Yi-Lun Liao

Yi-Lun Liao, Brandon Wood, Abhishek Das et al. · meta-ai

Equivariant Transformers such as Equiformer have demonstrated the efficacy of applying Transformers to the domain of 3D atomistic systems. However, they are limited to small degrees of equivariant representations due to their computational complexity. In this paper, we investigate whether these architectures can scale well to higher degrees. Starting from Equiformer, we first replace $SO(3)$ convolutions with eSCN convolutions to efficiently incorporate higher-degree tensors. Then, to better leverage the power of higher degrees, we propose three architectural improvements -- attention re-normalization, separable $S^2$ activation and separable layer normalization. Putting this all together, we propose EquiformerV2, which outperforms previous state-of-the-art methods on large-scale OC20 dataset by up to $9\%$ on forces, $4\%$ on energies, offers better speed-accuracy trade-offs, and $2\times$ reduction in DFT calculations needed for computing adsorption energies. Additionally, EquiformerV2 trained on only OC22 dataset outperforms GemNet-OC trained on both OC20 and OC22 datasets, achieving much better data efficiency. Finally, we compare EquiformerV2 with Equiformer on QM9 and OC20 S2EF-2M datasets to better understand the performance gain brought by higher degrees.

Yi-Lun Liao, Tess Smidt

Despite their widespread success in various domains, Transformer networks have yet to perform well across datasets in the domain of 3D atomistic graphs such as molecules even when 3D-related inductive biases like translational invariance and rotational equivariance are considered. In this paper, we demonstrate that Transformers can generalize well to 3D atomistic graphs and present Equiformer, a graph neural network leveraging the strength of Transformer architectures and incorporating SE(3)/E(3)-equivariant features based on irreducible representations (irreps). First, we propose a simple and effective architecture by only replacing original operations in Transformers with their equivariant counterparts and including tensor products. Using equivariant operations enables encoding equivariant information in channels of irreps features without complicating graph structures. With minimal modifications to Transformers, this architecture has already achieved strong empirical results. Second, we propose a novel attention mechanism called equivariant graph attention, which improves upon typical attention in Transformers through replacing dot product attention with multi-layer perceptron attention and including non-linear message passing. With these two innovations, Equiformer achieves competitive results to previous models on QM9, MD17 and OC20 datasets.

Chaitanya K. Joshi, Xiang Fu, Yi-Lun Liao et al. · baidu, cmu

Diffusion models are the standard toolkit for generative modelling of 3D atomic systems. However, for different types of atomic systems -- such as molecules and materials -- the generative processes are usually highly specific to the target system despite the underlying physics being the same. We introduce the All-atom Diffusion Transformer (ADiT), a unified latent diffusion framework for jointly generating both periodic materials and non-periodic molecular systems using the same model: (1) An autoencoder maps a unified, all-atom representations of molecules and materials to a shared latent embedding space; and (2) A diffusion model is trained to generate new latent embeddings that the autoencoder can decode to sample new molecules or materials. Experiments on MP20, QM9 and GEOM-DRUGS datasets demonstrate that jointly trained ADiT generates realistic and valid molecules as well as materials, obtaining state-of-the-art results on par with molecule and crystal-specific models. ADiT uses standard Transformers with minimal inductive biases for both the autoencoder and diffusion model, resulting in significant speedups during training and inference compared to equivariant diffusion models. Scaling ADiT up to half a billion parameters predictably improves performance, representing a step towards broadly generalizable foundation models for generative chemistry. Open source code: https://github.com/facebookresearch/all-atom-diffusion-transformer

Yi-Lun Liao, Alexander J. Hoffman, Sabrina C. Shen et al.

As $SE(3)$-equivariant graph neural networks mature as a core tool for 3D atomistic modeling, improving their efficiency, expressivity, and physical consistency has become a central challenge for large-scale applications. In this work, we introduce EquiformerV3, the third generation of the $SE(3)$-equivariant graph attention Transformer, designed to advance all three dimensions: efficiency, expressivity, and generality. Building on EquiformerV2, we have the following three key advances. First, we optimize the software implementation, achieving $1.75\times$ speedup. Second, we introduce simple and effective modifications to EquiformerV2, including equivariant merged layer normalization, improved feedforward network hyper-parameters, and attention with smooth radius cutoff. Third, we propose SwiGLU-$S^2$ activations to incorporate many-body interactions for better theoretical expressivity and to preserve strict equivariance while reducing the complexity of sampling $S^2$ grids. Together, SwiGLU-$S^2$ activations and smooth-cutoff attention enable accurate modeling of smoothly varying potential energy surfaces (PES), generalizing EquiformerV3 to tasks requiring energy-conserving simulations and higher-order derivatives of PES. With these improvements, EquiformerV3 trained with the auxiliary task of denoising non-equilibrium structures (DeNS) achieves state-of-the-art results on OC20, OMat24, and Matbench Discovery.

Yi-Lun Liao, Sertac Karaman, Vivienne Sze

Vision Transformer (ViT) demonstrates that Transformer for natural language processing can be applied to computer vision tasks and result in comparable performance to convolutional neural networks (CNN), which have been studied and adopted in computer vision for years. This naturally raises the question of how the performance of ViT can be advanced with design techniques of CNN. To this end, we propose to incorporate two techniques and present ViT-ResNAS, an efficient multi-stage ViT architecture designed with neural architecture search (NAS). First, we propose residual spatial reduction to decrease sequence lengths for deeper layers and utilize a multi-stage architecture. When reducing lengths, we add skip connections to improve performance and stabilize training deeper networks. Second, we propose weight-sharing NAS with multi-architectural sampling. We enlarge a network and utilize its sub-networks to define a search space. A super-network covering all sub-networks is then trained for fast evaluation of their performance. To efficiently train the super-network, we propose to sample and train multiple sub-networks with one forward-backward pass. After that, evolutionary search is performed to discover high-performance network architectures. Experiments on ImageNet demonstrate that ViT-ResNAS achieves better accuracy-MACs and accuracy-throughput trade-offs than the original DeiT and other strong baselines of ViT. Code is available at https://github.com/yilunliao/vit-search.

Yi-Lun Liao, Tess Smidt, Muhammed Shuaibi et al.

Understanding the interactions of atoms such as forces in 3D atomistic systems is fundamental to many applications like molecular dynamics and catalyst design. However, simulating these interactions requires compute-intensive ab initio calculations and thus results in limited data for training neural networks. In this paper, we propose to use denoising non-equilibrium structures (DeNS) as an auxiliary task to better leverage training data and improve performance. For training with DeNS, we first corrupt a 3D structure by adding noise to its 3D coordinates and then predict the noise. Different from previous works on denoising, which are limited to equilibrium structures, the proposed method generalizes denoising to a much larger set of non-equilibrium structures. The main difference is that a non-equilibrium structure does not correspond to local energy minima and has non-zero forces, and therefore it can have many possible atomic positions compared to an equilibrium structure. This makes denoising non-equilibrium structures an ill-posed problem since the target of denoising is not uniquely defined. Our key insight is to additionally encode the forces of the original non-equilibrium structure to specify which non-equilibrium structure we are denoising. Concretely, given a corrupted non-equilibrium structure and the forces of the original one, we predict the non-equilibrium structure satisfying the input forces instead of any arbitrary structures. Since DeNS requires encoding forces, DeNS favors equivariant networks, which can easily incorporate forces and other higher-order tensors in node embeddings. We study the effectiveness of training equivariant networks with DeNS on OC20, OC22 and MD17 datasets and demonstrate that DeNS can achieve new state-of-the-art results on OC20 and OC22 and significantly improve training efficiency on MD17.

Cheng-I Jeff Lai, Erica Cooper, Yang Zhang et al.

Are end-to-end text-to-speech (TTS) models over-parametrized? To what extent can these models be pruned, and what happens to their synthesis capabilities? This work serves as a starting point to explore pruning both spectrogram prediction networks and vocoders. We thoroughly investigate the tradeoffs between sparsity and its subsequent effects on synthetic speech. Additionally, we explored several aspects of TTS pruning: amount of finetuning data versus sparsity, TTS-Augmentation to utilize unspoken text, and combining knowledge distillation and pruning. Our findings suggest that not only are end-to-end TTS models highly prunable, but also, perhaps surprisingly, pruned TTS models can produce synthetic speech with equal or higher naturalness and intelligibility, with similar prosody. All of our experiments are conducted on publicly available models, and findings in this work are backed by large-scale subjective tests and objective measures. Code and 200 pruned models are made available to facilitate future research on efficiency in TTS.

Cheng-I Jeff Lai, Yang Zhang, Alexander H. Liu et al.

Self-supervised speech representation learning (speech SSL) has demonstrated the benefit of scale in learning rich representations for Automatic Speech Recognition (ASR) with limited paired data, such as wav2vec 2.0. We investigate the existence of sparse subnetworks in pre-trained speech SSL models that achieve even better low-resource ASR results. However, directly applying widely adopted pruning methods such as the Lottery Ticket Hypothesis (LTH) is suboptimal in the computational cost needed. Moreover, we show that the discovered subnetworks yield minimal performance gain compared to the original dense network. We present Prune-Adjust-Re-Prune (PARP), which discovers and finetunes subnetworks for much better performance, while only requiring a single downstream ASR finetuning run. PARP is inspired by our surprising observation that subnetworks pruned for pre-training tasks need merely a slight adjustment to achieve a sizeable performance boost in downstream ASR tasks. Extensive experiments on low-resource ASR verify (1) sparse subnetworks exist in mono-lingual/multi-lingual pre-trained speech SSL, and (2) the computational advantage and performance gain of PARP over baseline pruning methods. In particular, on the 10min Librispeech split without LM decoding, PARP discovers subnetworks from wav2vec 2.0 with an absolute 10.9%/12.6% WER decrease compared to the full model. We further demonstrate the effectiveness of PARP via: cross-lingual pruning without any phone recognition degradation, the discovery of a multi-lingual subnetwork for 10 spoken languages in 1 finetuning run, and its applicability to pre-trained BERT/XLNet for natural language tasks.

Tien-Ju Yang, Yi-Lun Liao, Vivienne Sze

Neural architecture search (NAS) typically consists of three main steps: training a super-network, training and evaluating sampled deep neural networks (DNNs), and training the discovered DNN. Most of the existing efforts speed up some steps at the cost of a significant slowdown of other steps or sacrificing the support of non-differentiable search metrics. The unbalanced reduction in the time spent per step limits the total search time reduction, and the inability to support non-differentiable search metrics limits the performance of discovered DNNs. In this paper, we present NetAdaptV2 with three innovations to better balance the time spent for each step while supporting non-differentiable search metrics. First, we propose channel-level bypass connections that merge network depth and layer width into a single search dimension to reduce the time for training and evaluating sampled DNNs. Second, ordered dropout is proposed to train multiple DNNs in a single forward-backward pass to decrease the time for training a super-network. Third, we propose the multi-layer coordinate descent optimizer that considers the interplay of multiple layers in each iteration of optimization to improve the performance of discovered DNNs while supporting non-differentiable search metrics. With these innovations, NetAdaptV2 reduces the total search time by up to $5.8\times$ on ImageNet and $2.4\times$ on NYU Depth V2, respectively, and discovers DNNs with better accuracy-latency/accuracy-MAC trade-offs than state-of-the-art NAS works. Moreover, the discovered DNN outperforms NAS-discovered MobileNetV3 by 1.8% higher top-1 accuracy with the same latency. The project website is http://netadapt.mit.edu.

Yi-Lun Liao, Yao-Cheng Yang, Yu-Chiang Frank Wang

Aiming at inferring 3D shapes from 2D images, 3D shape reconstruction has drawn huge attention from researchers in computer vision and deep learning communities. However, it is not practical to assume that 2D input images and their associated ground truth 3D shapes are always available during training. In this paper, we propose a framework for semi-supervised 3D reconstruction. This is realized by our introduced 2D-3D self-consistency, which aligns the predicted 3D models and the projected 2D foreground segmentation masks. Moreover, our model not only enables recovering 3D shapes with the corresponding 2D masks, camera pose information can be jointly disentangled and predicted, even such supervision is never available during training. In the experiments, we qualitatively and quantitatively demonstrate the effectiveness of our model, which performs favorably against state-of-the-art approaches in either supervised or semi-supervised settings.