How to evaluate by iter

[wzr0108]

Here is my config

evaluation = dict(interval=100, metric='accuracy', by_epoch=False)
optimizer = dict(
    type='AdamW',
    lr=0.001,
    weight_decay=0.05,
    eps=1e-08,
    betas=(0.9, 0.999),
    paramwise_cfg=dict(
        norm_decay_mult=0.0,
        bias_decay_mult=0.0,
        custom_keys=dict({
            '.absolute_pos_embed': dict(decay_mult=0.0),
            '.relative_position_bias_table': dict(decay_mult=0.0)
        })))
optimizer_config = dict(grad_clip=dict(max_norm=5.0))
lr_config = dict(
    policy='CosineAnnealing',
    by_epoch=False,
    min_lr_ratio=0.01,
    warmup='linear',
    warmup_ratio=0.001,
    warmup_iters=20,
    warmup_by_epoch=True)
runner = dict(type='EpochBasedRunner', max_epochs=12)
checkpoint_config = dict(interval=1)
log_config = dict(
    interval=50, hooks=[dict(type='TextLoggerHook', by_epoch=False),
                        dict(type='TensorboardLoggerHook')])
dist_params = dict(backend='nccl')
log_level = 'INFO'
load_from = None
resume_from = None
workflow = [('train', 1)]
work_dir = './work_dirs/debug'

But it did not evaluate at100 iter

[Ezra-Yu]

You are using the mmcls-0.x branch. Sorry that in this branch, it is not possible to evaluate by iter.

Welcome to try the mmpretain. you can modify it in

mmpretrain/configs/_base_/schedules/imagenet_bs256_200e_coslr_warmup.py

Lines 28 to 30 in e80418a

	train_cfg = dict(by_epoch=True, max_epochs=200, val_interval=1)
	val_cfg = dict()
	test_cfg = dict()

.

val_cfg = dict(by_epoch=False, ........)

refer to https://github.com/open-mmlab/mmengine/blob/7877f8515ffff98f24ad352d352f591dada8f33a/mmengine/runner/loops.py#L194