18 Nov
|
Arkion Identity Systems
|
Toronto
18 Nov
Arkion Identity Systems
Toronto
Develop machine learning evaluations remotely or hybrid in the Greater Toronto Area, using open-weight LLMs, datasets, and real GPUs. Measure whether training changes improve security and identity tasks.
This Fall/Winter internship focuses on taking models from raw data to measured results. You will work with PyTorch, Transformers, Datasets, and PEFT alongside the founding team. Your work includes dataset preparation, fine-tuning, evaluation design, and explicit reporting.
Key Responsibilities:
- Build evaluation sets and task-relevant benchmarks
- Prepare datasets for training and evaluation
- Run fine-tuning jobs with parameter-efficient methods
- Monitor GPU training and limited compute usage
- Report comparisons between models and training runs
Requirements:
- Hands-on experience adapting open-weight large language models
- Proficiency with PyTorch and Hugging Face ecosystem tools
- Knowledge of GPU memory and batch-size fundamentals
- Understanding of mixed-precision training
- Experience preparing datasets for model evaluation
Use machine learning experimentation, benchmark design, and model comparison to support security and identity applications.
#J-18808-Ljbffr
📌 Machine Learning Research Intern LLMs (Toronto)
🏢 Arkion Identity Systems
📍 Toronto