Print information about the VQA annotation file. :return:

Besides loading the `consolidated.*.pth` model weights, this function also tries to find tokenizer, 'meta.json', and 'config.json' un

torch.repeat_interleave(x, dim=2, repeats=n_rep)

Extract values from a 1-D numpy array for a batch of indices. :param arr: the 1-D numpy array. :param timesteps: a tensor of indices into

Extract values from a 1-D numpy array for a batch of indices. :param arr: the 1-D numpy array. :param timesteps: a tensor of indices into

Generate text responses based on input prompts, optionally using images and controlling generation parameters. :param prompts: A bat

r"""This method accepts a list of checkpoint paths, and load each checkpoint to the model in the order as given in the list. The behaviors of

Initialize the Attention module. Args: dim (int): Number of input dimensions. n_heads (int): Number of heads

Initialize the Attention module. Args: dim (int): Number of input dimensions. n_heads (int): Number of heads

Take the mean over all non-batch dimensions.

Take the mean over all non-batch dimensions.

convert the list of question-answer pairs to a string, which contains the conversation involving all the questions and answers. Whe

This function disables printing when not in master process

Apply the model to get p(x_{t-1} | x_t), as well as a prediction of the initial x, x_0. :param model: the model, which takes

Apply the model to get p(x_{t-1} | x_t), as well as a prediction of the initial x, x_0. :param model: the model, which takes

Compute the mean and variance of the diffusion posterior: q(x_{t-1} | x_t, x_0)

Generate text in a streaming manner for a single prompt, optionally using an image. :param prompt: The input text prompt for generat

Get a term for the variational lower-bound. The resulting units are bits (rather than nats, as one might expect). This allows

Get a term for the variational lower-bound. The resulting units are bits (rather than nats, as one might expect). This allows

r"""Load one tensor parallel state dict shard from the disk and post process according to format. Args: path (str): Path to the folde

Compute the mean and variance of the diffusion posterior: q(x_{t-1} | x_t, x_0)

Diffuse the data for a given number of diffusion steps. In other words, sample from q(x_t | x_0). :param x_start: the initial

Importance-sample timesteps for a batch. :param batch_size: the number of timesteps. :param device: the torch device to save

Returns the gradient norm of parameters ``param`` s, where the gradients are viewed as a single vector. The returned norm is in FP32 even if