def complexity(cx, w_in, w_out, stride, bm=None, gw=None, se_r=None):
err_str = "Vanilla block does not support bm, gw, and se_r options"
assert bm is None and gw is None and se_r is None, err_str
cx = nu.complexity_conv2d(cx, w_in, w_out, 3, stride, 1)
cx = nu.complexity_batchnorm2d(cx, w_out)
cx = nu.complexity_conv2d(cx, w_out, w_out, 3, 1, 1)
cx = nu.complexity_batchnorm2d(cx, w_out)
return cx
def complexity(cx, w_in, w_out, stride, w_b=None, num_gs=1):
err_str = "Basic transform does not support w_b and num_gs options"
assert w_b is None and num_gs == 1, err_str
cx = nu.complexity_conv2d(cx, w_in, w_out, 3, stride, 1)
cx = nu.complexity_batchnorm2d(cx, w_out)
cx = nu.complexity_conv2d(cx, w_out, w_out, 3, 1, 1)
cx = nu.complexity_batchnorm2d(cx, w_out)
return cx
def complexity(cx, w_in, w_out, stride, w_b, num_gs):
(s1, s3) = (stride, 1) if cfg.RESNET.STRIDE_1X1 else (1, stride)
cx = nu.complexity_conv2d(cx, w_in, w_b, 1, s1, 0)
cx = nu.complexity_batchnorm2d(cx, w_b)
cx = nu.complexity_conv2d(cx, w_b, w_b, 3, s3, 1, num_gs)
cx = nu.complexity_batchnorm2d(cx, w_b)
cx = nu.complexity_conv2d(cx, w_b, w_out, 1, 1, 0)
cx = nu.complexity_batchnorm2d(cx, w_out)
return cx
def complexity(cx, w_in, exp_r, kernel, stride, se_r, w_out):
w_exp = int(w_in * exp_r)
if w_exp != w_in:
cx = nu.complexity_conv2d(cx, w_in, w_exp, 1, 1, 0)
cx = nu.complexity_batchnorm2d(cx, w_exp)
padding = (kernel - 1) // 2
cx = nu.complexity_conv2d(cx, w_exp, w_exp, kernel, stride, padding,
w_exp)
cx = nu.complexity_batchnorm2d(cx, w_exp)
cx = SE.complexity(cx, w_exp, int(w_in * se_r))
cx = nu.complexity_conv2d(cx, w_exp, w_out, 1, 1, 0)
cx = nu.complexity_batchnorm2d(cx, w_out)
return cx
def complexity(cx, w_in, w_out, stride, bm, gw, se_r):
w_b = int(round(w_out * bm))
g = w_b // gw
cx = nu.complexity_conv2d(cx, w_in, w_b, 1, 1, 0)
cx = nu.complexity_batchnorm2d(cx, w_b)
cx = nu.complexity_conv2d(cx, w_b, w_b, 3, stride, 1, g)
cx = nu.complexity_batchnorm2d(cx, w_b)
if se_r:
w_se = int(round(w_in * se_r))
cx = SE.complexity(cx, w_b, w_se)
cx = nu.complexity_conv2d(cx, w_b, w_out, 1, 1, 0)
cx = nu.complexity_batchnorm2d(cx, w_out)
return cx
def complexity(cx, w_in, w_out, stride, trans_fun, w_b, num_gs):
proj_block = (w_in != w_out) or (stride != 1)
if proj_block:
h, w = cx["h"], cx["w"]
cx = nu.complexity_conv2d(cx, w_in, w_out, 1, stride, 0)
cx = nu.complexity_batchnorm2d(cx, w_out)
cx["h"], cx["w"] = h, w # parallel branch
cx = trans_fun.complexity(cx, w_in, w_out, stride, w_b, num_gs)
return cx
def complexity(cx, w_in, w_out, stride, bm=1.0, gw=1, se_r=None):
proj_block = (w_in != w_out) or (stride != 1)
if proj_block:
h, w = cx["h"], cx["w"]
cx = nu.complexity_conv2d(cx, w_in, w_out, 1, stride, 0)
cx = nu.complexity_batchnorm2d(cx, w_out)
cx["h"], cx["w"] = h, w # parallel branch
cx = BottleneckTransform.complexity(cx, w_in, w_out, stride, bm, gw,
se_r)
return cx
def complexity(cx, w_in, w_out, stride, bm=None, gw=None, se_r=None):
err_str = "Basic transform does not support bm, gw, and se_r options"
assert bm is None and gw is None and se_r is None, err_str
proj_block = (w_in != w_out) or (stride != 1)
if proj_block:
h, w = cx["h"], cx["w"]
cx = nu.complexity_conv2d(cx, w_in, w_out, 1, stride, 0)
cx = nu.complexity_batchnorm2d(cx, w_out)
cx["h"], cx["w"] = h, w # parallel branch
cx = BasicTransform.complexity(cx, w_in, w_out, stride)
return cx
def complexity(cx, w_in, w_out):
cx = nu.complexity_conv2d(cx, w_in, w_out, 7, 2, 3)
cx = nu.complexity_batchnorm2d(cx, w_out)
cx = nu.complexity_maxpool2d(cx, 3, 2, 1)
return cx
def complexity(cx, w_in, w_out):
cx = nu.complexity_conv2d(cx, w_in, w_out, 3, 1, 1)
cx = nu.complexity_batchnorm2d(cx, w_out)
return cx
def complexity(cx, w_in, w_out, nc):
cx = nu.complexity_conv2d(cx, w_in, w_out, 1, 1, 0)
cx = nu.complexity_batchnorm2d(cx, w_out)
cx["h"], cx["w"] = 1, 1
cx = nu.complexity_conv2d(cx, w_out, nc, 1, 1, 0, bias=True)
return cx