In this project, we aim to refine the capabilities of a pre-trained model known as hULMona, which was developed by a team of researchers including Obeida ElJundi, Wissam Antoun, Nour El Droubi, Hazem Hajj, Wassim El-Hajj, and Khaled Shaban. Our objective is to adapt hULMona to enhance its performance in specific applications, particularly in the realm of reviews classification.
Notebook: hULMona-ResClassification-v6.ipynb Purpose: This notebook utilizes the Fastai library to fine-tune the hULMona model on a dataset named Res.csv, which contains review data. The goal is to optimize the model's accuracy and efficiency in classifying the sentiment expressed in reviews.
Notebook: test-hULMona-Res2.ipynb
Following the fine-tuning process, this notebook is used to evaluate the performance of the updated model. The testing phase is crucial for assessing the effectiveness of the fine-tuning and for making any necessary adjustments to improve the model's classification capabilities. This structured approach ensures that the project is carried out in a systematic and professional manner, focusing on achieving the highest possible accuracy in sentiment classification tasks.