MCPcopy Create free account

hub / github.com/VAST-AI-Research/TripoSG / functions

Functions146 in github.com/VAST-AI-Research/TripoSG

⨍ Functions 146◇ Types & classes 26
↓ 25 callersFunction_upsample_like
(src,tar)
scripts/briarmbg.py:29
↓ 7 callersMethod__init__
(self, in_ch=3, mid_ch=12, out_ch=3, img_size=512)
scripts/briarmbg.py:39
↓ 6 callersMethoddecode
( self, z: torch.Tensor, sampled_points: torch.Tensor, return_dict: bool = Tru
triposg/models/autoencoders/autoencoder_kl_triposg.py:512
↓ 6 callersFunctionget_neighbor
(t, shift, axis)
triposg/inference_utils.py:220
↓ 6 callersFunctionsafe_where
(neighbor)
triposg/inference_utils.py:271
↓ 6 callersMethodset_timesteps
Sets the discrete timesteps used for the diffusion chain (to be run before inference). Args: num_inference_steps (`int`)
triposg/schedulers/scheduling_rectified_flow.py:166
↓ 4 callersMethodset_attn_processor
r""" Sets the attention processor to use to compute attention. Parameters: processor (`dict` of `AttentionProcessor` or o
triposg/models/transformers/triposg_transformer.py:580
↓ 3 callersFunctionhierarchical_extract_geometry
Args: geometric_func: device: bounds: dense_octree_depth: hierarchical_octree_depth: Returns:
triposg/inference_utils.py:129
↓ 3 callersMethodset_attn_processor
r""" Sets the attention processor to use to compute attention. Parameters: processor (`dict` of `AttentionProcessor` or o
triposg/models/autoencoders/autoencoder_kl_triposg.py:345
↓ 3 callersMethodset_topk
(self, topk)
triposg/models/transformers/triposg_transformer.py:275
↓ 2 callersMethod__init__
( self, in_channels: int = 3, dim: int = 512, num_attention_heads: int = 8,
triposg/models/autoencoders/autoencoder_kl_triposg.py:27
↓ 2 callersMethod_decode
( self, z: torch.Tensor, sampled_points: torch.Tensor, num_chunks: int = 50000
triposg/models/autoencoders/autoencoder_kl_triposg.py:481
↓ 2 callersMethod_encode
( self, x: torch.Tensor, num_tokens: int = 2048, seed: Optional[int] = None )
triposg/models/autoencoders/autoencoder_kl_triposg.py:439
↓ 2 callersFunctionflash_extract_geometry
( latents: torch.FloatTensor, vae: Callable, bounds: Union[Tuple[float], List[float], float] = 1.0
triposg/inference_utils.py:319
↓ 2 callersMethodsample
(self, generator: Optional[torch.Generator] = None)
triposg/models/autoencoders/vae.py:27
↓ 2 callersMethodset_chunk_feed_forward
(self, chunk_size: Optional[int], dim: int = 0)
triposg/models/transformers/triposg_transformer.py:283
↓ 2 callersMethodset_flash_decoder
(self)
triposg/models/autoencoders/autoencoder_kl_triposg.py:270
↓ 2 callersMethodstep
Predict the sample from the previous timestep by reversing the SDE. This function propagates the diffusion process from the learned m
triposg/schedulers/scheduling_rectified_flow.py:234
↓ 1 callersMethod__init__
( self, dim: int, num_attention_heads: int, use_self_attention: bool = True,
triposg/models/transformers/triposg_transformer.py:161
↓ 1 callersMethod_init_step_index
(self, timestep)
triposg/schedulers/scheduling_rectified_flow.py:226
↓ 1 callersMethod_sample_features
Sample points from features of the input point cloud. Args: x (torch.Tensor): The input point cloud. shape: (B, N, C)
triposg/models/autoencoders/autoencoder_kl_triposg.py:402
↓ 1 callersMethod_set_time_proj
( self, time_embedding_type: str, inner_dim: int, flip_sin_to_cos: bool,
triposg/models/transformers/triposg_transformer.py:475
↓ 1 callersMethod_t_to_sigma
(self, timestep)
triposg/schedulers/scheduling_rectified_flow.py:157
↓ 1 callersMethoddisable_slicing
r""" Disable sliced VAE decoding. If `enable_slicing` was previously enabled, this method will go back to computing decoding in one st
triposg/models/autoencoders/autoencoder_kl_triposg.py:395
↓ 1 callersMethodenable_slicing
r""" Enable sliced VAE decoding. When this option is enabled, the VAE will split the input tensor in slices to compute decoding in sev
triposg/models/autoencoders/autoencoder_kl_triposg.py:387
↓ 1 callersMethodencode
Encode a batch of point features into latents.
triposg/models/autoencoders/autoencoder_kl_triposg.py:460
↓ 1 callersMethodencode_image
(self, image, device, num_shapes_per_prompt)
triposg/pipelines/pipeline_triposg_scribble.py:149
↓ 1 callersMethodencode_image
(self, image, device, num_images_per_prompt)
triposg/pipelines/pipeline_triposg.py:138
↓ 1 callersMethodencode_text
(self, prompt, device, num_shapes_per_prompt)
triposg/pipelines/pipeline_triposg_scribble.py:140
↓ 1 callersFunctionexpand_edge_region_fast
(edge_coords, grid_size)
triposg/inference_utils.py:100
↓ 1 callersFunctionextract_near_surface_volume_fn
Args: input_tensor: shape [D, D, D], torch.float16 alpha: isosurface offset Returns: mask: shape [D, D, D], torch.int
triposg/inference_utils.py:203
↓ 1 callersFunctionfind_bounding_box
(gray_image)
scripts/image_process.py:13
↓ 1 callersFunctionfind_candidates_band
Returns the coordinates of all voxels in the occupancy_grid where |value| < band_threshold. Args: occupancy_grid (torch.Tensor): A 3
triposg/inference_utils.py:72
↓ 1 callersFunctiongenerate_dense_grid_points_2
( bbox_min: np.ndarray, bbox_max: np.ndarray, octree_resolution: int, indexing: str = "ij", )
triposg/inference_utils.py:300
↓ 1 callersFunctiongenerate_dense_grid_points_gpu
(bbox_min: torch.Tensor, bbox_max: torch.Tensor,
triposg/inference_utils.py:12
↓ 1 callersMethodget_dims
(self, input_dim)
triposg/models/embeddings.py:70
↓ 1 callersMethodindex_for_timestep
(self, timestep, schedule_timesteps=None)
triposg/schedulers/scheduling_rectified_flow.py:212
↓ 1 callersFunctionis_valid_alpha
(alpha, min_ratio = 0.01)
scripts/image_process.py:25
↓ 1 callersFunctionload_image
(img_path, bg_color=None, rmbg_net=None, padding_ratio=0.1)
scripts/image_process.py:20
↓ 1 callersFunctionload_surface
(data_path, num_pc=204800)
scripts/inference_vae.py:12
↓ 1 callersFunctionmesh_to_pymesh
(vertices, faces)
scripts/inference_triposg.py:48
↓ 1 callersFunctionparallel_zoom
(occupancy_grid, scale_factor)
triposg/inference_utils.py:123
↓ 1 callersFunctionprepare_image
(image_path, bg_color, rmbg_net=None)
scripts/image_process.py:143
↓ 1 callersMethodprepare_latents
( self, batch_size, num_tokens, num_channels_latents, dtype, d
triposg/pipelines/pipeline_triposg_scribble.py:162
↓ 1 callersMethodprepare_latents
( self, batch_size, num_tokens, num_channels_latents, dtype, d
triposg/pipelines/pipeline_triposg.py:151
↓ 1 callersFunctionpymesh_to_trimesh
(mesh)
scripts/inference_triposg.py:54
↓ 1 callersMethodqkv
(self, attn, q, k, v, attn_mask, dropout_p, is_causal)
triposg/models/attention_processor.py:27
↓ 1 callersMethodquery_geometry
( self, model_fn: callable, queries: torch.Tensor, sample: torch.Tensor,
triposg/models/autoencoders/autoencoder_kl_triposg.py:152
↓ 1 callersFunctionretrieve_timesteps
Calls the scheduler's `set_timesteps` method and retrieves timesteps from the scheduler after the call. Handles custom timesteps. Any kwargs
triposg/pipelines/pipeline_triposg_scribble.py:32
↓ 1 callersFunctionretrieve_timesteps
Calls the scheduler's `set_timesteps` method and retrieves timesteps from the scheduler after the call. Handles custom timesteps. Any kwargs
triposg/pipelines/pipeline_triposg.py:30
↓ 1 callersFunctionrmbg
(image: torch.Tensor)
scripts/image_process.py:34
↓ 1 callersFunctionrun_triposg
( pipe: Any, image_input: Union[str, Image.Image], rmbg_net: Any, seed: int, num_inference
scripts/inference_triposg.py:23
↓ 1 callersFunctionrun_triposg_scribble
( pipe: Any, image_input: Union[str, Image.Image], prompt: str, seed: int, num_inference_s
scripts/inference_triposg_scribble.py:20
↓ 1 callersMethodset_flash_processor
(self, processor)
triposg/models/autoencoders/autoencoder_kl_triposg.py:149
↓ 1 callersFunctionsimplify_mesh
(mesh: trimesh.Trimesh, n_faces)
scripts/inference_triposg.py:59
↓ 1 callersMethodtime_shift
(self, t: torch.Tensor)
triposg/schedulers/scheduling_rectified_flow.py:163
↓ 1 callersMethodtime_shift_dynamic
(self, mu: float, sigma: float, t: torch.Tensor)
triposg/schedulers/scheduling_rectified_flow.py:160
Method__call__
( self, image: PipelineImageInput, prompt: str, num_tokens: int = 512,
triposg/pipelines/pipeline_triposg_scribble.py:188
Method__call__
( self, image: PipelineImageInput, num_inference_steps: int = 50, num_tokens:
triposg/pipelines/pipeline_triposg.py:178
Method__call__
( self, attn: Attention, hidden_states: torch.Tensor, encoder_hidden_states: O
triposg/models/attention_processor.py:67
Method__call__
( self, attn: Attention, hidden_states: torch.Tensor, encoder_hidden_states: O
triposg/models/attention_processor.py:190
Method__call__
( self, attn: Attention, hidden_states: torch.Tensor, encoder_hidden_states: O
triposg/models/attention_processor.py:317
Method__init__
( self, vae: TripoSGVAEModel, transformer: TripoSGDiTModel, scheduler: FlowMat
triposg/pipelines/pipeline_triposg_scribble.py:102
Method__init__
( self, vae: TripoSGVAEModel, transformer: TripoSGDiTModel, scheduler: Flow
triposg/pipelines/pipeline_triposg.py:100
Method__init__
( self, num_train_timesteps: int = 1000, shift: float = 1.0, use_dynamic_shift
triposg/schedulers/scheduling_rectified_flow.py:100
Method__init__
(self, topk=True)
triposg/models/attention_processor.py:20
Method__init__
(self)
triposg/models/attention_processor.py:184
Method__init__
(self)
triposg/models/attention_processor.py:311
Method__init__
The initialization
triposg/models/embeddings.py:42
Method__init__
( self, num_attention_heads: int = 16, width: int = 2048, in_channels: int = 6
triposg/models/transformers/triposg_transformer.py:412
Method__init__
( self, parameters: torch.Tensor, deterministic: bool = False, feature_dim: in
triposg/models/autoencoders/vae.py:9
Method__init__
( self, in_channels: int = 3, out_channels: int = 1, dim: int = 512, n
triposg/models/autoencoders/autoencoder_kl_triposg.py:91
Method__init__
( self, in_channels: int = 3, # NOTE xyz instead of feature dim latent_channels: int
triposg/models/autoencoders/autoencoder_kl_triposg.py:221
Method__init__
(self,in_ch=3,out_ch=3,dirate=1,stride=1)
scripts/briarmbg.py:14
Method__init__
(self, in_ch=3, mid_ch=12, out_ch=3)
scripts/briarmbg.py:122
Method__init__
(self, in_ch=3, mid_ch=12, out_ch=3)
scripts/briarmbg.py:191
Method__init__
(self, in_ch=3, mid_ch=12, out_ch=3)
scripts/briarmbg.py:249
Method__init__
(self, in_ch=3, mid_ch=12, out_ch=3)
scripts/briarmbg.py:297
Method__init__
(self, in_ch=3, out_ch=1, kernel_size=3,
scripts/briarmbg.py:332
Method__init__
(self,config:dict={"in_ch":3,"out_ch":1})
scripts/briarmbg.py:357
Method__len__
(self)
triposg/schedulers/scheduling_rectified_flow.py:326
Method_set_gradient_checkpointing
(self, module, value=False)
triposg/models/transformers/triposg_transformer.py:472
Method_sigma_to_t
(self, sigma)
triposg/schedulers/scheduling_rectified_flow.py:154
Methodattention_kwargs
(self)
triposg/pipelines/pipeline_triposg_scribble.py:137
Methodattention_kwargs
(self)
triposg/pipelines/pipeline_triposg.py:131
Methodattn_processors
r""" Returns: `dict` of attention processors: A dictionary containing all attention processors used in the model with
triposg/models/transformers/triposg_transformer.py:552
Methodattn_processors
r""" Returns: `dict` of attention processors: A dictionary containing all attention processors used in the model with
triposg/models/autoencoders/autoencoder_kl_triposg.py:317
Methodbegin_index
The index for the first timestep. It should be set from pipeline with `set_begin_index` method.
triposg/schedulers/scheduling_rectified_flow.py:137
Functioncompute_density_for_timestep_sampling
( weighting_scheme: str, batch_size: int, logit_mean: float = 0.0, logit_std: float = 1.0,
triposg/schedulers/scheduling_rectified_flow.py:20
Functioncompute_loss_weighting
Computes loss weighting scheme for SD3 training. Courtesy: This was contributed by Rafie Walker in https://github.com/huggingface/diffusers/
triposg/schedulers/scheduling_rectified_flow.py:47
Methodcreate_custom_forward
(module)
triposg/models/transformers/triposg_transformer.py:681
Methodcustom_forward
(*inputs)
triposg/models/transformers/triposg_transformer.py:682
Methoddecode_progressive
(self)
triposg/pipelines/pipeline_triposg.py:135
Functiondefault_set_attn_proc_func
( name: str, hidden_size: int, cross_attention_dim: Optional[int], ori_attn_proc: object, )
triposg/models/transformers/__init__.py:6
Methoddisable_forward_chunking
(self)
triposg/models/transformers/triposg_transformer.py:764
Methoddisable_vae_slicing
r""" Disable sliced VAE decoding. If `enable_vae_slicing` was previously enabled, this method will go back to computing decoding in on
triposg/pipelines/pipeline_utils.py:18
Methoddisable_vae_tiling
r""" Disable tiled VAE decoding. If `enable_vae_tiling` was previously enabled, this method will go back to computing decoding in one
triposg/pipelines/pipeline_utils.py:33
Methoddo_classifier_free_guidance
(self)
triposg/pipelines/pipeline_triposg_scribble.py:129
Methoddo_classifier_free_guidance
(self)
triposg/pipelines/pipeline_triposg.py:123
Methodenable_forward_chunking
Sets the attention processor to use [feed forward chunking](https://huggingface.co/blog/reformer#2-chunked-feed-forward-layers).
triposg/models/transformers/triposg_transformer.py:730
next →1–100 of 146, ranked by callers