This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch, num_key_value_heads, seqlen, he

This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch, num_key_value_heads, seqlen, he

This method forwards all its arguments to Qwen2TokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please refer to the docstring o

Returns the smallest integer greater than or equal to 'number' that is divisible by 'factor'.

This method forwards all its arguments to Qwen2TokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to the docstring of this

Wraps modules and applies quantization noise to the weights for subsequent quantization with Iterative Product Quantization as described

Rotates half the hidden dims of the input.

Returns the closest integer to 'number' that is divisible by 'factor'.

Rescales the image so that the following conditions are met: 1. Both dimensions (height and width) are divisible by 'factor'. 2. The to

Creates a causal 4D mask of shape `(batch_size, 1, query_length, key_value_length)` from a 2D mask of shape `(batch_size, key_value_l

Creates a causal 4D mask of shape `(batch_size, 1, query_length, key_value_length)` from a 2D mask of shape `(batch_size, key_value_l

Applies Rotary Position Embedding with Multimodal Sections to the query and key tensors (https://qwenlm.github.io/blog/qwen2-vl/). Explanation:

Applies Rotary Position Embedding with Multimodal Sections to the query and key tensors (https://qwenlm.github.io/blog/qwen2-vl/). Explanation:

Disable the redundant torch default initialization to accelerate model creation.

Returns the largest integer less than or equal to 'number' that is divisible by 'factor'.

Return a list of indices so that each slice of `batch_size` consecutive indices correspond to elements of similar lengths. To do this, the in

Rotates half the hidden dims of the input.

Preprocess an image or batch of images. Copy of the `preprocess` method from `CLIPImageProcessor`. Args: images (`ImageI

Creates a dynamic module in the cache directory for modules.

Get the list of modules that are relatively imported in a module file. Args: module_file (`str` or `os.PathLike`): The module file t

Collects the state dict and dump to disk.

calculate the number of frames for video used for model inputs. Args: ele (dict): a dict contains the configuration of video.

Split a list of indices into `chunks` chunks of roughly equal lengths.

dynamic RoPE layers should recompute `inv_freq` in the following situations: 1 - growing beyond the cached sequence length (allow sca

dynamic RoPE layers should recompute `inv_freq` in the following situations: 1 - growing beyond the cached sequence length (allow sca

Get the number of images and videos for each sample to calculate the separation length of the sample tensor. These parameters are not