* @brief Width of a video frame in pixels */

Run one full epoch; return list of (jpg_header_bytes, cls) fingerprints.

Replaces the checkpoint state dictionary with a deserialized version. Parameters ---------- data : str A string p

Returns the wrapped `cudaStream_t` handle.

bbox_a and bbox_b are xy(z)XY(Z) format

* @brief Returns the order in which the data is accessed - it can be either host order * or a stream order (or unspecified). */

Register the data nodes as produced in current scope, otherwise if `global_scope` is True put them in the outermost scope.

Resets pipeline iterator If pipeline iterator reached the end then reset its state to the beginning.

* @brief Returns the type of the data in the input at given index. */

* @brief Utility to check for valid type */

! * \brief Deallocates the storage pointed to by `__p`. * * `__p` must have been returned by a prior call to `allocate(__bytes, * __alignm

Print information about the annotation file. :return:

Returns the next produced batch (in the order of schedule_batch calls) for the ``context_i``th callback. Parameters ---------

Create a new exception with a synthetic traceback and change the message of the initial one. Raise the old one with the changed type from the

Executes the operator immediately.

Distribute `work_batch` among workers. Parameters ---------- context_i : int Specifies which callback will be use

! * \brief Allocates storage of size at least `__bytes` bytes in stream order * on `__stream`. * * The returned storage is aligned to `def

* @name Shape access * @{ */ * @brief Get the number of samples this batch holds. */

Replaces placeholders in a Python template. AST Name and Tuple nodes always receive the context that inferred from the template. However, whe

Serializes the checkpoint state dictionary to a JSON string. Raises ------ RuntimeError If the state dictionary i

[span.obs], span observers

Wraps an existing object or creates a new stream. This function wraps a compatible stream object with a DALI Stream class or creates a new s

* @brief Return a span containing the shape of `sample` */

Convert a CHW ``torch.Tensor`` to a ``PIL.Image``. Mirrors ``torchvision.transforms.v2.functional.to_pil_image``. Parameters ------

* @brief Height of a video frame in pixels */

* @brief Synchronizes on the the stream where Tensor t was created */

Convert a DALI dynamic tensor to numpy array.

* @brief Returns the real size of the allocation */