Tasks:
- * Analyze success rate, convergence, generalization, and real-robot data needs
- * Compare training outcomes across data mixtures
- * Design and run mid-training and post-training experiments
- * Develop action benchmarks and post-training evaluations
- * Maintain training configurations, logs, checkpoints, and evaluation results
- * Organize ego, UMI, teleoperation, and real-robot training data
- * Write experimental reports
Perks/Benefits:
Skills/Tech stack required:
[Ablation Studies] [Computer Vision] [Conda] [Controlled experiments] [Deep learning] [Diffusion Policy] [Docker] [Fine Tuning] [Git] [Imitation Learning] [Linux] [Log Analysis] [Model Fine-tuning] [Multimodal Learning] [PyTorch] [Reinforcement Learning] [Robot Learning] [Training Log Analysis]
Educational requirements:
N/A
Role(s):
[Intern] [Learning Intern] [Machine Learning Intern] [Robotics Machine Learning Intern]