Speculative decoding accelerates inference in large language models (LLMs) by generating multiple draft tokens simultaneously.

Quantized training of Large Language Models (LLMs) remains an open challenge, as maintaining accuracy while performing all matrix multiplications in low precision has proven difficult. This is particularly the case when fine-tuning pre-trained models, which can have large weight, activation, and error (output gradient) outlier values that make lower-precision optimization difficult. To address this, we present HALO, a new quantization-aware training approach for Transformers that enables accurate and efficient low-precision training by combining 1) strategic placement of Hadamard rotations in both forward and backward passes, which mitigate outliers, 2) high-performance kernel support, and 3) FSDP integration for low-precision communication. Our approach ensures that all large matrix multiplications during the forward and backward passes are executed in lower precision.

Speculative decoding is a technique to leverage hardware concurrency in order to enable multiple steps of token generation in a single forward pass, thus improving the efficiency of large-scale autoregressive (AR) Transformer models. State-space models (SSMs) are already more efficient than ARTransformers, since their state summarizes all past data with no need to cache or re-process tokens in the sliding window context. However, their state can also comprise thousands of tokens; so, speculative decoding has recently been extended to SSMs. Existing approaches, however, do not leverage the tree-based verification methods, since current SSMs lack the means to compute a token tree efficiently. We propose the first scalable algorithm to perform tree-based speculative decoding in state-space models (SSMs) and hybrid architectures of SSMs and Transformer layers. We exploit the structure of accumulated state transition matrices to facilitate tree-based speculative decoding with minimal overhead relative to current SSM implementations. Along with the algorithm, we describe a hardware-aware implementation that improves naive application of ARTransformer tree-based speculative decoding methods to SSMs. Furthermore, we outperform vanilla speculative decoding with SSMs even with a baseline drafting model and tree structure on three different benchmarks, opening up opportunities for further speed up with SSM and hybrid model inference. Code can be find at: https://github.com/wyc1997/stree.

For large language models (LLMs), sparse autoencoders (SAEs) have been shown to decompose intermediate representations that often are not interpretable directly into sparse sums of interpretable features, facilitating better control and subsequent analysis. However, similar analysesTextand approaches have been lacking for text-toimage models. We investigate the possibility of using SAEs to learn interpretable features for SDXLTurbo, a few-step text-to-image diffusion model. To this end, SDXL Basewe train SAEs on the updates performed by transformer blocks within SDXL 25 steps Turbo's denoising U-net in its 1-step setting. Interestingly, we find that they generalize to 4-step SDXLTurbo and even to the multi-step SDXL base model (i.e., a different model) without additional training. In addition, we show that their learned features are interpretable, causally influence the generation process, and reveal specialization among the blocks.

With the growing requirement for natural human-computer interaction, speechbased systems receive increasing attention as speech is one of the most common forms of daily communication. However, the existing speech models still experience high latency when generating the first audio token during streaming, which poses a significant bottleneck for deployment. To address this issue, we propose VITA-Audio, an end-to-end large speech model with fast audio-text token generation. Specifically, we introduce a lightweight Multiple Cross-modal Token Prediction (MCTP) module that efficiently generates multiple audio tokens within a single model forward pass, which not only accelerates the inference but also significantly reduces the latency for generating the first audio in streaming scenarios. In addition, a four-stage progressive training strategy is explored to achieve model acceleration with minimal loss of speech quality. To our knowledge, VITA-Audio is the first multi-modal large language model capable of generating audio output during the first forward pass, enabling real-time conversational capabilities with minimal latency. VITA-Audio is fully reproducible and is trained on open-source data only. Experimental results demonstrate that our model achieves an inference speedup of 3 5 at 7B parameter scale, but also significantly outperforms opensource models of similar model size on multiple benchmarks for automatic speech recognition (ASR), text-to-speech (TTS), and spoken question answering (SQA).

Although model compression and parameter offloading are common strategies to address memory limitations, compression can degrade quality, and offloading maintains quality but suffers from slow inference. Speculative decoding presents a promising avenue to accelerate parameter offloading, utilizing a fast draft model to propose multiple draft tokens, which are then verified by the target LLM in parallel with a single forward pass. This method reduces the time-consuming data transfers in forward passes that involve offloaded weight transfers. Existing methods often rely on pretrained weights of the same family, but require additional training to align with custom-trained models. Moreover, approaches that involve draft model training usually yield only modest speedups.

Training large language models (LLMs) models directly in low-precision offers a way to address computational costs by improving both throughput and energy efficiency.

This work aims to compress the vision tokens of an LVLM into a representation that is simultaneously suitable for (a) generative and (b) discriminative tasks, (c) is nearly lossless, and (d) storage-efficient. To this end, we propose C&C, a novel compression method that leverages the LVLM itself for task-agnostic visual token compression. Unlike prior methods that perform token reduction on-the-fly, our approach offloads computation to a dedicated, upfront indexing stage, effectively decoupling compression from generation. This enables learning more powerful representations for generation during inference. At the core of C&C is a "doubleforward pass" training strategy. During the first forward pass, the LLM (of the LVLM) creates a bottleneck by compressing the dense visual tokens into a few summary tokens.

Fine-tuning pre-trained Large Language Models (LLMs) for downstream tasks using First-Order (FO) optimizers presents significant computational challenges. Parameter-Efficient Fine-Tuning~(PEFT) methods have been proposed to address these challenges by freezing most model parameters and training only a small subset. While PEFT is efficient, it may not outperform full fine-tuning when high task-specific performance is required. Zeroth-Order (ZO) methods offer an alternative for fine-tuning the entire pre-trained model by approximating gradients using only the forward pass, thus eliminating the computational burden of back-propagation, % in first-order methods, but they converge painfully slowly and are very sensitive to the choice of task prompts. We bridge these worlds with Bilevel ZOFO, a penalty based bilevel formulation that treats adapter parameters as a lower level learner coupled to an upper level ZO optimizer of the full backbone.

Although model compression and parameter offloading are common strategies to address memory limitations, compression can degrade quality, and offloading maintains quality but suffers from slow inference. Speculative decoding presents a promising avenue to accelerate parameter offloading, utilizing a fast draft model to propose multiple draft tokens, which are then verified by the target LLM in parallel with a single forward pass. This method reduces the time-consuming data transfers in forward passes that involve offloaded weight transfers. Existing methods often rely on pretrained weights of the same family, but require additional training to align with custom-trained models. Moreover, approaches that involve draft model training usually yield only modest speedups.