human_ml3d_bs128.py

# dataset settings
data_keys = ['motion', 'motion_mask', 'motion_length', 'clip_feat']
meta_keys = ['text', 'token']
train_pipeline = [
    dict(type='Normalize',
         mean_path='data/datasets/human_ml3d/mean.npy',
         std_path='data/datasets/human_ml3d/std.npy'),
    dict(type='Crop', crop_size=196),
    dict(type='ToTensor', keys=data_keys),
    dict(type='Collect', keys=data_keys, meta_keys=meta_keys)
]

data = dict(
    samples_per_gpu=128,
    workers_per_gpu=1,
    train=dict(type='RepeatDataset',
               dataset=dict(
                   type='TextMotionDataset',
                   dataset_name='human_ml3d',
                   data_prefix='data',
                   pipeline=train_pipeline,
                   ann_file='train.txt',
                   motion_dir='motions',
                   text_dir='texts',
                   token_dir='tokens',
                   clip_feat_dir='clip_feats',
               ),
               times=100),
    test=dict(type='TextMotionDataset',
              dataset_name='human_ml3d',
              data_prefix='data',
              pipeline=train_pipeline,
              ann_file='test.txt',
              motion_dir='motions',
              text_dir='texts',
              token_dir='tokens',
              clip_feat_dir='clip_feats',
              eval_cfg=dict(
                  shuffle_indexes=True,
                  replication_times=20,
                  replication_reduction='statistics',
                  evaluator_model=dict(
                      type='T2MContrastiveModel',
                      motion_encoder=dict(
                          input_size=263,
                          movement_hidden_size=512,
                          movement_latent_size=512,
                          motion_hidden_size=1024,
                          motion_latent_size=512,
                      ),
                      text_encoder=dict(word_size=300,
                                        pos_size=15,
                                        hidden_size=512,
                                        output_size=512,
                                        max_text_len=20),
                      init_cfg=dict(
                          type='Pretrained',
                          checkpoint='data/evaluators/human_ml3d/finest.tar')),
                  metrics=[
                      dict(type='R Precision', batch_size=32, top_k=3),
                      dict(type='Matching Score', batch_size=32),
                      dict(type='FID'),
                      dict(type='Diversity', num_samples=300),
                      dict(type='MultiModality',
                           num_samples=100,
                           num_repeats=30,
                           num_picks=10)
                  ]),
              test_mode=True))