def test_torch_l1filter_pruner(self):
"""
Filters with the minimum sum of the weights' L1 norm are pruned in this paper:
PRUNING FILTERS FOR EFFICIENT CONVNETS,
https://arxiv.org/abs/1608.08710
So if sparsity is 0.2, the expected masks should mask out filter 0, this can be verified through:
`all(torch.sum(mask1, (1, 2, 3)).numpy() == np.array([0., 27., 27., 27., 27.]))`
If sparsity is 0.6, the expected masks should mask out filter 0,1,2, this can be verified through:
`all(torch.sum(mask2, (1, 2, 3)).numpy() == np.array([0., 0., 0., 27., 27.]))`
"""
w = np.array([np.zeros((3, 3, 3)), np.ones((3, 3, 3)), np.ones((3, 3, 3)) * 2,
np.ones((3, 3, 3)) * 3, np.ones((3, 3, 3)) * 4])
model = TorchModel()
config_list = [{'sparsity': 0.2, 'op_names': ['conv1']}, {'sparsity': 0.6, 'op_names': ['conv2']}]
pruner = torch_compressor.L1FilterPruner(model, config_list)
model.conv1.weight.data = torch.tensor(w).float()
model.conv2.weight.data = torch.tensor(w).float()
layer1 = torch_compressor.compressor.LayerInfo('conv1', model.conv1)
mask1 = pruner.calc_mask(layer1, config_list[0])
layer2 = torch_compressor.compressor.LayerInfo('conv2', model.conv2)
mask2 = pruner.calc_mask(layer2, config_list[1])
assert all(torch.sum(mask1, (1, 2, 3)).numpy() == np.array([0., 27., 27., 27., 27.]))
assert all(torch.sum(mask2, (1, 2, 3)).numpy() == np.array([0., 0., 0., 27., 27.]))
def test_torch_l1filter_pruner(self):
"""
Filters with the minimum sum of the weights' L1 norm are pruned in this paper:
PRUNING FILTERS FOR EFFICIENT CONVNETS,
https://arxiv.org/abs/1608.08710
So if sparsity is 0.2, the expected masks should mask out filter 0, this can be verified through:
`all(torch.sum(mask1, (1, 2, 3)).numpy() == np.array([0., 27., 27., 27., 27.]))`
If sparsity is 0.6, the expected masks should mask out filter 0,1,2, this can be verified through:
`all(torch.sum(mask2, (1, 2, 3)).numpy() == np.array([0., 0., 0., 27., 27.]))`
"""
w = np.array([
np.zeros((3, 3, 3)),
np.ones((3, 3, 3)),
np.ones((3, 3, 3)) * 2,
np.ones((3, 3, 3)) * 3,
np.ones((3, 3, 3)) * 4
])
model = TorchModel()
optimizer = torch.optim.SGD(model.parameters(), lr=0.01, momentum=0.5)
config_list = [{
'sparsity': 0.2,
'op_types': ['Conv2d'],
'op_names': ['conv1']
}, {
'sparsity': 0.6,
'op_types': ['Conv2d'],
'op_names': ['conv2']
}]
pruner = torch_compressor.L1FilterPruner(model, config_list, optimizer)
model.conv1.module.weight.data = torch.tensor(w).float()
model.conv2.module.weight.data = torch.tensor(w).float()
mask1 = pruner.calc_mask(model.conv1)
mask2 = pruner.calc_mask(model.conv2)
assert all(
torch.sum(mask1['weight_mask'], (
1, 2, 3)).numpy() == np.array([0., 27., 27., 27., 27.]))
assert all(
torch.sum(mask2['weight_mask'], (
1, 2, 3)).numpy() == np.array([0., 0., 0., 27., 27.]))
def test_torch_l1filter_pruner(self):
"""
Filters with the minimum sum of the weights' L1 norm are pruned in this paper:
PRUNING FILTERS FOR EFFICIENT CONVNETS,
https://arxiv.org/abs/1608.08710
So if sparsity is 0.2 for conv1, the expected masks should mask out filter 0, this can be verified through:
`all(torch.sum(mask1, (1, 2, 3)).numpy() == np.array([0., 25., 25., 25., 25.]))`
If sparsity is 0.6 for conv2, the expected masks should mask out filter 0,1,2, this can be verified through:
`all(torch.sum(mask2, (1, 2, 3)).numpy() == np.array([0., 0., 0., 0., 0., 0., 125., 125., 125., 125.]))`
"""
w1 = np.array([np.ones((1, 5, 5)) * i
for i in range(5)]).astype(np.float32)
w2 = np.array([np.ones((5, 5, 5)) * i
for i in range(10)]).astype(np.float32)
model = TorchModel()
config_list = [{
'sparsity': 0.2,
'op_types': ['Conv2d'],
'op_names': ['conv1']
}, {
'sparsity': 0.6,
'op_types': ['Conv2d'],
'op_names': ['conv2']
}]
pruner = torch_compressor.L1FilterPruner(model, config_list)
model.conv1.module.weight.data = torch.tensor(w1).float()
model.conv2.module.weight.data = torch.tensor(w2).float()
mask1 = pruner.calc_mask(model.conv1)
mask2 = pruner.calc_mask(model.conv2)
assert all(
torch.sum(mask1['weight_mask'], (
1, 2, 3)).numpy() == np.array([0., 25., 25., 25., 25.]))
assert all(
torch.sum(mask2['weight_mask'], (1, 2, 3)).numpy() == np.array(
[0., 0., 0., 0., 0., 0., 125., 125., 125., 125.]))