XResNet1d

This is a modified version of fastai’s XResNet model in github

---

source

xresnet1d50_deeper

 xresnet1d50_deeper (c_in, c_out, act=<class
                     'torch.nn.modules.activation.ReLU'>, stride=1,
                     groups=1, reduction=None, nh1=None, nh2=None,
                     dw=False, g2=1, sa=False, sym=False,
                     norm_type=<NormType.Batch: 1>, act_cls=<class
                     'torch.nn.modules.activation.ReLU'>, ndim=2, ks=3,
                     pool=<function AvgPool>, pool_first=True,
                     padding=None, bias=None, bn_1st=True,
                     transpose=False, init='auto', xtra=None,
                     bias_std=0.01, dilation:Union[int,Tuple[int,int]]=1,
                     padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d34_deeper

 xresnet1d34_deeper (c_in, c_out, act=<class
                     'torch.nn.modules.activation.ReLU'>, stride=1,
                     groups=1, reduction=None, nh1=None, nh2=None,
                     dw=False, g2=1, sa=False, sym=False,
                     norm_type=<NormType.Batch: 1>, act_cls=<class
                     'torch.nn.modules.activation.ReLU'>, ndim=2, ks=3,
                     pool=<function AvgPool>, pool_first=True,
                     padding=None, bias=None, bn_1st=True,
                     transpose=False, init='auto', xtra=None,
                     bias_std=0.01, dilation:Union[int,Tuple[int,int]]=1,
                     padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d18_deeper

 xresnet1d18_deeper (c_in, c_out, act=<class
                     'torch.nn.modules.activation.ReLU'>, stride=1,
                     groups=1, reduction=None, nh1=None, nh2=None,
                     dw=False, g2=1, sa=False, sym=False,
                     norm_type=<NormType.Batch: 1>, act_cls=<class
                     'torch.nn.modules.activation.ReLU'>, ndim=2, ks=3,
                     pool=<function AvgPool>, pool_first=True,
                     padding=None, bias=None, bn_1st=True,
                     transpose=False, init='auto', xtra=None,
                     bias_std=0.01, dilation:Union[int,Tuple[int,int]]=1,
                     padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d50_deep

 xresnet1d50_deep (c_in, c_out, act=<class
                   'torch.nn.modules.activation.ReLU'>, stride=1,
                   groups=1, reduction=None, nh1=None, nh2=None, dw=False,
                   g2=1, sa=False, sym=False, norm_type=<NormType.Batch:
                   1>, act_cls=<class 'torch.nn.modules.activation.ReLU'>,
                   ndim=2, ks=3, pool=<function AvgPool>, pool_first=True,
                   padding=None, bias=None, bn_1st=True, transpose=False,
                   init='auto', xtra=None, bias_std=0.01,
                   dilation:Union[int,Tuple[int,int]]=1,
                   padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d34_deep

 xresnet1d34_deep (c_in, c_out, act=<class
                   'torch.nn.modules.activation.ReLU'>, stride=1,
                   groups=1, reduction=None, nh1=None, nh2=None, dw=False,
                   g2=1, sa=False, sym=False, norm_type=<NormType.Batch:
                   1>, act_cls=<class 'torch.nn.modules.activation.ReLU'>,
                   ndim=2, ks=3, pool=<function AvgPool>, pool_first=True,
                   padding=None, bias=None, bn_1st=True, transpose=False,
                   init='auto', xtra=None, bias_std=0.01,
                   dilation:Union[int,Tuple[int,int]]=1,
                   padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d18_deep

 xresnet1d18_deep (c_in, c_out, act=<class
                   'torch.nn.modules.activation.ReLU'>, stride=1,
                   groups=1, reduction=None, nh1=None, nh2=None, dw=False,
                   g2=1, sa=False, sym=False, norm_type=<NormType.Batch:
                   1>, act_cls=<class 'torch.nn.modules.activation.ReLU'>,
                   ndim=2, ks=3, pool=<function AvgPool>, pool_first=True,
                   padding=None, bias=None, bn_1st=True, transpose=False,
                   init='auto', xtra=None, bias_std=0.01,
                   dilation:Union[int,Tuple[int,int]]=1,
                   padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d152

 xresnet1d152 (c_in, c_out, act=<class
               'torch.nn.modules.activation.ReLU'>, stride=1, groups=1,
               reduction=None, nh1=None, nh2=None, dw=False, g2=1,
               sa=False, sym=False, norm_type=<NormType.Batch: 1>,
               act_cls=<class 'torch.nn.modules.activation.ReLU'>, ndim=2,
               ks=3, pool=<function AvgPool>, pool_first=True,
               padding=None, bias=None, bn_1st=True, transpose=False,
               init='auto', xtra=None, bias_std=0.01,
               dilation:Union[int,Tuple[int,int]]=1,
               padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d101

 xresnet1d101 (c_in, c_out, act=<class
               'torch.nn.modules.activation.ReLU'>, stride=1, groups=1,
               reduction=None, nh1=None, nh2=None, dw=False, g2=1,
               sa=False, sym=False, norm_type=<NormType.Batch: 1>,
               act_cls=<class 'torch.nn.modules.activation.ReLU'>, ndim=2,
               ks=3, pool=<function AvgPool>, pool_first=True,
               padding=None, bias=None, bn_1st=True, transpose=False,
               init='auto', xtra=None, bias_std=0.01,
               dilation:Union[int,Tuple[int,int]]=1,
               padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d50

 xresnet1d50 (c_in, c_out, act=<class 'torch.nn.modules.activation.ReLU'>,
              stride=1, groups=1, reduction=None, nh1=None, nh2=None,
              dw=False, g2=1, sa=False, sym=False,
              norm_type=<NormType.Batch: 1>, act_cls=<class
              'torch.nn.modules.activation.ReLU'>, ndim=2, ks=3,
              pool=<function AvgPool>, pool_first=True, padding=None,
              bias=None, bn_1st=True, transpose=False, init='auto',
              xtra=None, bias_std=0.01,
              dilation:Union[int,Tuple[int,int]]=1,
              padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d34

 xresnet1d34 (c_in, c_out, act=<class 'torch.nn.modules.activation.ReLU'>,
              stride=1, groups=1, reduction=None, nh1=None, nh2=None,
              dw=False, g2=1, sa=False, sym=False,
              norm_type=<NormType.Batch: 1>, act_cls=<class
              'torch.nn.modules.activation.ReLU'>, ndim=2, ks=3,
              pool=<function AvgPool>, pool_first=True, padding=None,
              bias=None, bn_1st=True, transpose=False, init='auto',
              xtra=None, bias_std=0.01,
              dilation:Union[int,Tuple[int,int]]=1,
              padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

---

source

xresnet1d18

 xresnet1d18 (c_in, c_out, act=<class 'torch.nn.modules.activation.ReLU'>,
              stride=1, groups=1, reduction=None, nh1=None, nh2=None,
              dw=False, g2=1, sa=False, sym=False,
              norm_type=<NormType.Batch: 1>, act_cls=<class
              'torch.nn.modules.activation.ReLU'>, ndim=2, ks=3,
              pool=<function AvgPool>, pool_first=True, padding=None,
              bias=None, bn_1st=True, transpose=False, init='auto',
              xtra=None, bias_std=0.01,
              dilation:Union[int,Tuple[int,int]]=1,
              padding_mode:str='zeros', device=None, dtype=None)

	Type	Default	Details
c_in
c_out
act	type	ReLU
stride	int	1
groups	int	1
reduction	NoneType	None
nh1	NoneType	None
nh2	NoneType	None
dw	bool	False
g2	int	1
sa	bool	False
sym	bool	False
norm_type	NormType	NormType.Batch
act_cls	type	ReLU
ndim	int	2
ks	int	3
pool	function	AvgPool
pool_first	bool	True
padding	NoneType	None
bias	NoneType	None
bn_1st	bool	True
transpose	bool	False
init	str	auto
xtra	NoneType	None
bias_std	float	0.01
dilation	typing.Union[int, typing.Tuple[int, int]]	1
padding_mode	str	zeros	TODO: refine this type
device	NoneType	None
dtype	NoneType	None

bs, c_in, seq_len = 2, 4, 32
c_out = 2
x = torch.rand(bs, c_in, seq_len)
archs = [
    xresnet1d18, xresnet1d34, xresnet1d50, 
    xresnet1d18_deep, xresnet1d34_deep, xresnet1d50_deep, xresnet1d18_deeper,
    xresnet1d34_deeper, xresnet1d50_deeper
#     # Long test
#     xresnet1d101, xresnet1d152,
]
for i, arch in enumerate(archs):
    print(i, arch.__name__)
    test_eq(arch(c_in, c_out, sa=True, act=Mish)(x).shape, (bs, c_out))

0 xresnet1d18
1 xresnet1d34
2 xresnet1d50
3 xresnet1d18_deep
4 xresnet1d34_deep
5 xresnet1d50_deep
6 xresnet1d18_deeper
7 xresnet1d34_deeper
8 xresnet1d50_deeper

m = xresnet1d34(4, 2, act=Mish)
test_eq(len(get_layers(m, is_bn)), 38)
test_eq(check_weight(m, is_bn)[0].sum(), 22)