Search NASA⌕ Search

Engineering topics

Dash, Sajal

Publications and source records attributed to Dash, Sajal.

At least 19 records

Bigpicc: a graph-based approach to identifying carcinogenic gene combinations from mutation data

Abstract Genome data from cancer patients represents relationships between the presence of a gene mutation and cancer occurrence in a patient. Different types of cancer in human are thought to be caused by combinations of two to nine gene mutations. Identifying these combinations through traditional exhaustive search requires the amount of computation that scales exponentially with the combination size and in most cases is intractable even for cutting-edge supercomputers. We propose a parameter-free heuristic approach that leverages the intrinsic topology of gene-patient mutations to identify carcinogenic combinations. The biological relevance of the identified combinations is measured by using them to predict the presence of tumor in previously unseen samples. The resulting classifiers for 16 cancer types perform on par with exhaustive search results, and score the average of 80.1% sensitivity and 91.6% specificity for the best choice of hit range per cancer type. Our approach is able to find higher-hit carcinogenic combinations targeting which would take years of computations using exhaustive search.

Biochemistry & Molecular Biology↗

FORGE: Pre-Training Open Foundation Models for Science

Large language models (LLMs) are poised to revolutionize the way we conduct scientific research. However, both model complexity and pre-training cost are impeding effective adoption for the wider science community. Identifying suitable scientific use cases, finding the optimal balance between model and data sizes, and scaling up model training are among the most pressing issues that need to be addressed. In this study, we provide practical solutions for building and using LLM-based foundation models targeting scientific research use cases. We present an end-to-end examination of the effectiveness of LLMs in scientific research, including their scaling behavior and computational requirements on Frontier, the first Exascale supercomputer. We have also developed for release to the scientific community a suite of open foundation models called FORGE with up to 26B parameters using 257B tokens from over 200M scientific articles, with performance either on par or superior to other state-of-the-art comparable models. We have demonstrated the use and effectiveness of FORGE on scientific downstream tasks. Our research establishes best practices that can be applied across various fields to take advantage of LLMs for scientific discovery.

Yin, Junqi↗

Tokenized Data for FORGE Foundation Models

This dataset comprises a vast corpus of 257 billion tokens, accompanied by the corresponding vocabulary file employed in the pre-training of FORGE foundation models. The primary data source for this corpus is scientific documents derived from diverse origins, and they have been tokenized using the Hugging Face BPE tokenizer. Further details about this research can be found in the publication titled FORGE: Pre-Training Open Foundation Models for Science authored by Junqi Yin, Sajal Dash, Feiyi Wang, and Mallikarjun (Arjun) Shankar, presented at SC'23. The data tokenization pipeline and resulting artifacts use CORE data [Ref: Knoth, P., and Zdrahal, Z. (2012). CORE: three access levels to underpin open access. D-Lib Magazine, 18(11/12)]. For use of these data sets for any purpose, please follow the guidelines provided in https://core.ac.uk/terms .

Yin, Junqi↗

Evaluating the Cloud for Capability Class Leadership Workloads

Cloud platforms offer a variety of benefits that are very appealing for a large scale HPC facility with a diverse and dynamic user base and workload set. At the same time, there is cause for concern about transitioning to the cloud. Incorporating cloud resources into existing HPC facilities or even fully transitioning to a cloud deployment poses significant challenges at the technical, organizational, and economic levels. Regardless, based on current trends it is highly likely that cloud platforms will become an integral component of many HPC centers in some form. To gain a better understanding of both the limitations and capabilities of current cloud infrastructures we evaluated the public offerings of the three leading cloud platforms (Amazon Web Services, Microsoft Azure, and Google Cloud Platform) using a selection of representative application workloads from our facility. Our findings show that while current HPC offerings are still nascent, significant progress is being made to address the present shortcomings. At the same time, significant challenges and questions remain about whether HPC cloud offerings will be able to deliver the full range of expected benefits.

97 MATHEMATICS AND COMPUTING↗

Evaluation of pre-training large language models on leadership-class supercomputers

Large language models (LLMs) have arisen rapidly to the center stage of artificial intelligence as the foundation models applicable to many downstream learning tasks. However, how to effectively build, train, and serve such models for many high-stake and first-principle-based scientific use cases are both of great interests and of great challenges. Moreover, pre-training LLMs with billions or even trillions of parameters can be prohibitively expensive not just for academic institutions, but also for well-funded industrial and government labs. Furthermore, the energy cost and the environmental impact of developing LLMs must be kept in mind. Here, in this work, we conduct a first-of-its-kind performance analysis to understand the time and energy cost of pre-training LLMs on the Department of Energy (DOE)’s leadership-class supercomputers. Employing state-of-the-art distributed training techniques, we evaluate the computational performance of various parallelization approaches at scale for a range of model sizes, and establish a projection model for the cost of full training. Our findings provide baseline results, best practices, and heuristics for pre-training such large models that should be valuable to HPC community at large. We also offer insights and optimization strategies for using the first exascale computing system, Frontier, to train models of the size of GPT-3 and beyond.

97 MATHEMATICS AND COMPUTING↗

Scaling Resolution of Gigapixel Whole Slide Images Using Spatial Decomposition on Convolutional Neural Networks

Gigapixel images are prevalent in scientific domains ranging from remote sensing, and satellite imagery to microscopy, etc. However, training a deep learning model at the natural resolution of those images has been a challenge in terms of both, overcoming the resource limit (e.g. HBM memory constraints), as well as scaling up to a large number of GPUs. In this paper, we trained Residual neural Networks (ResNet) on 22,528 x 22,528-pixel size images using a distributed spatial decomposition method on 2,304 GPUs on the Summit Supercomputer. We applied our method on a Whole Slide Imaging (WSI) dataset from The Cancer Genome Atlas (TCGA) database. WSI images can be in the size of 100,000 x 100,000 pixels or even larger, and in this work we studied the effect of image resolution on a classification task, while achieving state-of-the-art AUC scores. Moreover, our approach doesn't need pixel-level labels, since we're avoiding patching from the WSI images completely, while adding the capability of training arbitrary large-size images. This is achieved through a distributed spatial decomposition method, by leveraging the non-block fat-tree interconnect network of the Summit architecture, which enabled GPU-to-GPU direct communication. Finally, detailed performance analysis results are shown, as well as a comparison with a data-parallel approach when possible.

Tsaris, Aristeidis (aris)↗

A comparison of histopathology imaging comprehension algorithms based on multiple instance learning

Whole slide imaging (WSI), also called digital virtual microscopy, is a new imaging modality. It allows for the application of AI and machine learning methods to cancer pathology to help establish a means for the automatic diagnosis of cancer cases. However, designing machine-learning models for WSI is computationally challenging due to its required ultra-high resolution. The current state-of-the-art models use multiple instance learning (MIL). MIL is a weakly-supervised learning method in which the model uses an array of inferences from many smaller instances to make a final classification about the entire set. In the context of WSI, researchers divide the ultra-high-resolution image into many patches. The model then classifies the slide based on an array of inferences from the patches. Among several ways of making the final classification, attention-based mechanisms have resulted in superb accuracy scores. The Transformer, one attention-based algorithm, has reported substantial improvements for WSI comprehension tasks. In this project, we studied and compared several WSI comprehension algorithms. We used the following three datasets: CAMELYON16+17, TCGALung, and TCGA-Kidney. We found that attention-based MIL algorithms performed better than standard MIL algorithms for classifying WSI images, achieving a higher mean accuracy and AUC. However, none of the attention-based algorithms performed significantly better than the others, reporting accuracy scores that varied widely. Presumably, it is due to the limited availability of training samples in the data corpus. Since it is not easy to increase the samples from human subjects, some machine learning techniques like transfer learning could help mitigate this issue.

Saunders, Adam↗

Toward an Autonomous Workflow for Single Crystal Neutron Diffraction

The operation of the neutron facility relies heavily on beamline scientists. Some experiments can take one or two days with experts making decisions along the way. Leveraging the computing power of HPC platforms and AI advances in image analyses, here we demonstrate an autonomous workflow for the single-crystal neutron diffraction experiments. The workflow consists of three components: an inference service that provides real-time AI segmentation on the image stream from the experiments conducted at the neutron facility, a continuous integration service that launches distributed training jobs on Summit to update the AI model on newly collected images, and a frontend web service to display the AI tagged images to the expert. Ultimately, the feedback can be directly fed to the equipment at the edge in deciding the next-step experiment without requiring an expert in the loop. With the analyses of the requirements and benchmarks of the performance for each component, this effort serves as the first step toward an autonomous workflow for real-time experiment steering at ORNL neutron facilities.

Yin, Junqi↗

Distilling Knowledge from Ensembles of Cluster-Constrained-Attention Multiple-Instance Learners for Whole Slide Image Classification

The peculiar nature of whole slide imaging (WSI), digitizing conventional glass slides to obtain multiple high resolution images which capture microscopic details of a patient’s histopathological features, has garnered increased interest from the computer vision research community over the last two decades. Given the unique computational space and time complexity inherent to gigapixel-size whole slide image data, researchers have proposed novel machine learning algorithms to aid in the performance of diagnostic tasks in clinical pathology. One effective algorithm represents a Whole slide image as a bag of smaller image patches, which can be represented as low-dimension image patch embeddings. Weakly supervised deep-learning methods, such as cluster-constrained-attention multiple instance learning (CLAM), have shown promising results when combined with image patch embeddings. While traditional ensemble classifiers yield improved task performance, such methods come with a steep cost in model complexity. Through knowledge distillation, it is possible to retain some performance improvements from an ensemble, while minimizing costs to model complexity. In this work, we implement a weakly supervised ensemble using clustering-constrained-attention multiple-instance learners (CLAM), which uses attention and instance-level clustering to identify task salient regions and feature extraction in whole slides. By applying logit-based and attention-based knowledge distillation, we show it is possible to retain some performance improvements resulting from the ensemble at zero cost to model complexity.

Alamudun, Folami↗

Hvac: Removing I/O Bottleneck for Large-Scale Deep Learning Applications

Scientific communities are increasingly adopting deep learning (DL) models in their applications to accelerate scientific discovery processes. However, with rapid growth in the computing capabilities of HPC supercomputers, large-scale DL applications have to spend a significant portion of training time performing I/O to a parallel storage system. Previous research works have investigated optimization techniques such as prefetching and caching. Unfortunately, there exist non-trivial challenges to adopting the existing solutions on HPC supercomputers for large-scale DL training applications, which include non-performance and/or failures at extreme scale, lack of portability and generality in design, complex deployment methodology, and being limited to a specific application or dataset. To address these challenges, we propose High-Velocity AI Cache (HVAC), a distributed read-cache layer that targets and fully exploits the node-local storage or near node-local storage technology. HVAC seamlessly accelerates read I/O by aggregating node-local or near node-local storage, avoiding metadata lookups and file locking while preserving portability in the application code. We deploy and evaluate HVAC on 1,024 nodes (with over 6000 NVIDIA V100 GPUS) of the Summit supercomputer. In particular, we evaluate the scalability, efficiency, accuracy, and load distribution of HVAC compared to GPFS and XFS-on-NVMe. With four different DL applications, we observe an average 25 % performance improvement atop GPFS and 9% drop against XFS-on-NVMe, which scale linearly and are considered the performance upper bound. We envision HVAC as an important caching library for upcoming HPC supercomputers such as Frontier.

Khan, Awais↗

A Scalable Pipeline for Gigapixel Whole Slide Imaging Analysis on Leadership Class HPC Systems

Whole Slide Imaging (WSI) captures microscopic details of a patient's histopathological features at multiple res-olutions organized across different levels. Images produced by WSI are gigapixel-sized, and saving a single image in memory requires a few gigabytes which is scarce since a complicated model occupies tens of gigabytes. Performing a simple met-ric operation on these large images is also expensive. High-performance computing (HPC) can help us quickly analyze such large images using distributed training of complex deep learning models. One popular approach in analyzing these images is to divide a WSI image into smaller tiles (patches) and then train a simpler model with these reduced-sized but large numbers of patches. However, we need to solve three pre-processing challenges efficiently for pursuing this patch-based approach. 1) Creating small patches from a high-resolution image can result in a high number (hundreds of thousands per image) of patches. Storing and processing these images can be challenging due to a large number of I/O and arithmetic operations. To reduce I/Oand memory accesses, an optimal balance between the size and number of patches must exist to reduce I/O and memory accesses. 2) WSI images may have tiny annotated regions for cancer tissue and a significant portion with normal and fatty tissues; correct patch sampling should avoid dataset imbalance. 3) storing and retrieving many patches to and from disk storage might incur I/O latency while training a deep learning model. An efficient distributed data loader should reduce I/O latency during the training and inference steps. This paper explores these three challenges and provides empirical and algorithmic solutions deployed on the Summit supercomputer hosted at the Oak Ridge Leadership Computing Facility.

Dash, Sajal↗

Long Term Per-Component Power and Thermal Measurements of the OLCF Summit System

As we move into the exascale era, the power and energy footprints of high-performance computing (HPC) systems have grown significantly larger. Due to the harsh power and thermal conditions the system, components are exposed to extreme operating conditions. Operation of such modern HPC systems requires deep insights into long term system behavior to maintain its efficiency as well as its longevity. To help the HPC community to gain such insights, we provide a dataset that records the long-term power and thermal behavior of the 200PF pre-exascale supercomputer at the Oak Ridge Leadership Computing Facility (OLCF), Summit. This system is an IBM AC922 based system that has 9,252 IBM Power9 CPUs and 27,756 Nvidia V100 GPUs and can consume up to 13MW power at peak. Heat removal is performed using medium temperature direct liquid cooling and rear-door heat exchanger based secondary cooling loop. Originally extracted from a high-resolution (1Hz) per-component (GPUs, CPUs) measurements from the system, we primarily provide a dataset that has 10-second and 1-minute mean power and thermal measurements selected from five month-long segments over the course of 2020 (January and August), 2021 (February and August), and 2022 (January). For convenience, we also provide various sub datasets randomly sampled from the time and space (hosts) of the cluster. Further details and example code for analysis can be found in the following GitHub repository: https://github.com/at-aaims/summit_power_and_thermal_data

97 MATHEMATICS AND COMPUTING↗

Image transformers for classifying acute lymphoblastic leukemia

Cancer is the leading cause of death by disease in American children. Each year, nearly 16,000 children in the United States and over 300,000 children globally are diagnosed with cancer. Leukemia is a form of blood cancer that originates in the bone marrow and accounts for one-third of pediatric cancers. This disease occurs when the bone marrow contains 20% or more immature white blood cell blasts. Acute lymphoblastic leukemia is the most prevalent leukemia type found in children, with half of all annual cases in the U.S. diagnosed for subjects under 20 years of age. To diagnose acute lymphoblastic leukemia, pathologists often conduct a morphological bone marrow assessment. This assessment determines whether the immature white blood cell blasts in bone marrow display the correct morphological characteristics, such as size and appearance of nuclei. Pathologists also use immunophenotyping via multi-channel flow cytometry to test whether certain antigens are present on the surface of blast cells; the antigens are used to identify the cell lineage of acute lymphoblastic leukemia. These manual processes require well-trained personnel and medical professionals, thus being costly in time and expenses. Computerized decision support via machine learning can accelerate the diagnosis process and reduce the cost. Training a reliable classification model to distinguish between mature and immature white blood cells is essential to the decision support system. Here, we adopted the Vision Transformer model to classify white blood cells. The Vision Transformer achieved superb classification performance compared to state-of-the-art convolutional neural networks while requiring less computational resources for training. Additionally, the latent self-attention architecture provided attention maps for a given image, providing clues as to which portion(s) of the image were significant in decision-making. We applied the Vision Transformer model and a convolutional neural network model to an acute lymphoblastic leukemia classification dataset of 12,528 samples and achieved accuracies of 88.4% and 86.2%.

Cho, Priscilla↗

Mitigating Catastrophic Forgetting in Deep Learning in a Streaming Setting Using Historical Summary

Recent advancements in scientific equipment and the adaptation of electronics and the Internet of Things (IoT) in our everyday lives resulted in large and complex data production at a high rate. Making meaningful and timely knowledge discovery at a modest cost from this big data is difficult for computing power and storage limitations. Training deep learning models incrementally in a streaming setting can help us with overcoming these limitations. However, in a well-known phenomenon named catastrophic forgetting, incrementally trained models increasingly perform poorly on the past data. To mitigate catastrophic forgetting in training in a streaming setting, we propose constructing a historical summary over time and use the summary with newly arrived data during incremental training. We propose various data summarization techniques such as random sampling, micro clustering, coreset computation, and Auto Encoders to counteract catastrophic forgetting. We built a pipeline for incremental training with a historical summary for training deep learning models for streaming data. We demonstrate the effectiveness of historical summary in mitigating catastrophic forgetting using three case studies involving three different deep learning applications: an Artificial Neural Network (ANN) for classification task on MNIST dataset, a language model (RNN-LM) on the WikiText2 dataset, and a Convolutional Neural Network (CNN), ResNet50 to classify the ImageNet dataset. Through the training of the models, we observe that catastrophic forgetting is evident in ANN and CNN but not in an RNN. For the first task, our method recovers up to 47.9% lost accuracy due to catastrophic forgetting. For the third task, the historical summary recovers classification accuracy by up to 25%. For the second task, though there is not proof of catastrophic forgetting, the training performance (PPL) improves by up to 26% with historical summary.

Dash, Sajal↗

Comparative evaluation of deep learning workloads for leadership-class systems

Deep learning (DL) workloads and their performance at scale are becoming important factors to consider as we design, develop and deploy next-generation high-performance computing systems. Since DL applications rely heavily on DL frameworks and underlying compute (CPU/GPU) stacks, it is essential to gain a holistic understanding from compute kernels, models, and frameworks of popular DL stacks, and to assess their impact on science-driven, mission-critical applications. At Oak Ridge Leadership Computing Facility (OLCF), we employ a set of micro and macro DL benchmarks established through the Collaboration of Oak Ridge, Argonne, and Livermore (CORAL) to evaluate the AI readiness of our next-generation supercomputers. In this paper, we present our early observations and performance benchmark comparisons between the Nvidia V100 based Summit system with its CUDA stack and an AMD MI100 based testbed system with its ROCm stack. We take a layered perspective on DL benchmarking and point to opportunities for future optimizations in the technologies that we consider.

Yin, Junqi↗

Scaling Out a Combinatorial Algorithm for Discovering Carcinogenic Gene Combinations to Thousands of GPUs

Cancer is a leading cause of death in the US, second only to heart disease. It is primarily a result of a combination of an estimated two-nine genetic mutations (multi-hit combinations). Although a body of research has identified hundreds of cancer-causing genetic mutations, we don’t know the specific combination of mutations responsible for specific instances of cancer for most cancer types. An approximate algorithm for solving the weighted set cover problem was previously adapted to identify combinations of genes with mutations that may be responsible for individual instances of cancer. However, the algorithm’s computational requirement scales exponentially with the number of genes, making it impractical for identifying more than three-hit combinations, even after the algorithm was parallelized and scaled up to a V100 GPU. Since most cancers have been estimated to require more than three hits, we scaled out the algorithm to identify combinations of four or more hits using 1000 nodes (6000 V100 GPUs with ≈48×106 processing cores) on the Summit supercomputer at Oak Ridge National Laboratory. Efficiently scaling out the algorithm required a series of algorithmic innovations and optimizations for balancing an exponentially divergent workload across processors and for minimizing memory latency and inter-node communication. We achieved an average strong scaling efficiency of 90.14% (80.96%–97.96% for 200 to 1000 nodes), compared to a 100 node run, with 84.18% scaling efficiency for 1000 nodes. With experimental validation, the multi-hit combinations identified here could provide further insight into the etiology of different cancer subtypes and provide a rational basis for targeted combination therapy.

Dash, Sajal↗

iBLAST: Incremental BLAST of new sequences via automated e-value correction

Search results from local alignment search tools use statistical scores that are sensitive to the size of the database to report the quality of the result. For example, NCBI BLAST reports the best matches using similarity scores and expect values (i.e., e-values) calculated against the database size. Given the astronomical growth in genomics data throughout a genomic research investigation, sequence databases grow as new sequences are continuously being added to these databases. As a consequence, the results (e.g., best hits) and associated statistics (e.g., e-values) for a specific set of queries may change over the course of a genomic investigation. Thus, to update the results of a previously conducted BLAST search to find the best matches on an updated database, scientists must currently rerun the BLAST search against the entire updated database, which translates into irrecoverable and, in turn, wasted execution time, money, and computational resources. To address this issue, we devise a novel and efficient method to redeem past BLAST searches by introducing iBLAST. iBLAST leverages previous BLAST search results to conduct the same query search but only on the incremental (i.e., newly added) part of the database, recomputes the associated critical statistics such as e-values, and combines these results to produce updated search results. Our experimental results and fidelity analyses show that iBLAST delivers search results that are identical to NCBI BLAST at a substantially reduced computational cost, i.e., iBLAST performs (1 + δ )/ δ times faster than NCBI BLAST, where δ represents the fraction of database growth. We then present three different use cases to demonstrate that iBLAST can enable efficient biological discovery at a much faster speed with a substantially reduced computational cost.

59 BASIC BIOLOGICAL SCIENCES↗