Large Language Model 연구 및 개발 Internship

The role

EXAONE Lab conducts advanced research and development for large language models. This role involves research and development, experimentation, and applying cutting-edge AI technology to services.

What you will do

• Research and development and experimentation related to large language models
• Build training data collection and preprocessing pipelines
• Measure In-context learning and fine-tuning benchmark performance
• Prompt engineering for In-context learning
• Inference engine optimization
• Research and development of Multi-task instruction tuning and performance improvement techniques
• Research and development of Toxic classifier and large model safety methodologies
• Research reinforcement learning performance improvement methodologies for language models and distributed training development
• Research and development of large model compression methodologies

Required qualifications

• Undergraduate expected to graduate or graduated, or currently enrolled in Master's/PhD programs
• Deep learning, machine learning algorithm development experience is required
• Linux, Python, PyTorch, HuggingFace language model (BERT, T5, GPT, etc.) proficiency is required
• Computer science, computing, artificial intelligence, or machine learning major

Nice to have

• Language model pre-training development experience
• DeepSpeed, FastTransformer development experience
• Achieving high rankings on leaderboards, competition awards, open source activities
• ML production cycle (data collection, model development, model testing, serving) experience
• Large-scale data processing framework (Apache Beam, Amazon EMR, GCP Dataproc, etc.) development experience
• Language model related paper publication experience

About the company

EXAONE Lab is an organization that conducts advanced research and development related to large language models.