Flow Matching (FM) is a recent framework for generative modeling that has achieved state-of-the-art performance across various domains, including image, video, audio, speech, and biological structures. This guide offers a comprehensive and self-contained review of FM, covering its mathematical foundations, design choices, and extensions. By also providing a PyTorch package featuring relevant examples (e.g., image and text generation), this work aims to serve as a resource for both novice and experienced researchers interested in understanding, applying and further developing FM.

We introduce generator matching, a modality-agnostic framework for generative modeling using arbitrary Markov processes. Generators characterize the infinitesimal evolution of a Markov process, which we leverage for generative modeling in a similar vein to flow matching: we construct conditional generators which generate single data points, then learn to approximate the marginal generator which generates the full data distribution. We show that generator matching unifies various generative modeling methods, including diffusion models, flow matching and discrete diffusion models. Furthermore, it provides the foundation to expand the design space to new and unexplored Markov processes such as jump processes. Finally, generator matching enables the construction of superpositions of Markov generative processes and enables the construction of multimodal models in a rigorous manner. We empirically validate our method on protein and image structure generation, showing that superposition with a jump process improves image generation.

The design space of discrete-space diffusion or flow generative models are significantly less well-understood than their continuous-space counterparts, with many works focusing only on a simple masked construction. In this work, we aim to take a holistic approach to the construction of discrete generative models based on continuous-time Markov chains, and for the first time, allow the use of arbitrary discrete probability paths, or colloquially, corruption processes. Through the lens of optimizing the symmetric kinetic energy, we propose velocity formulas that can be applied to any given probability path, completely decoupling the probability and velocity, and giving the user the freedom to specify any desirable probability path based on expert knowledge specific to the data domain. Furthermore, we find that a special construction of mixture probability paths optimizes the symmetric kinetic energy for the discrete case. We find that we can outperform the mask construction even in text with kinetic-optimal mixture paths, while we can make use of domain-specific constructions of the probability path over the visual domain. Generative models over discrete spaces have not seen as much progress on the methodology side compared to continuous-space counterparts. For the most part, applications such as large language modeling rely solely on autoregressive models (Radford et al., 2019; Bommasani et al., 2021). The simplicity of autoregressive modeling has also motivated people to use them for multimodal generation, where other modalities, such as images and videos, are tokenized and modeled within an autoregressive framework (Van den Oord et al., 2016; Team, 2024; Sun et al., 2024). A promising framework that brings iterative refinement to the discrete case is to consider the use of Markov chains within a dynamical generative framework.

Tokenization is associated with many poorly understood shortcomings in language models (LMs), yet remains an important component for long sequence scaling purposes. We discover that, even when the two models are statistically equivalent, their predictive distributions over the next byte can be substantially different, a phenomenon we term as "tokenization bias". To fully characterize this phenomenon, we introduce the Byte-Token Representation Lemma, a framework that establishes a mapping between the learned token distribution and its equivalent byte-level distribution. From this result, we develop a next-byte sampling algorithm that eliminates tokenization bias without requiring further training or optimization. In other words, this enables zero-shot conversion of tokenized LMs into statistically equivalent token-free ones. We demonstrate its broad applicability with two use cases: fill-in-the-middle (FIM) tasks and model ensembles. In FIM tasks where input prompts may terminate mid-token, leading to out-of-distribution tokenization, our method mitigates performance degradation and achieves an approximately 18% improvement in FIM coding benchmarks, consistently outperforming the standard token healing fix. For model ensembles where each model employs a distinct vocabulary, our approach enables seamless integration, resulting in improved performance (up to 3.7%) over individual models across various standard baselines in reasoning, knowledge, and coding. Transformers form the backbone of all widely-used state-of-the-art language models (LMs) such as GPTs (Brown et al., 2020), Llama (Touvron et al., 2023), ans Mistral (Jiang et al., 2023a).

Taming the generation outcome of state of the art Diffusion and Flow-Matching (FM) models without having to re-train a task-specific model unlocks a powerful tool for solving inverse problems, conditional generation, and controlled generation in general. In this work we introduce D-Flow, a simple framework for controlling the generation process by differentiating through the flow, optimizing for the source (noise) point. We motivate this framework by our key observation stating that for Diffusion/FM models trained with Gaussian probability paths, differentiating through the generation process projects gradient on the data manifold, implicitly injecting the prior into the optimization process. We validate our framework on linear and non-linear controlled generation problems including: image and audio inverse problems and conditional molecule generation reaching state of the art performance across all.

We introduce SPIRIT-LM, a foundation multimodal language model that freely mixes text and speech. Our model is based on a pretrained text language model that we extend to the speech modality by continuously training it on text and speech units. Speech and text sequences are concatenated as a single set of tokens, and trained with a word-level interleaving method using a small automatically-curated speech-text parallel corpus. SPIRIT-LM comes in two versions: a BASE version that uses speech semantic units and an EXPRESSIVE version that models expressivity using pitch and style units in addition to the semantic units. For both versions, the text is encoded with subword BPE tokens. The resulting model displays both the semantic abilities of text models and the expressive abilities of speech models. Additionally, we demonstrate that SPIRIT-LM is able to learn new tasks in a few-shot fashion across modalities (i.e. ASR, TTS, Speech Classification).

Speech language models (SpeechLMs) process and generate acoustic data only, without textual supervision. In this work, we propose TWIST, a method for training SpeechLMs using a warm-start from a pretrained textual language models. We show using both automatic and human evaluations that TWIST outperforms a cold-start SpeechLM across the board. We empirically analyze the effect of different model design choices such as the speech tokenizer, the pretrained textual model, and the dataset size. We find that model and dataset scale both play an important role in constructing better-performing SpeechLMs. Based on our observations, we present the largest (to the best of our knowledge) SpeechLM both in terms of number of parameters and training data. We additionally introduce two spoken versions of the StoryCloze textual benchmark to further improve model evaluation and advance future research in the field. We make speech samples, code and models publicly available: https://pages.cs.huji.ac.il/adiyoss-lab/twist/ .

We tackle the task of conditional music generation. We introduce MusicGen, a single Language Model (LM) that operates over several streams of compressed discrete music representation, i.e., tokens. Unlike prior work, MusicGen is comprised of a single-stage transformer LM together with efficient token interleaving patterns, which eliminates the need for cascading several models, e.g., hierarchically or upsampling. Following this approach, we demonstrate how MusicGen can generate high-quality samples, both mono and stereo, while being conditioned on textual description or melodic features, allowing better controls over the generated output. We conduct extensive empirical evaluation, considering both automatic and human studies, showing the proposed approach is superior to the evaluated baselines on a standard text-to-music benchmark. Through ablation studies, we shed light over the importance of each of the components comprising MusicGen. Music samples, code, and models are available at https://github.com/facebookresearch/audiocraft

T, a masked generative sequence modeling method that operates directly over several streams of audio tokens. T is comprised of a single-stage, non-autoregressive transformer. During training, we predict spans of masked tokens obtained from a masking scheduler, while during inference we gradually construct the output sequence using several decoding steps. T, which will be then used for later decoding steps. T, in which we fuse between autoregressive and non-autoregressive models to generate the first few seconds in an autoregressive manner while the rest of the sequence is being decoded in parallel. T for the task of text-to-music and text-to-audio generation and conduct an extensive empirical evaluation, considering both objective metrics and human studies. The proposed approach is comparable to the evaluated baselines, while being significantly faster (x7 faster than the autoregressive baseline). Samples are available on our demo page https://pages.cs.huji.ac.il/adiyoss-lab/MAGNeT Recent developments in self-supervised representation learning (Hsu et al., 2021; Défossez et al., 2022), sequence modeling (Touvron et al., 2023; Rozière et al., 2023), and audio synthesis (Lee et al., 2022; Polyak et al., 2021) allow a great leap in performance when considering high quality conditional audio generation. Recently, Défossez et al. (2022); Zeghidour et al. (2021) proposed to apply a VQ-VAE directly on the raw waveform using residual vector quantization to obtain a multi-stream discrete representation of the audio signal. Later on, Kreuk et al. (2022a); Wang et al. (2023); Zhang et al. (2023); Copet et al. (2023); Kreuk et al. (2022b) presented a conditional language modeling on such audio signals representations. In parallel, Schneider et al. (2023); Huang et al. (2023b); Liu et al. (2023a) proposed training a conditional diffusion-based generative model operating on learned continuous representations of the audio signal obtained from a pre-trained auto-encoder model. Work was done as part of Alon's internship at FAIR.

We consider the task of generating diverse and realistic videos guided by natural audio samples from a wide variety of semantic classes. For this task, the videos are required to be aligned both globally and temporally with the input audio: globally, the input audio is semantically associated with the entire output video, and temporally, each segment of the input audio is associated with a corresponding segment of that video. We utilize an existing text-conditioned video generation model and a pre-trained audio encoder model. The proposed method is based on a lightweight adaptor network, which learns to map the audio-based representation to the input representation expected by the text-to-video generation model. As such, it also enables video generation conditioned on text, audio, and, for the first time as far as we can ascertain, on both text and audio. We validate our method extensively on three datasets demonstrating significant semantic diversity of audio-video samples and further propose a novel evaluation metric (AV-Align) to assess the alignment of generated videos with input audio samples. AV-Align is based on the detection and comparison of energy peaks in both modalities. In comparison to recent state-of-the-art approaches, our method generates videos that are better aligned with the input sound, both with respect to content and temporal axis. We also show that videos produced by our method present higher visual quality and are more diverse.