Select Git revision
benchmark_number_of_runs.py
-
Leon Michel Gorißen authored
- Added functionality to prepare data and execute training loops for various models (LLT, ITA). - Implemented data analysis and interpolation for trajectory data. - Integrated W&B sweeps to dynamically adjust training based on validation loss thresholds. - Enabled saving models upon reaching the loss threshold. - Improved logging and modular handling of multiple model configurations and UUIDs. - Removed old hardcoded training loops and replaced with flexible, scalable partial functions for different models.
Leon Michel Gorißen authored- Added functionality to prepare data and execute training loops for various models (LLT, ITA). - Implemented data analysis and interpolation for trajectory data. - Integrated W&B sweeps to dynamically adjust training based on validation loss thresholds. - Enabled saving models upon reaching the loss threshold. - Improved logging and modular handling of multiple model configurations and UUIDs. - Removed old hardcoded training loops and replaced with flexible, scalable partial functions for different models.