Return the best torch backend available.

Loads the class's state. Args: state_dict (dict): scaler state. Should be an object returned from a call to :meth:

Replace a module in the chain with a new module.

Instantiate an object from a config dict.

Load model from file. Args: path (str): file path

This method must be called once for each frame even with empty detections. NOTE:as in practical realtime MOT, the detector doesn't run

Load the LoRA weights from a dictionary of tensors. Expects the keys to be in the commonly found formats on CivitAI's hub.

Args: box: can be a k-tuple, k-list or an Nxk array/tensor, where k = 4 or 5 from_mode, to_mode (BoxMode) Re

Gets the contents of a url, but caches the response indefinitely. While we attempt to use the cached_path from huggingface transformers, we

Create a 1D, 2D, or 3D convolution module.

Generates images based on the provided prompts using the ImaginAIry API. Args: prompts (list[ImaginePrompt] | str | ImaginePrompt):

Retrieve the context object from the module's context provider.

Print a message if `verbose` is `True`.

Appends dimensions to the end of a tensor until it has target_dims dimensions.

Create a linear module.

Remove a module from the chain.

Create a 1D, 2D, or 3D convolution module.

Converts a denoiser output to a Karras ODE derivative.

Returns the state of the class as a :class:`dict`.