Is there a framework for managing, running, and recording experiments with different architectures and hyperparameters?

1

Is there a recommended framework for managing, running, and recording experiments with different architectures and hyperparameters, or should I roll one from scratch for whatever I’m doing?

I mean a free and open source framework like pytorch.

2

Optuna or Weights & Biases would be two frameworks which come to my mind.

1 Like