Search
8 articles for “Parameter- Efficient Fine-Tuning”
-
A Reviewed Study On Cpu-Optimized Parameter-Efficient Fine- Tuning For Large Language Models To Increase Accuracy Using Lora
Abstract: The fast proliferation of Large Language Models (LLMs) has increased the need to optimize the process of fine-tuning but the existing workflows that require a GPU are still expensive, intensive, and unavailable to most researchers. This paper is driven by the desire to have a more cost-efficient and democratized version by examining a CPU-efficient implementation of Parameter-Efficient Fine-Tuning (PEFT) based on Low-Rank Adaptation (LoRA). The major purpose of the study …
Published in Recent Trends in Parallel Computing · Vol. 13, Issue 1, 2026 Read article
-
Optimal Homotopy Analysis Method (OHAM) For the Approximate Series Solution of Non-linear Partial Differential Equation
Abstract: In this article, we have used the Optimal Homotopy Analysis Method (OHAM), which is a basically semi-analytic method to solve differential equations. The ability for the user to choose the convergence control parameter, auxiliary linear operator, auxiliary function, and starting approximation is what sets apart the OHAM technique. We guaranteed the efficacy and efficiency of the procedure by fine-tuning the convergence control parameter. We solved a non-linear partial differential equation …
Published in Research & Reviews: Discrete Mathematical Structures · Vol. 11, Issue 1, 2024 Read article
-
Stacked Generalization-Based Deep Learning Approach for Pneumonia Detection
Abstract: The proposed work focuses on a stacked generalization-based approach for diagnosing pneumonia from chest X-ray images. It utilizes regularization, early stopping, and data augmentation to deal with overfitting. It uses safe level SMOTE to deal with class imbalance and attention-based feature fusion to adaptively weigh features based on their importance. It uses two publicly available datasets (RSNA and Kermany) with ground truth provided by expert radiologists. The proposed work used …
Published in Journal of Image Processing & Pattern Recognition Progress · Vol. 12, Issue 3, 2025 · pp. 20–31 Read article
-
Optimization of Fuel Injection Parameters in LHR Engines Using Fish Oil Methyl Ester Blends for Enhanced Performance and Emission Reduction
Abstract: The need for alternative fuels has become increasingly critical due to the depletion of petroleum resources, growing automobile usage, and environmental concerns. Fish oil methyl ester, derived from fish oil, presents a promising alternative biodiesel, which can be produced from both edible and non-edible oils as well as animal fats. Diesel engines are valued for their efficiency, reliability, and durability, with performance and emissions being influenced significantly by factors such …
Published in Journal of Thermal Engineering and Applications · Vol. 11, Issue 3, 2024 · pp. 18–35 Read article
-
A Review on Parametric Optimization of WEDM Technique for OHNS Steel
Abstract: In this study, the Wire Electrical Discharge Machining (WEDM) process for OHNS (Oil Hardened Non-Shrinking) steel, a high-performance material frequently used in the production of dies, punches, and precision tooling components, is optimized parametrically and validated experimentally. A continuously moving wire electrode and a sequence of electrical discharges are used in WEDM, a non-traditional machining method, to erode material and produce intricate and precise profiles, particularly in materials that are …
Published in Trends in Mechanical Engineering & Technology · Vol. 15, Issue 2, 2025 · pp. 29–35 Read article
-
Performance Analysis of Deep CNN Architectures
Abstract: A Convolutional Neural Network (CNN) is an artificial neural network renowned for its remarkable ability to handle large image datasets effectively, particularly excelling in tasks such as image recognition and classification. The fundamental structure of a CNN relies on mathematical convolution operations, comprising essential components such as convolutional layers, activation functions, pooling layers, and fully connected layers. These components work synergistically to extract and learn hierarchical features from input data, …
Published in Journal of Communication Engineering & Systems · Vol. 14, Issue 2, 2024 · pp. 1–8 Read article
-
Performance Analysis of Dual Junction Solar Cell Devices utilizing Subcells of In0.51Ga0.49P and GaAs to Study Key Solar Cell Parameters via TCAD based Simulation
Abstract: In the present work, a multi-junction solar cell was designed to obtain better performance over single-junction solar cells. The proposed structure is composed of different layers of diverse semiconductor materials stacked on each other. Here, an In0.51Ga0.49P/GaAs double-junction solar cell was outlined as having a distinctive composite material (GaAs) as the tunneling junction. To optimize the solar cell's effectiveness, In0.47Ga0.15Al0.37P was used in the window layer and the back-surface field …
Published in Journal of Polymer & Composites · Vol. 13, Issue 1, 2025 · pp. 829–837 Read article
-
Improving Polymer Composite Properties Through Reinforcement Learning Guided Prototyping A Novel Approach for Material Engineering
Abstract: Innovative approaches integrating reinforcement learning (RL) and machine learning (ML) into the fields of polymer composite prototyping and soft actuator manufacturing for applications. This new an algorithm utilizing RL optimizes polymer composite fabrication parameters to enhance material properties efficiently. By iteratively adjusting parameters based on predefined objectives, the RL agent guides the prototyping process, promising to revolutionize polymer composite engineering. A finest control method for locked loop control of Shape …
Published in Journal of Polymer & Composites · Vol. 12, Issue 4, 2024 · pp. 208–218 Read article