parameter_grid,
# run_ids=run_ids,
n_samples_eval=50,
timeout_train=100,
n_gpus_train=batch_size if batch_size else 1,
timeout_eval=10,
n_gpus_eval=1,
# n_samples=50,
# timeout=10,
# n_gpus=1,
to_grid=False,
return_run_ids=True,
checkpoints_train=3,
params_to_ignore=['batch_size', 'use_mixed_precision', 'model_size'],
)
print(eval_results)
infer_grid(
job_name,
xpdnet_inference,
parameter_grid,
run_ids=run_ids,
timeout=10,
n_gpus=4,
to_grid=False,
params_to_ignore=[
'mask_type', 'batch_size', 'use_mixed_precision', 'model_size'
],
)
] + [
dict(
model_fun=model_fun,
model_kwargs=kwargs,
multicoil=False,
n_scales=n_scales,
res=res,
n_primal=n_primal,
contrast=contrast,
n_epochs=n_epochs,
refine_smaps=refine_smaps,
refine_big=refine_big,
af=4,
n_dual=n_dual,
n_iter=n_iter_per_model_size[model_size],
primal_only=primal_only,
n_dual_filters=n_dual_filters,
) for model_name, model_size, model_fun, kwargs, _, n_scales, res in
model_specs if model_size == 'big' and 'MWCNN' not in model_name
]
infer_grid(
job_name,
xpdnet_inference,
parameter_grid,
run_ids=run_ids,
timeout=10,
n_gpus=4,
to_grid=False,
)