用自己的数据集进行tsne可视化分析

[huadouxiansheng]

如题用自己的数据集写 t-SNE 可视化的数据配置文件，需要修改哪些配置，还是目前不支持imagenet数据集外的数据集
我尝试修改
dataset_type = 'mmcls.CustomDataset'
data_root = 'data/custom_dataset/'
file_client_args = dict(backend='disk')
name = 'custom_val'

extract_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='mmcls.ResizeEdge', scale=256, edge='short'),
dict(type='CenterCrop', crop_size=224),
dict(type='PackSelfSupInputs'),
]

extract_dataloader = dict(
batch_size=8,
num_workers=4,
dataset=dict(
type=dataset_type,
data_root='data/custom_dataset/val',
# ann_file='meta/val.txt',
data_prefix='val',
pipeline=extract_pipeline),
sampler=dict(type='DefaultSampler', shuffle=False),
)

pooling cfg

pool_cfg = dict(type='MultiPooling', in_indices=(1, 2, 3, 4))

但是显然有错

usage: visualize_tsne.py [-h] [--checkpoint CHECKPOINT] [--work-dir WORK_DIR]
[--launcher {none,pytorch,slurm,mpi}]
[--dataset-config DATASET_CONFIG]
[--max-num-class MAX_NUM_CLASS] [--seed SEED]
[--deterministic]
[--cfg-options CFG_OPTIONS [CFG_OPTIONS ...]]
[--n-components N_COMPONENTS]
[--perplexity PERPLEXITY]
[--early-exaggeration EARLY_EXAGGERATION]
[--learning-rate LEARNING_RATE] [--n-iter N_ITER]
[--n-iter-without-progress N_ITER_WITHOUT_PROGRESS]
[--init INIT]
config
visualize_tsne.py: error: unrecognized arguments: --dataset_config configs/benchmarks/classification/tsne_custom.py

[fangyixiao18]

https://mmselfsup.readthedocs.io/en/1.x/user_guides/visualization.html#visualize-t-sne
this is the newest version of tsne