code stringlengths 3 6.57k |
|---|
print(f'Loading "{pkl}"... ', end='', flush=True) |
dnnlib.util.open_url(pkl, verbose=False) |
legacy.load_network_pkl(f) |
print('Done.') |
CapturedException() |
print('Failed!') |
self._ignore_timing() |
isinstance(data, CapturedException) |
tuple(sorted(tweak_kwargs.items() |
self._networks.get(cache_key, None) |
copy.deepcopy(orig_net) |
self._tweak_network(net, **tweak_kwargs) |
net.to(self._device) |
CapturedException() |
self._ignore_timing() |
isinstance(net, CapturedException) |
_tweak_network(self, net) |
misc.named_params_and_buffers(net) |
name.endswith('.magnitude_ema') |
value.rsqrt() |
numpy() |
print(f'{name:<50s}{np.min(value) |
np.max(value) |
name.endswith('.weight') |
value.square() |
mean([1,2,3]) |
sqrt() |
numpy() |
print(f'{name:<50s}{np.min(value) |
np.max(value) |
_get_pinned_buf(self, ref) |
tuple(ref.shape) |
self._pinned_bufs.get(key, None) |
torch.empty(ref.shape, dtype=ref.dtype) |
pin_memory() |
to_device(self, buf) |
self._get_pinned_buf(buf) |
copy_(buf) |
to(self._device) |
to_cpu(self, buf) |
self._get_pinned_buf(buf) |
copy_(buf) |
clone() |
_ignore_timing(self) |
_apply_cmap(self, x, name='viridis') |
self._cmaps.get(name, None) |
matplotlib.cm.get_cmap(name) |
cmap(np.linspace(0, 1, num=1024) |
self.to_device(torch.from_numpy(cmap) |
clamp(0, hi) |
to(torch.int64) |
torch.nn.functional.embedding(x, cmap) |
self.get_network(pkl, 'G_ema') |
any('noise_const' in name for name, _buf in G.synthesis.named_buffers() |
hasattr(G.synthesis, 'input') |
hasattr(G.synthesis.input, 'transform') |
np.eye(3) |
np.linalg.inv(np.asarray(input_transform) |
CapturedException() |
G.synthesis.input.transform.copy_(torch.from_numpy(m) |
list(set(all_seeds) |
np.zeros([len(all_seeds) |
np.zeros([len(all_seeds) |
enumerate(all_seeds) |
np.random.RandomState(seed) |
rnd.randn(G.z_dim) |
rnd.randint(G.c_dim) |
self.to_device(torch.from_numpy(all_zs) |
self.to_device(torch.from_numpy(all_cs) |
G.mapping(z=all_zs, c=all_cs, truncation_psi=trunc_psi, truncation_cutoff=trunc_cutoff) |
dict(zip(all_seeds, all_ws) |
torch.stack([all_ws[seed] * weight for seed, weight in w0_seeds]) |
sum(dim=0, keepdim=True) |
len(stylemix_idx) |
dnnlib.EasyDict(noise_mode=noise_mode, force_fp32=force_fp32) |
torch.manual_seed(random_seed) |
self.run_synthesis_net(G.synthesis, w, capture_layer=layer_name, **synthesis_kwargs) |
tuple(sorted(synthesis_kwargs.items() |
torch.manual_seed(random_seed) |
self.run_synthesis_net(G.synthesis, w, **synthesis_kwargs) |
_apply_affine_transformation(out.to(torch.float32) |
to(torch.float32) |
max(min(base_channel, out.shape[0] - sel_channels) |
out.mean() |
sel.mean() |
out.std() |
sel.std() |
out.norm(float('inf') |
sel.norm(float('inf') |
img.norm(float('inf') |
clip(1e-8, 1e8) |
clamp(0, 255) |
to(torch.uint8) |
permute(1, 2, 0) |
sig.to(torch.float32) |
sig.mean(dim=[1,2], keepdim=True) |
torch.kaiser_window(sig.shape[1], periodic=False, beta=fft_beta, device=self._device) |
torch.kaiser_window(sig.shape[2], periodic=False, beta=fft_beta, device=self._device) |
torch.fft.fftn(sig, dim=[1,2]) |
abs() |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.