As deep learning models become increasingly complex, the deep learning compilers are critical for enhancing the system efficiency and unlocking hidden optimization opportunities. Although excellent speedups have been achieved in inference workloads, existing compilers face significant limitations in training. Firstly, the training computation graph involves intricate operations challenging to fuse, such as normalization, loss functions, and reductions, which limit optimization opportunities like kernel fusion. Secondly, the training graph's additional edges connecting forward and backward operators pose challenges in finding optimal and feasible partitions for kernel fusion. More importantly, existing compilers cannot either generate kernels with state-of-the-art performance on modern GPUs or accommodate diverse fusion patterns.

This research-to-practice full paper presents our innovative approach to integrating Generative Artificial Intelligence (GAI) chatbots into selected BSCS courses. These courses operate within an Online Learning Environment (OLE) based on asynchronous communication. We have detailed two distinct approaches to implement Retrieval-Augmented Generation (RAG) model-based GAI chatbots. The first approach utilizes Microsoft Copilot Studio, allowing us to create customized chatbots without the need for coding. The second approach involves Python coding to develop more advanced GAI chatbots. Both methods are straightforward and cost-effective, leveraging the latest advancements to enhance communication between students and faculty. By incorporating frequently asked questions and answers from various BSCS courses into these chatbots, we can enrich the learning experience and aid both students and faculty in achieving their goals. Our preliminary results suggest that with careful planning, design, and implementation, these chatbots can circumvent common issues such as misinformation, inaccuracy, ethical and safety concerns, and lack of contextual understanding, thereby enhancing the effectiveness of student learning in an asynchronous communication-based OLE.

During the last two and half years, the Covid-19 Pandemic has made online learning environment (OLE) a necessary modality for all educational degree programs globally. However, migrating a conventional ground campus only Bachelor of Science in Computer Engineering (BSCE) degree program into an online modality is a project more complicated than simply replicating what is common in an on-ground classroom into a virtual classroom. The major challenges include how to maintain instructional effectiveness in an online modality, how to enable students accessing lab equipment to conduct hands-on experiments in an OLE, how to support students expeditiously when they encounter difficulties in an asynchronous communication based OLE, and how to assess students work produced in an OLE, etc. In this paper, we share our experience on how we applied a systematic approach, guided by various learning pedagogies, migrating a ground campus only ABET accredited BSCE degree program into the online modality successfully. Our practice has demonstrated that this approach can assist us to quickly identify solutions, quite often unique to the OLE, based on the specific characteristics of online modality to effectively address various typical challenges encountered during the process of migration.

In recent years, to meet the growing needs of upgrading computer-related competence of working adults, many online Bachelor of Science of Computer Science (BSCS) or Bachelor of Science of Information Technology (BSIT) degree programs have been created. However, how to effectively support these online students is still a challenge. This is due to special needs such as customizing the learning to individual backgrounds, supporting various programming and technical hands-on labs, and promoting teamwork engagement, etc. We have conducted a project integrated three initiatives to create an Online Learning Environment (OLE) suited to BSCS and BSIT degree programs: (a) using the adaptive learning software to customize the learning content based on each student's existing academic foundation; (b) using Amazon's Web Services to set up the hands-on labs; and (c) using discussion board assignments to promote the engagement of student-student and student-faculty, to train students to be competent in teamwork. The initial results of our project have consistently shown that all three initiatives mentioned above have a positive effect on students' learning. Our work has demonstrated that it is possible to create an OLE, which is not only cost-effective but also functionality proper, to the online BSCS and BSIT degree programs.

BACKGROUND: The sequencing of the human genome has enabled us to access a comprehensive list of genes (both experimental and predicted) for further analysis. While a majority of the approximately 30,000 known and predicted human coding genes are characterized and have been assigned at least one function, there remains a fair number of genes (about 12,000) for which no annotation has been made. The recent sequencing of other genomes has provided us with a huge amount of auxiliary sequence data which could help in the characterization of the human genes. Clustering these sequences into families is one of the first steps to perform comparative studies across several genomes. RESULTS: Here we report a novel clustering algorithm (CLUGEN) that has been used to cluster sequences of experimentally verified and predicted proteins from all sequenced genomes using a novel distance metric which is a neural network score between a pair of protein sequences. This distance metric is based on the pairwise sequence similarity score and the similarity between their domain structures. The distance metric is the probability that a pair of protein sequences are of the same Interpro family/domain, which facilitates the modelling of transitive homology closure to detect remote homologues. The hierarchical average clustering method is applied with the new distance metric. CONCLUSION: Benchmarking studies of our algorithm versus those reported in the literature shows that our algorithm provides clustering results with lower false positive and false negative rates. The clustering algorithm is applied to cluster several eukaryotic genomes and several dozens of prokaryotic genomes.