Hi, I’m trying to create the following sweep:
name: ResNet
method: random
metric:
name: val_loss
goal: minimize
parameters:
learning_rate:
distribution: log_uniform_values
min: 3e-5
max: 3e-3
batch_size:
values: [128, 256, 512]
epochs:
values: [200]
filters_1:
values: [32, 64, 128]
kernel_size_1:
values: [3, 5, 7]
filters_2:
values: [32, 64, 128]
kernel_size_2:
values: [3, 5, 7]
filters_3:
values: [32, 64, 128]
kernel_size_3:
values: [3, 5, 7]
pooling:
values: [‘max’, ‘average’]
dense_units:
values: [0, 128, 256]
When running: 'wandb sweep --entity <> —project <> sweep.yaml I get the following error:
wandb: ERROR Error while calling W&B API: Invalid sweep config: invalid hyperparameter configuration: learning_rate (<Response [400]>)
Error: Invalid sweep config: invalid hyperparameter configuration: learning_rate
Last week I was able to create sweeps with the same configuration for learning_rate …
I would appreciate your help to solve it. Thanks