A class that keeps all conversation history.

Layers used in mapping text embeddings to visual outputs.

PatchEmbed from Hydra

PatchEmbed from Hydra

The model can behave as an encoder (with only self-attention) as well as a decoder, in which case a layer of cross-attention is added between

The model can behave as an encoder (with only self-attention) as well as a decoder, in which case a layer of cross-attention is added between

Base class for causal language model (or autoregressive) outputs. Args: loss (`torch.FloatTensor` of shape `(1,)`, *optional*, retur

LoRA for LLaMa model

Track a series of values and provide access to smoothed values over a window or the global series average.

Convert the video into 3 smaller clips spatially. Must be used after the temporal crops to get spatial crops, and should be used with

Fake file-like stream object that redirects writes to a logger instance.

Construct the embeddings from word and position embeddings.

Construct the embeddings from word and position embeddings.

Collate examples for supervised fine-tuning.

similar to normal implementation of distributed sampler, except implementation is at the batch sampler level, instead of just the sampler lev

similar to normal implementation of distributed sampler, except implementation is at the batch sampler level, instead of just the sampler lev

This is a modality-blended batch sampler which allows to sample a batch data from different dataset alternatively.

This is a modality-blended batch sampler which allows to sample a batch data from different dataset alternatively.

Grouping Block to group similar segments together. Args: dim (int): Dimension of the input. out_dim (int): Dimension of the outpu

A Transformer layer with Grouping Block for one stage. Args: dim (int): Number of input channels. num_input_token (int): Input re

Sampler that samples indices in a way that groups together features of the dataset of roughly the same length while keeping a bit of randomne

Multi-headed attention from 'Attention Is All You Need' paper

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`LlamaDecoderLayer`] Args: config: LlamaConfig