling. When no file is supplied, text and binary formats return their serialized value. Pickle and unrestricted Torch checkpoints are executable formats; do not load them from an untrusted source.

Video shape boundary

coerce_video_frames is the normalization boundary for paths, Torch tensors, NumPy arrays, PIL frame lists, and tensor frame lists. It returns a uint8 NumPy array in T × H × W × C layout. video_tensor_to_uint8_frames handles normalized C × T × H × W or single-batch tensors and makes the value range explicit. read_video adds decoder metadata, while load_frames_from_video is for selected frame indices.

Use materialize_video_input when a subprocess or external runtime requires a local filename. Use TileProcessor only for codec/model functions that explicitly support overlapping spatiotemporal tiles; it changes execution layout but blends overlaps back into one output.

Complete reference

The blocks below are the generated signatures for this category. Use the on-page symbol index to jump; source links open the defining implementation behind each lazy export.

56 public symbols

def checkpoint_root_path(*parts: str | Path,specific_env: str | None = None,env: Mapping[str, str] | None = None) -> Path
worldfoundry.core.io.paths.checkpoint_root_pathfrom worldfoundry.core.io.paths import checkpoint_root_path
source

Overview

Resolves a target model checkpoint directory path with nested subfolders. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Source docstring

Resolves a target model checkpoint directory path with nested subfolders.

Avoids host-specific hardcoding by querying general and model-specific variables.

Parameters

partsstr | Path
Child path components appended to the resolved checkpoint root.
specific_envstr | None
Optional model-specific environment variable that takes precedence over `WORLDFOUNDRY_CKPT_DIR`.default: None
envMapping[str, str] | None
Environment mapping used instead of `os.environ`.default: None

Returns: PathResolved checkpoint path without creating it.

def coerce_video_frames(video_input)
worldfoundry.core.coerce_video_framesfrom worldfoundry.core import coerce_video_frames
source

Overview

Normalize common video inputs into a uint8 THWC numpy array. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

video_input
class Color()
worldfoundry.core.Colorfrom worldfoundry.core import Color
source

Overview

Small color helper compatible with legacy model-runtime config printers. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Methods

smethred(value)source

Overview

Public staticmethod on this type.

Parameters

value
smethgreen(value)source

Overview

Public staticmethod on this type.

Parameters

value
smethyellow(value)source

Overview

Public staticmethod on this type.

Parameters

value
smethblue(value)source

Overview

Public staticmethod on this type.

Parameters

value
smethmagenta(value)source

Overview

Public staticmethod on this type.

Parameters

value
smethcyan(value)source

Overview

Public staticmethod on this type.

Parameters

value
def copy_uri(src: str | os.PathLike[str],dst: str | os.PathLike[str],**storage_options) -> str
worldfoundry.core.copy_urifrom worldfoundry.core import copy_uri
source

Overview

Copy one URI to another using storage-aware byte streams. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str.

Parameters

srcstr | os.PathLike[str]
dststr | os.PathLike[str]
storage_options

Returns: str

def crop_and_resize(image: Image.Image,height: int,width: int) -> Image.Image
worldfoundry.core.crop_and_resizefrom worldfoundry.core import crop_and_resize
source

Overview

Center-crop and resize a PIL image to `(width, height). Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Image.Image`.

Parameters

imageImage.Image
heightint
widthint

Returns: Image.Image

def dump_serialized(obj: Any,file: str | Path | IO[Any] | None = None,file_format: str | None = None,encoding: str = 'utf-8',**kwargs: Any) -> str | bytes | None
worldfoundry.core.dump_serializedfrom worldfoundry.core import dump_serialized
source

Overview

Dump an object to a URI, file object, or string when `file is None. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str | bytes | None`.

Parameters

objAny
filestr | Path | IO[Any] | None
default: None
file_formatstr | None
default: None
encodingstr
default: 'utf-8'
kwargsAny

Returns: str | bytes | None

def exists_uri(uri: str | os.PathLike[str], **storage_options) -> bool
worldfoundry.core.exists_urifrom worldfoundry.core import exists_uri
source

Overview

Return whether a WorldFoundry URI (local or remote scheme) currently exists.

Parameters

uristr | os.PathLike[str]
storage_options

Returns: bool

def extract_frames_from_video_url(video_url: str)
worldfoundry.core.extract_frames_from_video_urlfrom worldfoundry.core import extract_frames_from_video_url
source

Overview

Decode a remote video URL and return RGB PIL frames. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

video_urlstr
def file_sha256(path: str | os.PathLike[str], , chunk_size: int = 1024  1024) -> str
worldfoundry.core.file_sha256from worldfoundry.core import file_sha256
source

Overview

Compute a file's SHA-256 digest without loading the whole artifact into memory. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str.

Parameters

pathstr | os.PathLike[str]
chunk_sizeint
default: 1024 * 1024

Returns: str

def get_video_details(video_path: str | Path) -> tuple[int, float, float]
worldfoundry.core.get_video_detailsfrom worldfoundry.core import get_video_details
source

Overview

Return `(total_frames, fps, duration_seconds) for a local video. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: tuple[int, float, float]`.

Parameters

video_pathstr | Path

Returns: tuple[int, float, float]

def hf_download_or_fpath(path: str | PathLike[str] | None) -> str | Any
worldfoundry.core.hf_download_or_fpathfrom worldfoundry.core import hf_download_or_fpath
source

Overview

Backwards-compatible alias for :func:resolve_hf_path. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str | Any.

Parameters

pathstr | PathLike[str] | None

Returns: str | Any

def hfd_root_path(parts: str | Path, env: Mapping[str, str] | None = None) -> Path
worldfoundry.core.io.paths.hfd_root_pathfrom worldfoundry.core.io.paths import hfd_root_path
source

Overview

Resolves the hfd-style local downloader checkpoint directory. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Parameters

partsstr | Path
envMapping[str, str] | None
default: None

Returns: Path

def infer_serialization_format(file: str | Path | IO[Any] | None,file_format: str | None = None) -> str
worldfoundry.core.infer_serialization_formatfrom worldfoundry.core import infer_serialization_format
source

Overview

Infer a normalized serialization format key. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str.

Parameters

filestr | Path | IO[Any] | None
file_formatstr | None
default: None

Returns: str

def is_remote_uri(uri: str | os.PathLike[str]) -> bool
worldfoundry.core.is_remote_urifrom worldfoundry.core import is_remote_uri
source

Overview

Return whether a path needs a non-local storage backend. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: bool.

Parameters

uristr | os.PathLike[str]

Returns: bool

def join_uri(base: str | os.PathLike[str], parts: str | os.PathLike[str]) -> str
worldfoundry.core.join_urifrom worldfoundry.core import join_uri
source

Overview

Join path components without losing remote URI schemes. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str.

Parameters

basestr | os.PathLike[str]
partsstr | os.PathLike[str]

Returns: str

def jsonable(value: Any) -> Any
worldfoundry.core.jsonablefrom worldfoundry.core import jsonable
source

Overview

Recursively convert common runtime objects into JSON-safe values. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Any.

Parameters

valueAny
Dataclass, `to_dict object, path, mapping, sequence, tensor/ array-like object with tolist`, callable, primitive, or fallback object.

Notes

Mapping keys are stringified, sets become lists, callables become a module/qualified-name record, and otherwise unsupported objects fall back to `repr`. This is evidence serialization, not a reversible object codec.

Returns: AnyA tree containing only JSON-compatible primitives and containers.

def load_frames_from_video(video_path: str | Path,indices: Iterable[int],video_decode_backend: str = 'decord',eval_: bool = True)
worldfoundry.core.load_frames_from_videofrom worldfoundry.core import load_frames_from_video
source

Overview

Load selected RGB frames into a torch tensor using decord or OpenCV. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

video_pathstr | Path
indicesIterable[int]
video_decode_backendstr
default: 'decord'
eval_bool
default: True
def load_serialized(file: str | Path | IO[Any],file_format: str | None = None,encoding: str = 'utf-8',**kwargs: Any) -> Any
worldfoundry.core.load_serializedfrom worldfoundry.core import load_serialized
source

Overview

Load a structured object from a URI or file object. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Any.

Parameters

filestr | Path | IO[Any]
file_formatstr | None
default: None
encodingstr
default: 'utf-8'
kwargsAny

Returns: Any

def load_video_frames(video_path: str | Path)
worldfoundry.core.load_video_framesfrom worldfoundry.core import load_video_frames
source

Overview

Decode a local or remote video into a uint8 THWC numpy array. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

video_pathstr | Path
def local_model_root_path(env: Mapping[str, str] | None = None) -> Path
worldfoundry.core.io.paths.local_model_root_pathfrom worldfoundry.core.io.paths import local_model_root_path
source

Overview

Resolves the root directory containing local model weights, configs, and adapters. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Parameters

envMapping[str, str] | None
default: None

Returns: Path

class LowMemoryImageFolder(folder, file_list = None)
worldfoundry.core.LowMemoryImageFolderfrom worldfoundry.core import LowMemoryImageFolder
source

Overview

Lazy image-folder reader that loads frames on demand. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

folder
file_list
default: None
class LowMemoryVideo(file_name: str)
worldfoundry.core.LowMemoryVideofrom worldfoundry.core import LowMemoryVideo
source

Overview

Lazy video reader that loads frames on demand. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

file_namestr
def materialize_hf_snapshot(repo_id_or_path: str | os.PathLike[str],revision: str | None = None,cache_dir: str | os.PathLike[str] | None = None,allow_patterns: str | Sequence[str] | None = None,ignore_patterns: str | Sequence[str] | None = None,required_files: Sequence[str] = (),local_files_only: bool | None = None,token: str | bool | None = None) -> Path
worldfoundry.core.materialize_hf_snapshotfrom worldfoundry.core import materialize_hf_snapshot
source

Overview

Return a local snapshot for either a path or Hugging Face repo id. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Source docstring

Return a local snapshot for either a path or Hugging Face repo id.

Remote downloads are serialized on rank zero and then reopened in local-only mode, avoiding partial-cache races in multi-GPU jobs.

Parameters

repo_id_or_pathstr | os.PathLike[str]
revisionstr | None
default: None
cache_dirstr | os.PathLike[str] | None
default: None
allow_patternsstr | Sequence[str] | None
default: None
ignore_patternsstr | Sequence[str] | None
default: None
required_filesSequence[str]
default: ()
local_files_onlybool | None
default: None
tokenstr | bool | None
default: None

Returns: Path

def materialize_video_input(video_input,output_dir: Optional[str] = None,filename: str = 'input.mp4',fps: int = 24) -> str
worldfoundry.core.materialize_video_inputfrom worldfoundry.core import materialize_video_input
source

Overview

Return a local video path, materializing in-memory inputs when needed. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str.

Parameters

video_input
output_dirOptional[str]
default: None
filenamestr
default: 'input.mp4'
fpsint
default: 24

Returns: str

def maybe_download_hf_repo_on_rank0(repo_id_or_path: str,revision: str | None = None,cache_dir: str | os.PathLike[str] | None = None,allow_patterns: str | Sequence[str] | None = None,ignore_patterns: str | Sequence[str] | None = None,token: str | bool | None = None) -> None
worldfoundry.core.maybe_download_hf_repo_on_rank0from worldfoundry.core import maybe_download_hf_repo_on_rank0
source

Overview

Download a remote HF repo snapshot from rank 0 when downloads are allowed. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: None.

Source docstring

Download a remote HF repo snapshot from rank 0 when downloads are allowed.

Local paths and explicit offline/local-only modes are no-ops. For remote repositories, rank 0 preloads the snapshot while other distributed ranks wait for its success/failure signal. A filesystem lock serializes independent processes that share the same HF cache directory.

Parameters

repo_id_or_pathstr
revisionstr | None
default: None
cache_dirstr | os.PathLike[str] | None
default: None
allow_patternsstr | Sequence[str] | None
default: None
ignore_patternsstr | Sequence[str] | None
default: None
tokenstr | bool | None
default: None

Returns: None

def merge_video_audio(video_path: str, audio_path: str) -> None
worldfoundry.core.merge_video_audiofrom worldfoundry.core import merge_video_audio
source

Overview

Merge video and audio with ffmpeg; overwrite `video_path on success. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: None`.

Parameters

video_pathstr
audio_pathstr

Returns: None

def open_uri(uri: str | os.PathLike[str],mode: str = 'rb',encoding: str = 'utf-8',**storage_options) -> Generator[BinaryIO | TextIO, None, None]
worldfoundry.core.open_urifrom worldfoundry.core import open_uri
source

Overview

Open a local or remote URI. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Generator[BinaryIO | TextIO, None, None].

Source docstring

Open a local or remote URI.

`fsspec` is used when available. Without it, local files are supported for all modes and HTTP(S) URLs are supported for reads.

Parameters

uristr | os.PathLike[str]
modestr
default: 'rb'
encodingstr
default: 'utf-8'
storage_options

Returns: Generator[BinaryIO | TextIO, None, None]

def package_module_root(package: str) -> Path
worldfoundry.core.package_module_rootfrom worldfoundry.core import package_module_root
source

Overview

Resolve the source directory for an importable package. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Parameters

packagestr

Returns: Path

class ParallelHelper()
worldfoundry.core.ParallelHelperfrom worldfoundry.core import ParallelHelper
source

Overview

Distribute video tiles by cost and reconstruct their global ordering. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Source docstring

Distribute video tiles by cost and reconstruct their global ordering.

In non-distributed execution every tile remains local. With a process group, larger tiles are round-robin balanced across ranks and decoded frame tensors are gathered with explicit count, dtype, and shape metadata.

Methods

smethsplit_tile_list(tile_numel_dict: OrderedDict[int, int],parallel_group: torch.distributed.ProcessGroup = None) -> List[int]source

Overview

Splits the given tile size into a list of sizes that each rank should handle.

Source docstring

Splits the given tile size into a list of sizes that each rank should handle.

This method takes into account the number of ranks in a distributed setting. If the distributed environment is not initialized, it returns a list of integers from 0 to tile_size - 1, representing each tile index.

If the distributed environment is initialized, it calculates the base tile size for each rank and distributes any remaining tiles among the ranks.

Parameters

tile_numel_dictOrderedDict[int, int]
Dict of index and numel of tiles.
parallel_grouptorch.distributed.ProcessGroup
Distributed decoding group. Defaults to None.default: None

Returns: List[int]List[int]: A list of tile indices assigned to the current rank. List[int]: A list of global tile indices.

smethgather_frames(frames: List[torch.Tensor],global_tile_idxs: List[int],parallel_group: torch.distributed.ProcessGroup = None) -> List[torch.Tensor]source

Overview

Gathers frame data from all ranks in a distributed environment.

Source docstring

Gathers frame data from all ranks in a distributed environment.

This method collects frames from all ranks and combines them into a single list. If the distributed environment is not initialized, it simply returns the input frames.

Parameters

framesList[torch.Tensor]
A list of frames (tensors) from the current rank.
global_tile_idxsList[int]
A list of global tile indices.
parallel_grouptorch.distributed.ProcessGroup
Distributed decoding group. Defaults to None.default: None

Returns: List[torch.Tensor]List[torch.Tensor]: A list of frames (tensors) from all ranks.

smethindex_undot(index: int, loop_size: List[int]) -> List[int]source

Overview

Converts a single index into a list of indices, representing the position in a multi-dimensional space.

Source docstring

Converts a single index into a list of indices, representing the position in a multi-dimensional space.

This method takes an integer index and a list of loop sizes, and converts the index into a list of indices that correspond to the position in a multi-dimensional space.

Parameters

indexint
The single index to be converted.
loop_sizeList[int]
A list of integers representing the size of each dimension in the multi-dimensional space.

Returns: List[int]List[int]: A list of integers representing the position in the multi-dimensional space.

smethindex_dot(index: List[int], loop_size: List[int]) -> intsource

Overview

Converts a list of indices into a single index, representing the position in a multi-dimensional space.

Source docstring

Converts a list of indices into a single index, representing the position in a multi-dimensional space.

This method takes a list of indices and a list of loop sizes, and converts the list of indices into a single index that corresponds to the position in a multi-dimensional space.

Parameters

indexList[int]
A list of integers representing the position in the multi-dimensional space.
loop_sizeList[int]
A list of integers representing the size of each dimension in the multi-dimensional space.

Returns: intint: A single integer representing the position in the multi-dimensional space.

def parse_uri_scheme(uri: str | os.PathLike[str]) -> str
worldfoundry.core.parse_uri_schemefrom worldfoundry.core import parse_uri_scheme
source

Overview

Return the lowercase URI scheme, using `file for local paths. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str`.

Parameters

uristr | os.PathLike[str]

Returns: str

def project_root(start: str | Path | None = None) -> Path
worldfoundry.core.io.paths.project_rootfrom worldfoundry.core.io.paths import project_root
source

Overview

Walks upward from a starting path to locate the root repository containing pyproject.toml. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Source docstring

Walks upward from a starting path to locate the root repository containing pyproject.toml.

This helper provides robust local development support, falling back to a package-relative root if executed from a system-wide python site-packages deployment.

Parameters

startstr | Path | None
File or directory from which to search upward. `None` starts from this module's installed source path.default: None

Returns: PathFirst ancestor containing `pyproject.toml`, or the package-relative fallback when no repository marker is found.

def read_binary_uri(uri: str | os.PathLike[str], **storage_options) -> bytes
worldfoundry.core.read_binary_urifrom worldfoundry.core import read_binary_uri
source

Overview

Read all bytes from a URI through Core’s storage helpers.

Parameters

uristr | os.PathLike[str]
storage_options

Returns: bytes

def read_image_as_video_tensor(image_path: str | Path,resolution: Iterable[int],num_video_frames: int,resize: bool = True)
worldfoundry.core.read_image_as_video_tensorfrom worldfoundry.core import read_image_as_video_tensor
source

Overview

Load an image and materialize a `[1,C,T,H,W]` uint8 conditioning video tensor. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

image_pathstr | Path
resolutionIterable[int]
num_video_framesint
resizebool
default: True
def read_text_uri(uri: str | os.PathLike[str],encoding: str = 'utf-8',**storage_options) -> str
worldfoundry.core.read_text_urifrom worldfoundry.core import read_text_uri
source

Overview

Read all text from a URI through Core’s storage helpers.

Parameters

uristr | os.PathLike[str]
encodingstr
default: 'utf-8'
storage_options

Returns: str

def read_video(video_path: str | Path, , return_metadata: bool = True)
worldfoundry.core.read_videofrom worldfoundry.core import read_video
source

Overview

Decode a video into frames and optional metadata. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

video_pathstr | Path
return_metadatabool
default: True
def resize_video_tensor_to_resolution(video_tensor, resolution: Iterable[int])
worldfoundry.core.resize_video_tensor_to_resolutionfrom worldfoundry.core import resize_video_tensor_to_resolution
source

Overview

Resize and center-crop a `[T,C,H,W] tensor to (target_h, target_w)`. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

video_tensor
resolutionIterable[int]
def resolve_data_path(parts: str | Path) -> Path
worldfoundry.core.io.paths.resolve_data_pathfrom worldfoundry.core.io.paths import resolve_data_path
source

Overview

Resolves a subpath under the internal worldfoundry/data asset directory. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Parameters

partsstr | Path

Returns: Path

def resolve_hf_path(path: str | PathLike[str] | None) -> str | Any
worldfoundry.core.resolve_hf_pathfrom worldfoundry.core import resolve_hf_path
source

Overview

Resolve a possibly `hf://-prefixed path to a local filesystem path. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str | Any`.

Source docstring

Resolve a possibly `hf://`-prefixed path to a local filesystem path.

Accepts either:

* a local path (returned unchanged if it exists), or * `hf://<owner>/<repo>[/<subpath>]` — resolves an already materialized WorldFoundry-local snapshot and returns the requested file or directory.

Runtime I/O is deliberately offline. Repository acquisition belongs to the explicit preparation workflow, never to model inference.

Parameters

pathstr | PathLike[str] | None

Returns: str | Any

def resolve_hf_snapshot_path(value: str | os.PathLike[str],required_files: Sequence[str] = (),local_files_only_env: str = 'WORLDFOUNDRY_HF_LOCAL_FILES_ONLY',local_files_only: bool | None = None) -> Path
worldfoundry.core.resolve_hf_snapshot_pathfrom worldfoundry.core import resolve_hf_snapshot_path
source

Overview

Resolve a repo id, HF cache repo root, or local path to a usable snapshot. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Parameters

valuestr | os.PathLike[str]
required_filesSequence[str]
default: ()
local_files_only_envstr
default: 'WORLDFOUNDRY_HF_LOCAL_FILES_ONLY'
local_files_onlybool | None
default: None

Returns: Path

def resolve_local_checkpoint_file(model_id_or_path: str | Path,filename: str,env: Mapping[str, str] | None = None) -> Path
worldfoundry.core.resolve_local_checkpoint_filefrom worldfoundry.core import resolve_local_checkpoint_file
source

Overview

Resolve one checkpoint file from explicit or WorldFoundry-local storage. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Source docstring

Resolve one checkpoint file from explicit or WorldFoundry-local storage.

An explicit file path is accepted directly. Directory paths and repository identifiers are resolved through :func:resolve_local_hf_model_path, so incomplete hfd transfers are rejected and this function never contacts a model hub.

Parameters

model_id_or_pathstr | Path
filenamestr
envMapping[str, str] | None
default: None

Returns: Path

def resolve_local_hf_model_path(model_id_or_path: str | Path,required_files: Sequence[str] = (),env: Mapping[str, str] | None = None) -> Path
worldfoundry.core.resolve_local_hf_model_pathfrom worldfoundry.core import resolve_local_hf_model_path
source

Overview

Resolve a Hugging Face model strictly from WorldFoundry-local storage. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Source docstring

Resolve a Hugging Face model strictly from WorldFoundry-local storage.

The resolver understands direct export directories, hfd-style names such as `owner--repo, and the Hub cache snapshots/<revision>` layout. It never contacts the network, which keeps model initialization deterministic and works in inference environments that do not install PyTorch.

Parameters

model_id_or_pathstr | Path
required_filesSequence[str]
default: ()
envMapping[str, str] | None
default: None

Returns: Path

def resolve_worldfoundry_path(value: str | Path,env: Mapping[str, str] | None = None) -> Path
worldfoundry.core.io.paths.resolve_worldfoundry_pathfrom worldfoundry.core.io.paths import resolve_worldfoundry_path
source

Overview

Expands structural WorldFoundry path tokens (e.g. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: Path.

Source docstring

Expands structural WorldFoundry path tokens (e.g. ${WORLDFOUNDRY_CKPT_DIR}) and home markers (~).

Performs precise regex-free variable mapping replacement while preserving subfolder hierarchies.

Parameters

valuestr | Path
Path containing optional `$NAME or ${NAME}` tokens.
envMapping[str, str] | None
Environment mapping used to build/override WorldFoundry tokens.default: None

Returns: PathExpanded `Path`. The target is not created or required to exist.

def save_frames(frames, save_path)
worldfoundry.core.save_framesfrom worldfoundry.core import save_frames
source

Overview

Write a sequence of PIL frames to numbered PNG files. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

frames
save_path
def save_image_or_video_tensor(tensor,save_path,fps: int = 24,quality: int | None = None,ffmpeg_params: list[str] | None = None,value_range: str | tuple[float, float] = 'auto',image_format: str = 'JPEG',video_format: str = 'mp4',**kwargs) -> str | None
worldfoundry.core.save_image_or_video_tensorfrom worldfoundry.core import save_image_or_video_tensor
source

Overview

Save a normalized `[C,T,H,W] or [B,C,T,H,W] tensor as image/video. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: str | None`.

Source docstring

Save a normalized `[C,T,H,W] or [B,C,T,H,W]` tensor as image/video.

A single-frame tensor is saved as an image; multi-frame tensors are saved as videos. Local paths and URI-like targets supported by `worldfoundry.core.io` storage helpers are both accepted.

Parameters

tensor
save_path
fpsint
default: 24
qualityint | None
default: None
ffmpeg_paramslist[str] | None
default: None
value_rangestr | tuple[float, float]
default: 'auto'
image_formatstr
default: 'JPEG'
video_formatstr
default: 'mp4'
kwargs

Returns: str | None

def save_video(frames,save_path,fps,quality = 9,ffmpeg_params = None)
worldfoundry.core.save_videofrom worldfoundry.core import save_video
source

Overview

Write a sequence of PIL frames to a video file. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

frames
save_path
fps
quality
default: 9
ffmpeg_params
default: None
def save_video_frames(video_frames,output_path: str | Path,fps: int = 16,**kwargs) -> None
worldfoundry.core.save_video_framesfrom worldfoundry.core import save_video_frames
source

Overview

Write a THWC uint8 frame array/list to a video path or URI. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: None.

Parameters

video_frames
output_pathstr | Path
fpsint
default: 16
kwargs

Returns: None

def save_video_with_audio(frames,save_path,audio_path,fps = 16,quality = 9,ffmpeg_params = None)
worldfoundry.core.save_video_with_audiofrom worldfoundry.core import save_video_with_audio
source

Overview

Write PIL frames, then mux an external audio track with ffmpeg. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

frames
Iterable of PIL images.
save_path
Destination video path, replaced after muxing.
audio_path
Existing audio file.
fps
Output frame rate.default: 16
quality
ImageIO encoder quality.default: 9
ffmpeg_params
Optional ImageIO ffmpeg arguments.default: None

Notes

`merge_video_audio` reports ffmpeg failures and removes its temporary file. Use lower-level helpers when the caller needs structured process error handling.

def search_for_images(folder)
worldfoundry.core.search_for_imagesfrom worldfoundry.core import search_for_images
source

Overview

Return naturally sorted JPG/PNG paths from one image-sequence folder. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Source docstring

Return naturally sorted JPG/PNG paths from one image-sequence folder.

Embedded digit runs are compared numerically, so `2.png sorts before 10.png`. The search is non-recursive.

Parameters

folder
class TileProcessor(encode_fn,decode_fn,tile_sample_min_height: int = 256,tile_sample_min_width: int = 256,tile_sample_min_length: int = 16,spatial_downsample_factor: int = 8,temporal_downsample_factor: int = 1,spatial_tile_overlap_factor: float = 0.25,temporal_tile_overlap_factor: float = 0,sr_ratio = 1,first_frame_as_image: bool = False,parallel_group: torch.distributed.ProcessGroup = None)
worldfoundry.core.TileProcessorfrom worldfoundry.core import TileProcessor
source

Overview

Encode or decode large videos through overlapping spatiotemporal tiles. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Source docstring

Encode or decode large videos through overlapping spatiotemporal tiles.

The processor derives latent tile sizes from codec downsample factors, schedules tiles locally or across `parallel_group, and linearly blends temporal/vertical/horizontal overlaps to suppress seams. encode_fn and decode_fn` remain model-owned callables.

Parameters

encode_fn
The encoding function used for tile sampling.
decode_fn
The decoding function used for tile reconstruction.
tile_sample_min_heightint
default: 256
tile_sample_min_widthint
default: 256
tile_sample_min_lengthint
The minimum length of the sampled tiles. Defaults to 16.default: 16
spatial_downsample_factorint
The actual spataial downsample factor of given encode_fn. Defaults to 8.default: 8
temporal_downsample_factorint
The actual temporal downsample factor of the latent space tiles. Defaults to 1.default: 1
spatial_tile_overlap_factorfloat
default: 0.25
temporal_tile_overlap_factorfloat
default: 0
sr_ratio
default: 1
first_frame_as_imagebool
default: False
parallel_grouptorch.distributed.ProcessGroup
Distributed decoding group. Defaults to None.default: None

Methods

methblend_t(a: torch.Tensor,b: torch.Tensor,blend_extent: int) -> torch.Tensorsource

Overview

Public method on this type.

Parameters

atorch.Tensor
btorch.Tensor
blend_extentint

Returns: torch.Tensor

methblend_v(a: torch.Tensor,b: torch.Tensor,blend_extent: int) -> torch.Tensorsource

Overview

Public method on this type.

Parameters

atorch.Tensor
btorch.Tensor
blend_extentint

Returns: torch.Tensor

methblend_h(a: torch.Tensor,b: torch.Tensor,blend_extent: int) -> torch.Tensorsource

Overview

Public method on this type.

Parameters

atorch.Tensor
btorch.Tensor
blend_extentint

Returns: torch.Tensor

methtiled_encode(x: torch.FloatTensor, verbose: bool = False)source

Overview

Public method on this type.

Parameters

xtorch.FloatTensor
verbosebool
default: False
methtiled_decode(z: torch.FloatTensor, verbose: bool = False)source

Overview

Public method on this type.

Parameters

ztorch.FloatTensor
verbosebool
default: False
def video_tensor_to_uint8_frames(video_tensor,value_range: str | tuple[float, float] = 'auto') -> 'object'
worldfoundry.core.video_tensor_to_uint8_framesfrom worldfoundry.core import video_tensor_to_uint8_frames
source

Overview

Convert a normalized torch video tensor to uint8 THWC frames. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: 'object'.

Source docstring

Convert a normalized torch video tensor to uint8 THWC frames.

`value_range="auto" treats tensors with negative values as [-1, 1] and non-negative floating tensors as [0, 1]`.

Parameters

video_tensor
value_rangestr | tuple[float, float]
default: 'auto'

Returns: 'object'

class VideoData(video_file = None,image_folder = None,height = None,width = None,**kwargs)
worldfoundry.core.VideoDatafrom worldfoundry.core import VideoData
source

Overview

Lazy video or image-folder dataset with optional resize. Belongs to Core I/O and media (paths, URIs, images, video, serialization).

Parameters

video_file
default: None
image_folder
default: None
height
default: None
width
default: None
kwargs

Methods

methraw_data()source

Overview

Public method on this type.

methset_length(length)source

Overview

Public method on this type.

Parameters

length
methset_shape(height, width)source

Overview

Public method on this type.

Parameters

height
width
methshape()source

Overview

Public method on this type.

methsave_images(folder)source

Overview

Public method on this type.

Parameters

folder
def worldfoundry_path_tokens(env: Mapping[str, str] | None = None) -> dict[str, str]
worldfoundry.core.io.paths.worldfoundry_path_tokensfrom worldfoundry.core.io.paths import worldfoundry_path_tokens
source

Overview

Generates the dictionary of logical path-token replacements used across the system. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: dict[str, str].

Source docstring

Generates the dictionary of logical path-token replacements used across the system.

Builds dynamic mappings for artifact, checkpoint, data, conda, and repo paths. Prioritizes explicit environment overrides (such as WORLDFOUNDRY_HOME or WORLDFOUNDRY_CACHE_DIR) and falls back to user-home cache directories when variables are unset.

Parameters

envMapping[str, str] | None
Environment mapping to resolve instead of `os.environ`. Passing a mapping makes resolution deterministic in tests.default: None

Returns: dict[str, str]Token name to expanded path-string mapping.

def write_binary_uri(uri: str | os.PathLike[str],data: bytes | bytearray | memoryview | io.BytesIO,**storage_options) -> None
worldfoundry.core.write_binary_urifrom worldfoundry.core import write_binary_uri
source

Overview

Write bytes to a URI, creating local parent directories as needed. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: None.

Parameters

uristr | os.PathLike[str]
databytes | bytearray | memoryview | io.BytesIO
storage_options

Returns: None

def write_text_uri(uri: str | os.PathLike[str],data: str,encoding: str = 'utf-8',**storage_options) -> None
worldfoundry.core.write_text_urifrom worldfoundry.core import write_text_uri
source

Overview

Write text to a URI, creating local parent directories as needed. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: None.

Parameters

uristr | os.PathLike[str]
datastr
encodingstr
default: 'utf-8'
storage_options

Returns: None

def write_video(video_frames,output_path: str | Path,fps: int = 16,quality: int | None = None,format: str | None = None,**kwargs) -> None
worldfoundry.core.write_videofrom worldfoundry.core import write_video
source

Overview

Write a THWC video array/list to a local path or remote URI. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: None.

Parameters

video_frames
output_pathstr | Path
fpsint
default: 16
qualityint | None
default: None
formatstr | None
default: None
kwargs

Returns: None

def write_video_torchvision(filename: str | Path,video_array: Any,fps: float,*args: Any,**kwargs: Any) -> None
worldfoundry.core.write_video_torchvisionfrom worldfoundry.core import write_video_torchvision
source

Overview

Write RGB video frames with a `torchvision.io.write_video-compatible signature. Belongs to Core I/O and media (paths, URIs, images, video, serialization). Annotated return type: None`.

Parameters

filenamestr | Path
video_arrayAny
fpsfloat
argsAny
kwargsAny

Returns: None