Upload aagcn/j_pretrained_otp2.py with huggingface_hub

aagcn/j_pretrained_otp2.py

@@ -0,0 +1,100 @@
+model = dict(
+    type='RecognizerGCN',
+    backbone=dict(type='AAGCN', graph_cfg=dict(layout='coco', mode='spatial')),
+    cls_head=dict(type='GCNHead', num_classes=28, in_channels=256))
+dataset_type = 'PoseDataset'
+train_pipeline = [
+    dict(type='PreNormalize2D'),
+    dict(type='GenSkeFeat', dataset='coco', feats=['j']),
+    dict(type='UniformSample', clip_len=100),
+    dict(type='PoseDecode'),
+    dict(type='FormatGCNInput', num_person=2),
+    dict(type='Collect', keys=['keypoint', 'label'], meta_keys=[]),
+    dict(type='ToTensor', keys=['keypoint'])
+]
+val_pipeline = [
+    dict(type='PreNormalize2D'),
+    dict(type='GenSkeFeat', dataset='coco', feats=['j']),
+    dict(type='UniformSample', clip_len=100, num_clips=1),
+    dict(type='PoseDecode'),
+    dict(type='FormatGCNInput', num_person=2),
+    dict(type='Collect', keys=['keypoint', 'label'], meta_keys=[]),
+    dict(type='ToTensor', keys=['keypoint'])
+]
+test_pipeline = [
+    dict(type='PreNormalize2D'),
+    dict(type='GenSkeFeat', dataset='coco', feats=['j']),
+    dict(type='UniformSample', clip_len=100, num_clips=10),
+    dict(type='PoseDecode'),
+    dict(type='FormatGCNInput', num_person=2),
+    dict(type='Collect', keys=['keypoint', 'label'], meta_keys=[]),
+    dict(type='ToTensor', keys=['keypoint'])
+]
+data = dict(
+    videos_per_gpu=16,
+    workers_per_gpu=1,
+    persistent_workers=True,
+    test_dataloader=dict(videos_per_gpu=1),
+    train=dict(
+        type='RepeatDataset',
+        times=5,
+        dataset=dict(
+            type='PoseDataset',
+            ann_file='data/pyskl_otp2.pkl',
+            pipeline=[
+                dict(type='PreNormalize2D'),
+                dict(type='GenSkeFeat', dataset='coco', feats=['j']),
+                dict(type='UniformSample', clip_len=100),
+                dict(type='PoseDecode'),
+                dict(type='FormatGCNInput', num_person=2),
+                dict(type='Collect', keys=['keypoint', 'label'], meta_keys=[]),
+                dict(type='ToTensor', keys=['keypoint'])
+            ],
+            split='train')),
+    val=dict(
+        type='PoseDataset',
+        ann_file='data/pyskl_otp2.pkl',
+        pipeline=[
+            dict(type='PreNormalize2D'),
+            dict(type='GenSkeFeat', dataset='coco', feats=['j']),
+            dict(type='UniformSample', clip_len=100, num_clips=1),
+            dict(type='PoseDecode'),
+            dict(type='FormatGCNInput', num_person=2),
+            dict(type='Collect', keys=['keypoint', 'label'], meta_keys=[]),
+            dict(type='ToTensor', keys=['keypoint'])
+        ],
+        split='val'),
+    test=dict(
+        type='PoseDataset',
+        ann_file='data/pyskl_otp2.pkl',
+        pipeline=[
+            dict(type='PreNormalize2D'),
+            dict(type='GenSkeFeat', dataset='coco', feats=['j']),
+            dict(type='UniformSample', clip_len=100, num_clips=10),
+            dict(type='PoseDecode'),
+            dict(type='FormatGCNInput', num_person=2),
+            dict(type='Collect', keys=['keypoint', 'label'], meta_keys=[]),
+            dict(type='ToTensor', keys=['keypoint'])
+        ],
+        split='test'))
+optimizer = dict(
+    type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005, nesterov=True)
+optimizer_config = dict(grad_clip=None)
+lr_config = dict(policy='CosineAnnealing', min_lr=0, by_epoch=False)
+total_epochs = 24
+checkpoint_config = dict(interval=2)
+workflow = [('train', 1)]
+evaluation = dict(
+    interval=1, metrics=['top_k_accuracy', 'mean_class_accuracy'], topk=(1, 5))
+log_config = dict(interval=100, hooks=[dict(type='TextLoggerHook')])
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+work_dir = './work_dirs/aagcn/otp2'
+seed = 42
+deterministic = True
+cudnn_benchmark = False
+cudnn_deterministic = True
+load_from = 'https://download.openmmlab.com/mmaction/pyskl/ckpt/aagcn/aagcn_pyskl_ntu120_xsub_hrnet/j.pth'
+resume_from = None
+find_unused_parameters = False
+gpu_ids = range(0, 1)