Browsing by Author "Rahman, Tahmid"

Now showing 1 - 2 of 2

Leveraging RLHF with Instruction Fine-tuning for Improving LLM Response in Bangla Conversations
(Department of Computer Science and Engineering(CSE), Islamic University of Technology(IUT), Board Bazar, Gazipur-1704, Bangladesh, 2024-11-30) Rahman, Tahmid; Mahmud, Shahriar; Nasrum, Nur
In the realm of Bangla conversational agents, this research endeavors to elevate the responsiveness of Large Language Models (LLMs) through the synergistic ap plication of Reinforcement Learning from Human Feedback (RLHF) and instruc tion fine-tuning. The primary objectives encompass the creation of a domain specific Bangla conversational dataset, an evaluation of existing LLMs using an instruction-tuned dataset, and the introduction of a novel human-centric bench marking framework. This work uses a multi-step process to improve the effi cacy of Large Language Models (LLMs) in the context of Bangla Conversational Agents. The technique consists of many key steps, each of which contributes to the refining and optimization of model performance. To address the shortage of domain-specific datasets for Bangla Conversational Agents, we start by construct ing a Bangla Conversational Dataset. The dataset is then fine-tuned with the use of an Instruction-Tuned Format. This structuring makes the data more suitable for training language models, allowing them to better comprehend and respond to precise commands in the Bangla conversational environment. Existing LLMs go through the next phase of our procedure, Supervised Fine-Tuning (SFT), us ing the instruction-tuned dataset. This fine-tuning procedure guarantees that the models are tailored to the variety and complexity of Bangla talks, maximizing their performance in accordance with the dataset’s unique instructions. Following fine-tuning, we do a detailed examination and comparison of the LLM models. This stage gives insight into the effectiveness of the fine-tuned models and enables the selection of the most promising candidate. This iterative procedure entails modifying the model with human feedback to improve its performance in a more dynamic and sophisticated way.
Power Flow Analysis Using Neural Networks
(Department of Electrical and Electronic Engineering (EEE), Islamic University of Technology(IUT), Board Bazar, Gazipur-1704, Bangladesh, 2025-10-25) Sayad, Rafsan; Akash, Al-Mahdi; Rahman, Tahmid
Solving power flow equations is a crucial part of power system state estimation. Due to the complexity of the relationship among the power system variables, power flow equations are solved mainly through iterative approaches. But due to the slow nature of the iterative approaches, it is only natural to examine alternatives for solving power flow equations. In this study, first, we have proceeded to solve power flow equations using a classical neural network based approach. The obvious shortcoming of a classical neural network is the amount of data required to properly train a model. So, secondly, we have proceeded to solve power flow equations using physics guided neural network based approaches in order to address the shortcoming of the classical neural network. We developed and examined two different physics guided neural networks models. Finally, we examined the feasibility of hybrid quantum classical neural networks in solving power flow equations and explored the applications of quantum neural networks in the field of power systems. Our study shows that classical neural networks surpasses iterative approaches in terms of speed by a big margin with reasonable accuracy. Our study also suggests that classical neural networks perform the best in terms of accuracy among the three neural network based approaches tried. Physics guided neural networks solve the problem of availability of data with decent accuracy. And the hybrid quantum classical neural network shows decent accuracy and a speed up in the training process.