-
Stable LM 2 1.6B Technical Report
Authors:
Marco Bellagente,
Jonathan Tow,
Dakota Mahan,
Duy Phung,
Maksym Zhuravinskyi,
Reshinth Adithyan,
James Baicoianu,
Ben Brooks,
Nathan Cooper,
Ashish Datta,
Meng Lee,
Emad Mostaque,
Michael Pieler,
Nikhil Pinnaparju,
Paulo Rocha,
Harry Saini,
Hannah Teufel,
Niccolo Zanichelli,
Carlos Riquelme
Abstract:
We introduce StableLM 2 1.6B, the first in a new generation of our language model series. In this technical report, we present in detail the data and training procedure leading to the base and instruction-tuned versions of StableLM 2 1.6B. The weights for both models are available via Hugging Face for anyone to download and use. The report contains thorough evaluations of these models, including z…
▽ More
We introduce StableLM 2 1.6B, the first in a new generation of our language model series. In this technical report, we present in detail the data and training procedure leading to the base and instruction-tuned versions of StableLM 2 1.6B. The weights for both models are available via Hugging Face for anyone to download and use. The report contains thorough evaluations of these models, including zero- and few-shot benchmarks, multilingual benchmarks, and the MT benchmark focusing on multi-turn dialogues. At the time of publishing this report, StableLM 2 1.6B was the state-of-the-art open model under 2B parameters by a significant margin. Given its appealing small size, we also provide throughput measurements on a number of edge devices. In addition, we open source several quantized checkpoints and provide their performance metrics compared to the original model.
△ Less
Submitted 27 February, 2024;
originally announced February 2024.
-
Networked Restless Multi-Armed Bandits for Mobile Interventions
Authors:
Han-Ching Ou,
Christoph Siebenbrunner,
Jackson Killian,
Meredith B Brooks,
David Kempe,
Yevgeniy Vorobeychik,
Milind Tambe
Abstract:
Motivated by a broad class of mobile intervention problems, we propose and study restless multi-armed bandits (RMABs) with network effects. In our model, arms are partially recharging and connected through a graph, so that pulling one arm also improves the state of neighboring arms, significantly extending the previously studied setting of fully recharging bandits with no network effects. In mobil…
▽ More
Motivated by a broad class of mobile intervention problems, we propose and study restless multi-armed bandits (RMABs) with network effects. In our model, arms are partially recharging and connected through a graph, so that pulling one arm also improves the state of neighboring arms, significantly extending the previously studied setting of fully recharging bandits with no network effects. In mobile interventions, network effects may arise due to regular population movements (such as commuting between home and work). We show that network effects in RMABs induce strong reward coupling that is not accounted for by existing solution methods. We propose a new solution approach for networked RMABs, exploiting concavity properties which arise under natural assumptions on the structure of intervention effects. We provide sufficient conditions for optimality of our approach in idealized settings and demonstrate that it empirically outperforms state-of-the art baselines in three mobile intervention domains using real-world graphs.
△ Less
Submitted 28 January, 2022;
originally announced January 2022.
-
GraphVAMPNet, using graph neural networks and variational approach to markov processes for dynamical modeling of biomolecules
Authors:
Mahdi Ghorbani,
Samarjeet Prasad,
Jeffery B. Klauda,
Bernard R. Brooks
Abstract:
Finding low dimensional representation of data from long-timescale trajectories of biomolecular processes such as protein-folding or ligand-receptor binding is of fundamental importance and kinetic models such as Markov modeling have proven useful in describing the kinetics of these systems. Recently, an unsupervised machine learning technique called VAMPNet was introduced to learn the low dimensi…
▽ More
Finding low dimensional representation of data from long-timescale trajectories of biomolecular processes such as protein-folding or ligand-receptor binding is of fundamental importance and kinetic models such as Markov modeling have proven useful in describing the kinetics of these systems. Recently, an unsupervised machine learning technique called VAMPNet was introduced to learn the low dimensional representation and linear dynamical model in an end-to-end manner. VAMPNet is based on variational approach to Markov processes (VAMP) and relies on neural networks to learn the coarse-grained dynamics. In this contribution, we combine VAMPNet and graph neural networks to generate an end-to-end framework to efficiently learn high-level dynamics and metastable states from the long-timescale molecular dynamics trajectories. This method bears the advantages of graph representation learning and uses graph message passing operations to generate an embedding for each datapoint which is used in the VAMPNet to generate a coarse-grained representation. This type of molecular representation results in a higher resolution and more interpretable Markov model than the standard VAMPNet enabling a more detailed kinetic study of the biomolecular processes. Our GraphVAMPNet approach is also enhanced with an attention mechanism to find the important residues for classification into different metastable states.
△ Less
Submitted 12 January, 2022;
originally announced January 2022.
-
Variational embedding of protein folding simulations using gaussian mixture variational autoencoders
Authors:
Mahdi Ghorbani,
Samarjeet Prasad,
Jeffery B. Klauda,
Bernard R. Brooks
Abstract:
Conformational sampling of biomolecules using molecular dynamics simulations often produces large amount of high dimensional data that makes it difficult to interpret using conventional analysis techniques. Dimensionality reduction methods are thus required to extract useful and relevant information. Here we devise a machine learning method, Gaussian mixture variational autoencoder (GMVAE) that ca…
▽ More
Conformational sampling of biomolecules using molecular dynamics simulations often produces large amount of high dimensional data that makes it difficult to interpret using conventional analysis techniques. Dimensionality reduction methods are thus required to extract useful and relevant information. Here we devise a machine learning method, Gaussian mixture variational autoencoder (GMVAE) that can simultaneously perform dimensionality reduction and clustering of biomolecular conformations in an unsupervised way. We show that GMVAE can learn a reduced representation of the free energy landscape of protein folding with highly separated clusters that correspond to the metastable states during folding. Since GMVAE uses a mixture of Gaussians as the prior, it can directly acknowledge the multi-basin nature of protein folding free-energy landscape. To make the model end-to-end differentialble, we use a Gumbel-softmax distribution. We test the model on three long-timescale protein folding trajectories and show that GMVAE embedding resembles the folding funnel with folded states down the funnel and unfolded states outer in the funnel path. Additionally, we show that the latent space of GMVAE can be used for kinetic analysis and Markov state models built on this embedding produce folding and unfolding timescales that are in close agreement with other rigorous dynamical embeddings such as time independent component analysis (TICA).
△ Less
Submitted 27 August, 2021;
originally announced August 2021.
-
Recommendations for Planning Inclusive Astronomy Conferences
Authors:
Inclusive Astronomy 2 Local Organizing Committee,
:,
Brian Brooks,
Keira Brooks,
Lea Hagen,
Nimish Hathi,
Samantha Hoffman,
James Paranilam,
Laura Prichard
Abstract:
The Inclusive Astronomy (IA) conference series aims to create a safe space where community members can listen to the experiences of marginalized individuals in astronomy, discuss actions being taken to address inequities, and give recommendations to the community for how to improve diversity, equity, and inclusion in astronomy. The first IA was held in Nashville, TN, USA, 17-19 June, 2015. The Inc…
▽ More
The Inclusive Astronomy (IA) conference series aims to create a safe space where community members can listen to the experiences of marginalized individuals in astronomy, discuss actions being taken to address inequities, and give recommendations to the community for how to improve diversity, equity, and inclusion in astronomy. The first IA was held in Nashville, TN, USA, 17-19 June, 2015. The Inclusive Astronomy 2 (IA2) conference was held in Baltimore, MD, USA, 14-15 October, 2019. The Inclusive Astronomy 2 (IA2) Local Organizing Committee (LOC) has put together a comprehensive document of recommendations for planning future Inclusive Astronomy conferences based on feedback received and lessons learned. While these are specific to the IA series, many parts will be applicable to other conferences as well. Please find the recommendations and accompanying letter to the community here: https://outerspace.stsci.edu/display/IA2/LOC+Recommendations.
△ Less
Submitted 21 July, 2020;
originally announced July 2020.
-
Link community detection through global optimization and the inverse resolution limit of partition density
Authors:
Juyong Lee,
Zhong-Yuan Zhang,
Jooyoung Lee,
Bernard R. Brooks,
Yong-Yeol Ahn
Abstract:
We investigate the possibility of global optimization-based overlapping community detection, using link community framework. We first show that partition density, the original quality function used in link community detection method, is not suitable as a quality function for global optimization because it prefers breaking communities into triangles except in highly limited conditions. We analytica…
▽ More
We investigate the possibility of global optimization-based overlapping community detection, using link community framework. We first show that partition density, the original quality function used in link community detection method, is not suitable as a quality function for global optimization because it prefers breaking communities into triangles except in highly limited conditions. We analytically derive those conditions and confirm it with computational results on direct optimization of various synthetic and real-world networks. To overcome this limitation, we propose alternative approaches combining the weighted line graph transformation and existing quality functions for node-based communities. We suggest a new line graph weighting scheme, a normalized Jaccard index. Computational results show that community detection using the weighted line graphs generated with the normalized Jaccard index leads to a more accurate community structure.
△ Less
Submitted 19 January, 2016;
originally announced January 2016.
