Updates the cache with the new `key_states` and `value_states` for the layer `layer_idx`. Parameters: key_states (`torch

Get the encoded token span corresponding to a word in a sequence of the batch. Token spans are returned as a [`~tokenization_utils_b

Creates an index Args: indices (`torch.LongTensor`, same shape as a *values* Tensor to which the indices refer):

r""" If you are not familiar with adapters and PEFT methods, we invite you to read more about them on the PEFT official documentation:

Very simple data collator that simply collates batches of dict-like objects and performs special handling for potential keys named:

r""" Instantiate a Bark processor associated with a pretrained model. Args: pretrained_model_name_or_path (`str` or `os.P

This is the gelu implementation from the original ESM repo. Using F.gelu yields subtly wrong results.

Get environment device properties.

Initializes a video session for inference. If a video is provided (async inference), the video will be processed and stored on the `v

Checks if `mlu` is available via an `cndev-based` check which won't trigger the drivers and leave mlu uninitialized.

Checks if `torch_musa` is installed and potentially if a MUSA is in the environment

Converts the output of [`DPTForSemanticSegmentation`] into semantic segmentation maps. Args: outputs ([`DPTForSemanticSe

Rescales `image` by `scale`. Args: image (`np.ndarray`): The image to rescale. scale (`float`): The

Converts `image` to a PIL Image. Optionally rescales it and puts the channel dimension back as the last axis if needed. Args: im

Converts `image` to a PIL Image. Optionally rescales it and puts the channel dimension back as the last axis if needed. Args

Updates the cache with the new `key_states` and `value_states` for the layer `layer_idx`. Parameters: key_states (`torch

Get the character span in the original string corresponding to given word in a sequence of the batch. Character spans are returned a

The Mask2Former Loss. The loss is computed very similar to DETR. The process happens in two steps: 1) we compute hungarian assignment

Returns whether this model can generate sequences with `.generate()` from the `GenerationMixin`. Under the hood, on classes where th

Computes the output image size given the input image size and the desired output size with multiple of divisible_size. Args: image_s

Returns a boolean indicating whether a pair of generate outputs are similar. Two `generate` call outputs are considered similar in the follow

Whether or not the current process is the local process, based on `xr.global_ordinal()` (for TPUs) first, then on `local_rank`.

Prepares a sequence of input id, entity id and entity span, or a pair of sequences of inputs ids, entity ids, entity spans so that it

Verify that a tensor is synchronized (or not synchronized) across all processes in the mesh's process group. Handles both regular tensors and

Basic Tokenization of a piece of text. Split on "white spaces" only, for sub-word tokenization, see WordPieceTokenizer. Args

Verify that out_indices and out_features are valid for the given stage_names.