Videos

mlop supports logging many different video types to the platform, you can currently log media that are of the format of:

MP4 - MPEG-4 Part 14 container format
GIF - Graphics Interchange Format
WebM - Open web media format
M4A - MPEG-4 Audio format
OGG - Free, open container format

To use video logging you first need to instantiate the mlop.Video class

video = mlop.Video(
    data: Union[str, np.ndarray],
    rate: int | None = 30,
    caption: str | None = None,
    format: str | None = None,
)
mlop.log({"video/file/0-0": video})

Parameter	Type	Description
`data`	`Union[str np.ndarray]`	The video data to log. Can be a path to a video file or a NumPy array.
`rate`	`int`	The frame rate of the video. Defaults to 30.
`caption`	`str`	A caption for the video.
`format`	`str`	The format of the video. Defaults to `mp4`.

Supported Formats

mlop uses moviepy and imageio as its backend for video logging, with the following supported formats.

Examples

Logging Videos from File Paths

path = "video.mp4"
mlop.log({"video": mlop.Video(path)})

Logging Videos from NumPy Arrays

data = np.random.randint(low=0, high=256, size=(10, 3, 112, 112), dtype=np.uint8)
mlop.log({"video": mlop.Video(data)})

Videos

Examples

Logging Videos from File Paths

Logging Videos from NumPy Arrays

On this page