Neuri: Diversifying dnn generation via inductive rule inference
J Liu, J Peng, Y Wang, L Zhang - Proceedings of the 31st ACM Joint …, 2023 - dl.acm.org
Deep Learning (DL) is prevalently used in various industries to improve decision-making
and automate processes, driven by the ever-evolving DL libraries and compilers. The …
and automate processes, driven by the ever-evolving DL libraries and compilers. The …
D3: Differential Testing of Distributed Deep Learning with Model Generation
J Wang, HV Pham, Q Li, L Tan, Y Guo… - IEEE Transactions …, 2024 - ieeexplore.ieee.org
Deep Learning (DL) techniques have been widely deployed in many application domains.
The growth of DL models' size and complexity demands distributed training of DL models …
The growth of DL models' size and complexity demands distributed training of DL models …
Detecting Numerical Deviations in Deep Learning Models Introduced by the TVM Compiler
Z Xia, Y Chen, P Nie, Z Wang - 2024 IEEE 35th International …, 2024 - ieeexplore.ieee.org
Deep learning (DL) compilers are crucial for deploying DL models and speeding up their
inferences. Meanwhile, they may introduce numerical deviations, and finally undefined or …
inferences. Meanwhile, they may introduce numerical deviations, and finally undefined or …
Emerging Platforms Meet Emerging LLMs: A Year-Long Journey of Top-Down Development
S Feng, J Liu, R Lai, CF Ruan, Y Yu, L Zhang… - arXiv preprint arXiv …, 2024 - arxiv.org
Deploying machine learning (ML) on diverse computing platforms is crucial to accelerate
and broaden their applications. However, it presents significant software engineering …
and broaden their applications. However, it presents significant software engineering …
OPASS: Orchestrating TVM's Passes for Lowering Memory Footprints of Computation Graphs
P Nie, Z Wang, C Wan, Z Lin, H Jiang… - 2024 IEEE …, 2024 - ieeexplore.ieee.org
Deep learning (DL) compilers, such as TVM and TensorFlow, encompass a variety of
passes for optimizing computation graphs (ie, DL models). Despite the efforts on developing …
passes for optimizing computation graphs (ie, DL models). Despite the efforts on developing …
Vortex under Ripplet: An Empirical Study of RAG-enabled Applications
Y Shao, Y Huang, J Shen, L Ma, T Su… - arXiv preprint arXiv …, 2024 - arxiv.org
Large language models (LLMs) enhanced by retrieval-augmented generation (RAG)
provide effective solutions in various application scenarios. However, developers face …
provide effective solutions in various application scenarios. However, developers face …