Functions — github.com/Sin3DM/Sin3DM

hub / github.com/Sin3DM/Sin3DM / functions

Functions339 in github.com/Sin3DM/Sin3DM

⨍ Functions 339◇ Types & classes 50

↓ 1 callersMethod_load_ema_parameters

(self, rate)

src/diffusion/train_util.py:116

↓ 1 callersMethod_load_lpips_weights

(self)

evaluation/lpips.py:64

↓ 1 callersMethod_load_optimizer_state

(self)

src/diffusion/train_util.py:132

↓ 1 callersMethod_optimize_fp16

(self, opt: th.optim.Optimizer)

src/diffusion/fp16_util.py:189

↓ 1 callersMethod_optimize_normal

(self, opt: th.optim.Optimizer)

src/diffusion/fp16_util.py:210

↓ 1 callersMethod_predict_xstart_from_xprev

(self, x_t, t, xprev)

src/diffusion/gaussian_diffusion.py:336

↓ 1 callersMethod_prior_bpd

Get the prior KL term for the variational lower-bound, measured in bits-per-dim. This term can't be optimized, as it only de

src/diffusion/gaussian_diffusion.py:858

↓ 1 callersMethod_sample_and_visualize

(self)

src/diffusion/train_util.py:174

↓ 1 callersMethod_sample_batch

(self, batch_size)

src/encoding/model.py:114

↓ 1 callersMethod_set_optimizer

set optimizer and lr scheduler used in training

src/encoding/model.py:129

↓ 1 callersMethod_update_ema

(self)

src/diffusion/train_util.py:237

↓ 1 callersMethod_visualize_batch

(self, step)

src/encoding/model.py:283

↓ 1 callersMethod_warmed_up

(self)

src/diffusion/resample.py:153

↓ 1 callersFunctionadd_diffusion_training_options

(parser)

src/utils/parser_util.py:39

↓ 1 callersFunctionadd_encoding_training_options

(parser)

src/utils/parser_util.py:13

↓ 1 callersFunctionadd_sampling_options

(parser)

src/utils/parser_util.py:56

↓ 1 callersFunctionbetas_for_alpha_bar

Create a beta schedule that discretizes the given alpha_t_bar function, which defines the cumulative product of (1-beta) over time from t = [

src/diffusion/gaussian_diffusion.py:46

↓ 1 callersFunctioncalculate_frechet_distance

Numpy implementation of the Frechet Distance. The Frechet distance between two multivariate Gaussians X_1 ~ N(mu_1, C_1) and X_2 ~ N(mu_2, C_2

evaluation/ssfid.py:11

↓ 1 callersFunctioncalculate_lpips_given_images

(group_of_images, lpips_model=None)

evaluation/lpips.py:89

↓ 1 callersFunctioncalculate_multiview_lpips_given_paths

(gen_render_dirs, device="cuda:0")

evaluation/lpips.py:109

↓ 1 callersFunctioncheck_overflow

(value)

src/diffusion/fp16_util.py:236

↓ 1 callersFunctioncheckpoint

Evaluate a function without caching intermediate activations, allowing for reduced memory at the expense of extra compute in the backward pas

src/diffusion/nn.py:124

↓ 1 callersMethodcondition_mean

Compute the mean for the previous step, given a function cond_fn that computes the gradient of a conditional log probability with res

src/diffusion/gaussian_diffusion.py:357

↓ 1 callersMethodcondition_score

Compute what the p_mean_variance output would have been, should the model's score function be conditioned by cond_fn. See co

src/diffusion/gaussian_diffusion.py:372

↓ 1 callersMethodconvert_to_fp16

Convert the torso of the model to float16.

src/diffusion/unet_triplane.py:451

↓ 1 callersFunctioncreate_gaussian_diffusion

( *, steps=1000, learn_sigma=False, sigma_small=False, noise_schedule="linear", use_kl

src/diffusion/script_util.py:22

↓ 1 callersFunctioncreate_named_schedule_sampler

Create a ScheduleSampler from a library of pre-defined samplers. :param name: the name of the sampler. :param diffusion: the diffusion o

src/diffusion/resample.py:8

↓ 1 callersMethodddim_sample

Sample x_{t-1} from the model using DDIM. Same usage as p_sample().

src/diffusion/gaussian_diffusion.py:538

↓ 1 callersMethodddim_sample_loop_progressive

Use DDIM to sample from the model and yield intermediate samples from each timestep of DDIM. Same usage as p_sample_loop_pro

src/diffusion/gaussian_diffusion.py:680

↓ 1 callersMethoddecode

(self, x, feat_maps, aabb=None)

src/encoding/networks.py:192

↓ 1 callersMethoddecode

(self, x, feat_maps, aabb=None)

src/encoding/networks.py:299

↓ 1 callersMethoddecode_voxel

(self, save_dir, triplane_feat, reso, n_faces=10000, only_largest_cc=True)

src/encoding/model.py:476

↓ 1 callersFunctiondiffusion_model_path

(exp_tag, ema, step)

src/utils/parser_util.py:229

↓ 1 callersFunctiondiscretized_gaussian_log_likelihood

Compute the log-likelihood of a Gaussian distribution discretizing to a given image. :param x: the target images. It is assumed that thi

src/diffusion/losses.py:50

↓ 1 callersMethodencode

(self, vol)

src/encoding/networks.py:61

↓ 1 callersMethodencode

(self, vol)

src/encoding/networks.py:164

↓ 1 callersMethodencode

(self, vol)

src/encoding/networks.py:271

↓ 1 callersFunctioneval_Div_given_paths

(data_paths, device="cpu")

evaluation/patch_utils.py:169

↓ 1 callersFunctioneval_LP_Fscore

compute LP-F-score over two set of patches. Args: gen_patches (torch.Tensor): patches from generated shape ref_patches (torch.Ten

evaluation/patch_utils.py:100

↓ 1 callersFunctioneval_LP_IoU

compute LP-IoU over two set of patches. Args: gen_patches (torch.Tensor): patches from generated shape ref_patches (torch.Tensor)

evaluation/patch_utils.py:77

↓ 1 callersFunctioneval_LP_given_paths

(data_paths, ref_path, patch_size=11, stride=5, patch_num=1000, device="cpu")

evaluation/patch_utils.py:125

↓ 1 callersFunctioneval_SSFID_given_paths

(data_paths, ref_path, model_out_layer=2, device="cpu")

evaluation/ssfid.py:81

↓ 1 callersFunctionevaluate_tsdf_prediction

(pred_sdf, gt_sdf, sdf_threshold)

src/encoding/model.py:491

↓ 1 callersFunctionfind_ema_checkpoint

(main_checkpoint, step, rate)

src/diffusion/train_util.py:323

↓ 1 callersFunctionfind_input_path

(example)

src/app.py:91

↓ 1 callersMethodforward

(self, x)

src/diffusion/nn.py:13

↓ 1 callersMethodforward_backward

(self, batch, cond)

src/diffusion/train_util.py:198

↓ 1 callersMethodgeo_parameters

(self)

src/encoding/networks.py:48

↓ 1 callersFunctionget_activations

Calculates the activations of the pool_3 layer for all images. Params: -- files : List of image files paths -- model : Instan

evaluation/sifid.py:45

↓ 1 callersFunctionget_data_iterator

(featmaps_data, sizes, batch_size=1)

src/utils/triplane_util.py:64

↓ 1 callersFunctionget_networks

(cfg)

src/encoding/networks.py:7

↓ 1 callersFunctionget_rank_without_mpi_import

()

src/diffusion/logger.py:403

↓ 1 callersFunctioninterpolate

(attr, rast, attr_idx, rast_db=None)

src/encoding/utils3d.py:224

↓ 1 callersMethodload_ckpt

load saved checkpoint

src/encoding/model.py:159

↓ 1 callersMethodlog_loss_dict

(self, diffusion, ts, losses)

src/diffusion/train_util.py:283

↓ 1 callersMethodlog_step

(self)

src/diffusion/train_util.py:249

↓ 1 callersFunctionlogkv

Log a value of some diagnostic Call this once for each diagnostic quantity, each iteration If called many times, last value will be used.

src/diffusion/logger.py:212

↓ 1 callersFunctionmake_output_format

(format, ev_dir, log_suffix="")

src/diffusion/logger.py:191

↓ 1 callersMethodmake_watertight_copy

(self, resolution=100_000, is_watertight=False)

data/mesh_sampler.py:80

↓ 1 callersMethodmake_watertight_copy

(self, resolution=100_000, is_watertight=False)

data/mesh_sampler_pbr.py:87

↓ 1 callersFunctionmaster_params_to_model_params

Copy the master parameter data back into the model parameters.

src/diffusion/fp16_util.py:65

↓ 1 callersFunctionmaster_params_to_state_dict

( model, param_groups_and_shapes, master_params, use_fp16 )

src/diffusion/fp16_util.py:95

↓ 1 callersMethodmaster_params_to_state_dict

(self, master_params)

src/diffusion/fp16_util.py:227

↓ 1 callersFunctionmesh_decimation

(v, f, face_count=10000)

src/encoding/utils3d.py:216

↓ 1 callersFunctionmodel_grads_to_master_grads

Copy the gradients from the model parameters into the master parameters from make_master_params().

src/diffusion/fp16_util.py:52

↓ 1 callersFunctionmpi_weighted_mean

Copied from: https://github.com/openai/baselines/blob/ea25b9e8b234e6ee1bca43083f8f3cf974143998/baselines/common/mpi_util.py#L110 Perform a we

src/diffusion/logger.py:412

↓ 1 callersMethodnormalize

(self, reso=256, enlarge_scale=1.03, mult=8)

data/mesh_sampler.py:75

↓ 1 callersMethodnormalize

(self, reso=256, enlarge_scale=1.03, mult=8)

data/mesh_sampler_pbr.py:82

↓ 1 callersMethodoptimize

(self, opt: th.optim.Optimizer)

src/diffusion/fp16_util.py:183

↓ 1 callersMethodp_sample

Sample x_{t-1} from the model at the given timestep. :param model: the model to sample from. :param x: the current tensor at

src/diffusion/gaussian_diffusion.py:396

↓ 1 callersMethodp_sample_loop

Generate samples from the model. :param model: the model module. :param shape: the shape of the samples, (N, C, H, W).

src/diffusion/gaussian_diffusion.py:442

↓ 1 callersMethodp_sample_loop_progressive

Generate samples from the model and yield intermediate samples from each timestep of diffusion. Arguments are the same as p_

src/diffusion/gaussian_diffusion.py:488

↓ 1 callersFunctionpairwise_IoU_dist

average pairwise 1-IoU for a list of 3D shape volume

evaluation/patch_utils.py:30

↓ 1 callersFunctionparam_grad_or_zeros

(param)

src/diffusion/fp16_util.py:141

↓ 1 callersFunctionparse_resume_step_from_filename

Parse filenames of the form path/to/modelNNNNNN.pt, where NNNNNN is the checkpoint's number of steps.

src/diffusion/train_util.py:296

↓ 1 callersFunctionprofile_kv

(scopename)

src/diffusion/logger.py:294

↓ 1 callersMethodq_mean_variance

Get the distribution q(x_t | x_0). :param x_start: the [N x C x ...] tensor of noiseless inputs. :param t: the number of dif

src/diffusion/gaussian_diffusion.py:172

↓ 1 callersFunctionreadOBJ

(filePath, location, rotation_euler, scale)

rendering/blender_render_pbr.py:59

↓ 1 callersFunctionread_metarial_params_from_mtl

(path)

src/encoding/utils3d.py:28

↓ 1 callersMethodrun_loop

(self)

src/diffusion/train_util.py:144

↓ 1 callersMethodrun_step

(self, batch, cond)

src/diffusion/train_util.py:163

↓ 1 callersMethodsample

Importance-sample timesteps for a batch. :param batch_size: the number of timesteps. :param device: the torch device to save

src/diffusion/resample.py:42

↓ 1 callersFunctionsample_args

()

src/utils/parser_util.py:148

↓ 1 callersFunctionsample_grid_points_aabb

(aabb, resolution)

src/encoding/utils3d.py:13

↓ 1 callersMethodsample_watertight_surf

(self, n=1_000_000)

data/mesh_sampler.py:99

↓ 1 callersMethodsample_watertight_surf

(self, n=1_000_000)

data/mesh_sampler_pbr.py:106

↓ 1 callersMethodsave_ckpt

save checkpoint for future restore

src/encoding/model.py:141

↓ 1 callersFunctionsave_mesh_with_pbr

(fname, pointnp_px3, tcoords_px2, facenp_fx3, facetex_fx3, albedo_img, metallic_img, r

src/encoding/utils3d.py:137

↓ 1 callersFunctionsave_mesh_with_tex

(fname, pointnp_px3, tcoords_px2, facenp_fx3, facetex_fx3, tex_img, mtl_str=None, Kd=[1

src/encoding/utils3d.py:46

↓ 1 callersFunctionsave_mesh_with_tex_to_glb

(fname, pointnp_px3, tcoords_px2, facenp_fx3, facetex_fx3, tex_img)

src/encoding/utils3d.py:91

↓ 1 callersFunctionsdfgrid_to_mesh

(sdf_grid, save_path=None, only_largest_cc=True, is_voxel=False)

src/encoding/utils3d.py:196

↓ 1 callersFunctionsetLight_sun

(location, rotation_euler, strength, shadow_soft_size = 0.05)

rendering/blender_render.py:101

↓ 1 callersFunctionsetLight_threeArea

( radius = 4, height = 10, intensity = 100, softness = 1)

rendering/blender_render_pbr.py:138

↓ 1 callersMethodset_comm

(self, comm)

src/diffusion/logger.py:385

↓ 1 callersMethodset_level

(self, level)

src/diffusion/logger.py:382

↓ 1 callersFunctionspace_timesteps

Create a list of timesteps to use from an original diffusion process, given the number of timesteps we want to take from equally-sized portio

src/diffusion/respace.py:7

↓ 1 callersFunctionstate_dict_to_master_params

(model, state_dict, use_fp16)

src/diffusion/fp16_util.py:116

↓ 1 callersMethodstate_dict_to_master_params

(self, state_dict)

src/diffusion/fp16_util.py:232

↓ 1 callersMethodtex_parameters

(self)

src/encoding/networks.py:51

↓ 1 callersFunctiontrain_args

()

src/utils/parser_util.py:102

← previous next →101–200 of 339, ranked by callers