Sinusoid position encoding table

Computes spatio-temporal features from given features. Parameters: features (torch.Tensor): Input features to process. Returns:

Creates a dictionary of task objects from either a name of task, config, or prepared Task object. :param task_name_list: List[Union[str, Dict, Ta

Whether the task has a training set

Whiten values with masked values.

Given a list of sources, each is a conversation list. This transform: 1. Add signal '### ' at the beginning each sentence, with end signal '\

Given a list of sources, each is a conversation list. This transform: 1. Add signal '### ' at the beginning each sentence, with end signal '\

Take a single document and the LMM results and evaluates, returning a dict where keys are the names of submetrics and values are the values of

Perform per-sample random masking by per-sample shuffling. Per-sample shuffling is done by argsort random noise. x: [N, L, D]

Register fast dLLM cache hooks to the model. Args: model: The LLaDA model to register hooks to Returns: Fas

Rotates half the hidden dims of the input.

Rotates half the hidden dims of the input.

Takes possible nested list and recursively converts all inner component to strings

Given the task name, returns a sanitized version of it.

Saves all of the model weights

Predict the sample at the previous timestep by reversing the SDE. Core function to propagate the diffusion process from the learned model ou

Selects the best resolution from a list of possible resolutions based on the original size. Args: original_size (tuple): The origina

Sets the progress bar config for the pipeline

Print the colour legend.

Parses something like args1=val1,arg2=val2 Into a dictionary

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

r""" Returns the state dictionary of the model. We add the state dictionary of the value head to the state dictionary of the wrapped m

Perform a single step of training. Args: epoch (int): The current epoch. global_step (int): The current glob

Check if the current generation sequence has finished.

Check if the current generation sequences have finished.

Returns the text encoder used for encoding text inputs

:return: Iterable[obj] A iterable of any object, that doc_to_text can handle

Unpads a PyTorch tensor of a padded and resized image. Args: tensor (torch.Tensor): The image tensor, assumed to be in CxHxW format.

Check whether the text violates OpenAI moderation API.

regex_pattern: The basic regex pattern to use. If fails to match, we will use the customized match procedure - step 1

Calculates the running mean and standard deviation of a data stream. Reference: https://github.com/OpenLMLab/MOSS-RLHF/blob/40b91eb2f

Add Gumbel noise for categorical sampling.

Add speaker and start/end signal on each round.

Add speaker and start/end signal on each round.

Apply rotary position embedding to query and key tensors.

helper function solely used in validating get_task_dict output. Takes the output of lmms_eval.evaluator_utils.get_subtask_list and returns a l

Create a sliced version of fast_dllm_cache for block processing.

Create fast attention forward method with caching support.