MCPcopy Create free account

hub / github.com/darioShar/DLPM / functions

Functions785 in github.com/darioShar/DLPM

⨍ Functions 785◇ Types & classes 166
↓ 27 callersMethodlog
(self, data_type, data)
dlpm/NeptuneLogger.py:56
↓ 24 callersFunctionnormalization
Make a standard normalization layer. :param channels: number of input channels. :return: an nn.Module for normalization.
dlpm/models/nn.py:93
↓ 22 callersMethodload
(self, p = None, checkpoint_dir = None, logger = None, ep
bem/Experiments.py:152
↓ 21 callersFunctionmatch_last_dims
(data, size)
dlpm/methods/LIM/functions/sampler.py:8
↓ 20 callersFunctionncsn_conv3x3
3x3 convolution with PyTorch initialization. Same as NCSNv1/NCSNv2.
dlpm/methods/LIM/models/layers.py:108
↓ 20 callersFunctionncsn_conv3x3
3x3 convolution with PyTorch initialization. Same as NCSNv1/NCSNv2.
dlpm/models/layers.py:108
↓ 18 callersMethod__init__
(self, in_dim, num_units, init_scale=0.1)
dlpm/methods/LIM/models/layers.py:549
↓ 18 callersMethod__init__
(self, in_dim, num_units, init_scale=0.1)
dlpm/models/layers.py:549
↓ 18 callersMethodgenerate
(self, *args, **kwargs)
bem/datasets/Data.py:75
↓ 17 callersMethodlog
(self, data_type, data)
bem/Logger.py:21
↓ 14 callersMethodsave
(self, curr_epoch = None, files='all', save_new_eval=False)
bem/Experiments.py:173
↓ 14 callersMethodupdate_constants
(self, shape)
dlpm/methods/dlpm.py:171
↓ 13 callersMethodget_t_to_batch_size
(self, x_t, t)
dlpm/methods/dlpm.py:165
↓ 12 callersMethodupdate
(self, module)
bem/utils_ema.py:22
↓ 11 callersMethodget
if there is no matching key, return one of two closest values if linear_approx=True, return linear approximation of two closest values.
dlpm/methods/LIM/torchlevy/torch_dictionary.py:22
↓ 11 callersMethodget
if there is no matching key, return one of two closest values if linear_approx=True, return linear approximation of two closest values.
bem/datasets/torchlevy/torch_dictionary.py:22
↓ 10 callersFunctionconv_nd
Create a 1D, 2D, or 3D convolution module.
dlpm/models/nn.py:22
↓ 10 callersMethodmarginal_std
(self, t)
dlpm/methods/LIM/functions/sde.py:46
↓ 9 callersMethodbeta
(self, t)
dlpm/methods/LIM/functions/sde.py:27
↓ 9 callersMethodload_state_dict
(self, state_dict)
dlpm/methods/LIM/models/ema.py:48
↓ 8 callersMethoddiffusion_coeff
(self, t)
dlpm/methods/LIM/functions/sde.py:43
↓ 8 callersMethodmarginal_log_mean_coeff
(self, t)
dlpm/methods/LIM/functions/sde.py:35
↓ 8 callersFunctionupfirdn2d
(input, kernel, up=1, down=1, pad=(0, 0))
dlpm/methods/LIM/op/upfirdn2d.py:145
↓ 7 callersMethod__init__
(self, in_dim, num_units, init_scale=0.1)
dlpm/methods/LIM/models/layerspp.py:61
↓ 7 callersMethod__init__
(self, num_features, bias=True)
dlpm/methods/LIM/models/normalization.py:150
↓ 7 callersMethod__init__
(self, in_dim, num_units, init_scale=0.1)
dlpm/models/layerspp.py:61
↓ 7 callersMethod__init__
(self, num_features, bias=True)
dlpm/models/normalization.py:150
↓ 7 callersFunctiondefault_init
The same initialization used in DDPM.
dlpm/methods/LIM/models/layerspp.py:71
↓ 7 callersFunctiondefault_init
The same initialization used in DDPM.
dlpm/models/layerspp.py:71
↓ 7 callersMethodsample
(self)
dlpm/methods/LIM/diffusion.py:302
↓ 6 callersFunctionmatch_last_dims
(data, size)
dlpm/methods/dlpm.py:47
↓ 6 callersMethodsample
(self, alpha, beta=0, size=1, loc=0, scale=1, type=torch.float32, reject_threshold:int=None, is
bem/datasets/torchlevy/levy.py:176
↓ 6 callersMethodscore
(self, x: torch.Tensor)
dlpm/methods/LIM/torchlevy/levy_gaussian.py:24
↓ 6 callersMethodscore
(self, x: torch.Tensor)
bem/datasets/torchlevy/levy_gaussian.py:24
↓ 6 callersMethodsetParams
(self, *args, **kwargs)
bem/datasets/Data.py:60
↓ 6 callersMethodstate_dict
(self)
dlpm/methods/LIM/models/ema.py:45
↓ 5 callersMethod__init__
(self, channels, use_conv, dims=2)
dlpm/models/unet.py:58
↓ 5 callersFunction_shape
(x, dim)
dlpm/methods/LIM/models/up_or_down_sampling.py:191
↓ 5 callersFunction_shape
(x, dim)
dlpm/models/up_or_down_sampling.py:191
↓ 5 callersFunctionddpm_conv3x3
3x3 convolution with DDPM initialization.
dlpm/methods/LIM/models/layers.py:118
↓ 5 callersFunctionddpm_conv3x3
3x3 convolution with DDPM initialization.
dlpm/models/layers.py:118
↓ 5 callersFunctionis_image_dataset
(name)
bem/datasets/__init__.py:102
↓ 5 callersFunctionnonlinearity
(x)
dlpm/methods/LIM/models/ddpm.py:34
↓ 5 callersFunctionnonlinearity
(x)
dlpm/models/ddpm.py:34
↓ 5 callersMethodpredict_eps
(self, x_t, t, xstart)
dlpm/methods/dlpm.py:198
↓ 5 callersFunctionscore_model
(x, t)
dlpm/methods/LIM/functions/sampler.py:54
↓ 4 callersFunctionLIM_sampler
(#args, #config, ddim, x, y,
dlpm/methods/LIM/functions/sampler.py:15
↓ 4 callersFunctionNormalize
(in_channels)
dlpm/methods/LIM/models/ddpm.py:38
↓ 4 callersFunctionNormalize
(in_channels)
dlpm/models/ddpm.py:38
↓ 4 callersMethod__init__
Build pretrained InceptionV3 Parameters ---------- output_blocks : list of int Indices of blocks to return featur
dlpm/methods/LIM/evaluate/inception.py:31
↓ 4 callersMethod__init__
(self, config)
dlpm/methods/LIM/models/ddpm.py:200
↓ 4 callersMethod__init__
(self, p)
dlpm/models/ddpm.py:200
↓ 4 callersMethod__init__
Build pretrained InceptionV3 Parameters ---------- output_blocks : list of int Indices of blocks to return featur
bem/evaluate/inception.py:31
↓ 4 callersFunction_between_minus_1_1_with_quantile
(x, quantile, scale_to_minus_1_1 = True)
bem/datasets/Distributions.py:93
↓ 4 callersFunction_setup_kernel
(k)
dlpm/methods/LIM/models/up_or_down_sampling.py:181
↓ 4 callersFunction_setup_kernel
(k)
dlpm/models/up_or_down_sampling.py:181
↓ 4 callersFunctiondefault_init
The same initialization used in DDPM.
dlpm/methods/LIM/models/layers.py:88
↓ 4 callersFunctiondefault_init
The same initialization used in DDPM.
dlpm/models/layers.py:88
↓ 4 callersMethodevaluate
(self, evaluate_emas = False, **kwargs)
bem/TrainingManager.py:160
↓ 4 callersFunctionget_dataset
(args, config)
dlpm/methods/LIM/datasets/__init__.py:44
↓ 4 callersFunctioninverse_affine_transform
(x)
bem/datasets/__init__.py:108
↓ 4 callersMethodload_original_data
(self, nsamples)
bem/GenerationManager.py:65
↓ 4 callersMethodload_state_dict
(self, state_dict)
bem/utils_ema.py:62
↓ 4 callersMethodpdf
calculate pdf through zolotarev thm ref. page 7, https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID2894444_code545.pdf?abstrac
dlpm/methods/LIM/torchlevy/levy.py:13
↓ 4 callersMethodpdf
calculate pdf through zolotarev thm ref. page 7, https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID2894444_code545.pdf?abstrac
bem/datasets/torchlevy/levy.py:13
↓ 4 callersFunctiontqdm
(x)
dlpm/methods/LIM/evaluate/fid_score.py:39
↓ 4 callersFunctiontqdm
(x)
bem/evaluate/fid_score.py:39
↓ 3 callersMethod_get_scatter_marker_specific_kwargs
(self, marker)
bem/GenerationManager.py:134
↓ 3 callersFunction_get_score_dict_cft
(alpha, sigma_1, sigma_2, beta, t0, Fs, mode='fraction')
dlpm/methods/LIM/torchlevy/levy_gaussian_origin.py:51
↓ 3 callersFunction_get_score_dict_cft
(alpha, sigma_1, sigma_2, beta, t0, Fs, is_fdsm)
dlpm/methods/LIM/torchlevy/levy_gaussian.py:48
↓ 3 callersFunction_get_score_dict_cft
(alpha, sigma_1, sigma_2, beta, t0, Fs, mode='fraction')
bem/datasets/torchlevy/levy_gaussian_origin.py:51
↓ 3 callersFunction_get_score_dict_cft
(alpha, sigma_1, sigma_2, beta, t0, Fs, is_fdsm)
bem/datasets/torchlevy/levy_gaussian.py:48
↓ 3 callersMethod_reset_attributes
(self, p, checkpoint_dir, logger)
bem/Experiments.py:93
↓ 3 callersMethodcompute_Gamma_t
(self, t, Sigma_t_1, Sigma_t)
dlpm/methods/dlpm.py:250
↓ 3 callersFunctionfrom_flattened_numpy
Form a torch tensor with the given `shape` from a flattened numpy array `x`.
dlpm/methods/LIM/evaluate/likelihood.py:29
↓ 3 callersFunctionfrom_flattened_numpy
Form a torch tensor with the given `shape` from a flattened numpy array `x`.
bem/evaluate/likelihood.py:29
↓ 3 callersMethodget_exp_hash
(self, p, verbose=False)
bem/utils_exp.py:52
↓ 3 callersMethodget_param_from_config
(config_path, config_file)
bem/utils_exp.py:256
↓ 3 callersFunctionget_timestep_embedding
This matches the implementation in Denoising Diffusion Probabilistic Models: From Fairseq. Build sinusoidal embeddings. This matches
dlpm/methods/LIM/models/ddpm.py:5
↓ 3 callersFunctionget_timestep_embedding
This matches the implementation in Denoising Diffusion Probabilistic Models: From Fairseq. Build sinusoidal embeddings. This matches
dlpm/models/ddpm.py:5
↓ 3 callersMethodget_timesteps
(self, steps)
dlpm/methods/dlpm.py:103
↓ 3 callersFunctionlinear
Create a linear module.
dlpm/models/nn.py:35
↓ 3 callersFunctionloss_fn
(model, sde, x0, t, e, config, y=None
dlpm/methods/LIM/functions/loss.py:12
↓ 3 callersMethodpredict_xstart
(self, x_t, t, eps)
dlpm/methods/dlpm.py:191
↓ 3 callersMethodsave_eval_metrics
(self, eval_path)
bem/TrainingManager.py:287
↓ 3 callersMethodstate_dict
(self)
bem/utils_ema.py:59
↓ 3 callersMethodterminate
(self)
bem/Experiments.py:125
↓ 3 callersFunctionto_flattened_numpy
Flatten a torch tensor `x` and convert it to numpy.
dlpm/methods/LIM/evaluate/likelihood.py:25
↓ 3 callersFunctionto_flattened_numpy
Flatten a torch tensor `x` and convert it to numpy.
bem/evaluate/likelihood.py:25
↓ 3 callersMethodupdate
(self, module)
dlpm/methods/LIM/models/ema.py:16
↓ 3 callersFunctionzero_module
Zero out the parameters of a module and return it.
dlpm/models/nn.py:68
↓ 2 callersMethod__init__
(self, nfeatures: int, diffusion_steps: int, nblocks: int = 2, nunits: int = 64)
dlpm/models/Deprecated.py:35
↓ 2 callersMethod__init__
(self, nunits, dropout_rate, skip_connection,
dlpm/models/DiffusionBlocks.py:14
↓ 2 callersMethod__repr__
(self)
dlpm/methods/LIM/datasets/vision.py:34
↓ 2 callersMethod__repr__
(self)
bem/datasets/vision.py:34
↓ 2 callersFunction_add_channels
(img, total_channels=3)
bem/datasets/tinyimagenet.py:48
↓ 2 callersMethod_format_transform_repr
(self, transform, head)
dlpm/methods/LIM/datasets/vision.py:49
↓ 2 callersMethod_format_transform_repr
(self, transform, head)
dlpm/methods/LIM/datasets/vision.py:70
↓ 2 callersMethod_format_transform_repr
(self, transform, head)
bem/datasets/vision.py:49
↓ 2 callersMethod_format_transform_repr
(self, transform, head)
bem/datasets/vision.py:70
next →1–100 of 785, ranked by callers