How do I fix an `anaconda 400 error` during a sweep?

An anaconda 400 error during a sweep often means you didn’t log the metric you’re optimizing. You might see:

wandb: ERROR Error while calling W&B API: anaconda 400 error:
{"code": 400, "message": "TypeError: bad operand type for unary -: 'NoneType'"}

The sweep controller can’t find the metric it needs to compare runs. Log the exact metric named in your sweep configuration. For example, if your sweep config specifies:

metric:
  name: validation_loss
  goal: minimize

Add a matching wandb.log call in your training loop:

# After you compute validation_loss for an epoch or step:
# wandb.log({"validation_loss": value})

The metric name must match exactly between your sweep configuration and your wandb.log call. For more information, see Sweeps troubleshooting and Define sweep configuration.

Sweeps Metrics