Notifications
Clear all
Weights and Biases Reviews
1
Posts
1
Users
0
Reactions
0
Views
Topic starter
16/07/2026 5:06 am
Hey everyone! I've been learning about Weights & Biases for ML experiment tracking, and it looks super slick. But I'm wondering about its flexibility.
My team runs a lot of complex simulations (think physics or agent-based models) that aren't traditional ML training runs. We need to track parameters, metrics over time, and log outputs. Has anyone here used W&B successfully for something like that? I'm curious if the existing tools (like `wandb.log` for metrics or logging artifacts) are a good fit, or if it feels like forcing a square peg into a round hole.
Also, if you've done it, any tips on setting it up? Like, do you just treat each simulation run as an "experiment"? Would love to hear about real use cases! 😊