MachineUnlearning.Config(
        study_name='security_machine_unlearning',
        experiment_name='exp000',
        doc='',
        base_dir='/opt/scratch',
        seed=42,
        device=None,
        dataset=UnlearningData.Config(
            base_dir='/opt/scratch',
            working_dir=PosixPath('/opt/scratch/datasets/cifar10'),
            batch_size=128,
            eval_batch_size=128,
            num_streams=4,
            translate_pad=4,
            mean=(0.4914, 0.4822, 0.4465),
            std=(0.2023, 0.1994, 0.201),
            augment=PartialConfig(
                priml.baselines.mlsbench.dl.vision.data.pad_crop_flip_normalize
            ),
            forget_class=0
        ),
        pretrain=VisionTrainStep.Config(
        │   model=ResNet.Config(
        │       channels_in=3,
        │       channels_out=10,
        │       image_size=32,
        │       channels_hidden=(16, 32, 64),
        │       blocks_per_stage=(3, 3, 3),
        │       block=BasicBlock.Config(
        │           channels_in=-1,
        │           channels_out=-1,
        │           stride=1,
        │           activation=priml.model.swiglu.relu,
        │           norm=<class 'priml.baselines.mlsbench.dl.vision.model.BatchNorm'>
        │       ),
        │       activation=priml.model.swiglu.relu,
        │       norm=<class 'priml.baselines.mlsbench.dl.vision.model.BatchNorm'>,
        │       pool=GlobalAvgPool.Config(),
        │       init_conv=priml.baselines.mlsbench.dl.vision.model.kaiming_normal_fan_out,
        │       init_head=priml.baselines.mlsbench.dl.vision.model.kaiming_normal_fan_in,
        │       init_bias=torch.nn.init.zeros_,
        │       init=priml.baselines.mlsbench.dl.vision.model.keep_initialization
        │   ),
        │   optimizer=PartialConfig(
        │       <class 'torch.optim.sgd.SGD'>,
        │       lr=0.1,
        │       momentum=0.9,
        │       weight_decay=0.0005
        │   ),
        │   loss=PartialConfig(
        │       priml.baselines.mlsbench.dl.vision.train_step.cross_entropy
        │   ),
        │   accumulate_grad_batches=1,
        │   drop_partial_accumulation_on_epoch_end=True,
        │   train_budget_steps=inf,
        │   train_budget_sec=inf,
        │   train_budget_epochs=80,
        │   lr_schedule=PartialConfig(priml.math.schedules.constant),
        │   parallelism=NoParallel.Config(device=None),
        │   model_quantization=NoModelQuantization.Config(),
        │   activation_memoization=DefaultActivationStorage.Config(),
        │   compile=None,
        │   ema=NoEMA.Config(),
        │   gradient_clip_norm=inf,
        │   skip_step_on_nonfinite_grad=False,
        │   device_init='eager',
        │   dtype_autocast=None,
        │   autocast_cache_enabled=False,
        │   schedule=PartialConfig(
        │       priml.baselines.mlsbench.dl.vision.train_step.cosine_annealing
        │   )
        ),
        unlearn_epochs=20,
        unlearn_optimizer=PartialConfig(<class 'torch.optim.adam.Adam'>, lr=0.001),
        method=RetainFinetune.Config()
    )
