Currently, I'm an NLP Researcher at Huawei Turkey R&D Center. I specialize in Natural Language Processing and have a background in Computer Science from Yildiz Technical University. I'm passionate about AI research and have contributed to a TUBITAK-funded project focusing on Curriculum Learning.
- π At Huawei, I curate and synthesize data for fine-tuning large language models in multiple languages, leveraging model and data parallelism with mixed precision. I've also enhanced finetuning processes and created a versatile model serving platform.
- π± I'm continually learning and improving my skills, with recent interests in reinforcement learning and advanced aspects of efficient deep learning model finetuning and serving techniques.
- π¬ Feel free to ask me about my research, projects, or my current role at Huawei. I'm always eager to share and learn.
- π« How to reach me: You can send me an email at mrbesheralkurdi@outlook.com.
Here are some of my academic publications:
- Scaling Up Paraphrase Generation Datasets with Machine Translation and Semantic Similarity Filtering
- Semantic Similarity Based Filtering for Turkish Paraphrase Dataset Creation
- Linguistic-based Data Augmentation Approach for Offensive Language Detection
I love experimenting with things. Here are a few projects I've worked on recently:
Find me on Youtube, Kaggle, and LinkedIn. I'd love to connect and collaborate!