Fine-tune Llama 3.18B in legal domain to classify arguments in court decisions

Seeking an experienced NLP and ML engineer to fine-tune our pretrained Czech legal LLM (Llama 3.1 8B) to pursue multi-label classififcation on a paragraph level in legal domain. We have: 1) Pretrained Llama 3.1 8B on 300k Czech court decisions 2) 270 annotated Czech court decisions (9k paragraphs) for fine-tuning 3) 230k Czech court decisions ready for analysis 4) All data preprocessed and ready to use Your task: 1) Fine-tune the Llama model on our annotated data 2) Use the model for the 230k decisions 3) Provide complete documentation and statistics Required: Experience with LLM fine-tuning and legal NLP Note: We reserve the right to withdraw this position at any time and to select or reject any applicants at our discretion. This posting does not constitute an obligation to fill the position.