MCPcopy Create free account

hub / github.com/casterpollux/MiniMax-bmo / functions

Functions35 in github.com/casterpollux/MiniMax-bmo

↓ 3 callersMethod__init__
( self, dim: int, time_freq_dim: int, time_proj_dim: int )
transformer_minimax_remover.py:66
↓ 3 callersMethodresize
Official resize method - exact copy from official code with dimension validation
pipeline_minimax_remover_bmo.py:91
↓ 3 callersMethodresolve_model_path
Resolve model path with intelligent fallbacks Args: model_path: User-provided path (can be relative or absol
minimax_mask_node_bmo.py:315
↓ 2 callersMethodauto_download_models
Automatically download models if they don't exist Returns the paths to the downloaded models
minimax_mask_node_bmo.py:123
↓ 1 callersMethodauto_resize_for_compatibility
Automatically resize images and masks to be compatible with each other and the VAE This handles any input resolution combination and
pipeline_minimax_remover_bmo.py:213
↓ 1 callersFunctiondownload_minimax_models
Download Minimax Remover models if not already present
download_models.py:22
↓ 1 callersFunctiondownload_sam_model
Download SAM model if not already present
download_models.py:6
↓ 1 callersMethodexpand_masks
Official mask expansion method - exact copy from official code
pipeline_minimax_remover_bmo.py:70
↓ 1 callersFunctionget_comfyui_base_path
Get the ComfyUI base path for proper imports
minimax_mask_node_bmo.py:25
↓ 1 callersMethodget_torch_device
()
minimax_mask_node_bmo.py:58
↓ 1 callersFunctionlazy_import_diffusers
Lazy import diffusers components to avoid import issues
minimax_mask_node_bmo.py:100
↓ 1 callersMethodload_models
Load MiniMax-Remover models from paths with auto-download support
minimax_mask_node_bmo.py:356
↓ 1 callersFunctionmain
()
download_models.py:80
↓ 1 callersFunctionmain
()
setup_comfyui_integration_bmo.py:12
↓ 1 callersMethodprepare_latents
( self, batch_size: int, num_channels_latents: int = 16, height: int = 720,
pipeline_minimax_remover_bmo.py:44
MethodINPUT_TYPES
(cls)
minimax_mask_node_bmo.py:260
Method__call__
( self, attn: Attention, hidden_states: torch.Tensor, rotary_emb: Optional[tor
transformer_minimax_remover.py:22
Method__call__
BMO inference method - supports any resolution input with automatic compatibility
pipeline_minimax_remover_bmo.py:290
Method__init__
(self)
minimax_mask_node_bmo.py:118
Method__init__
(self)
transformer_minimax_remover.py:18
Method__init__
( self, attention_head_dim: int, patch_size: Tuple[int, int, int], max_seq_len: int, theta: float = 10
transformer_minimax_remover.py:96
Method__init__
( self, dim: int, ffn_dim: int, num_heads: int, qk_norm: str = "rms_no
transformer_minimax_remover.py:139
Method__init__
( self, patch_size: Tuple[int] = (1, 2, 2), num_attention_heads: int = 40, att
transformer_minimax_remover.py:200
Method__init__
( self, transformer: Transformer3DModel, vae: AutoencoderKLWan, scheduler: Uni
pipeline_minimax_remover_bmo.py:25
Methodapply_rotary_emb
(hidden_states: torch.Tensor, freqs: torch.Tensor)
transformer_minimax_remover.py:47
Methodcurrent_timestep
(self)
pipeline_minimax_remover_bmo.py:131
Methoddiagnose_inputs
Diagnostic function to check input compatibility Helps users identify potential dimension issues before processing
minimax_mask_node_bmo.py:589
Methodforward
( self, timestep: torch.Tensor, )
transformer_minimax_remover.py:80
Methodforward
(self, hidden_states: torch.Tensor)
transformer_minimax_remover.py:116
Methodforward
( self, hidden_states: torch.Tensor, temb: torch.Tensor, rotary_emb: torch.Ten
transformer_minimax_remover.py:170
Methodforward
( self, hidden_states: torch.Tensor, timestep: torch.LongTensor )
transformer_minimax_remover.py:247
Methodinterrupt
(self)
pipeline_minimax_remover_bmo.py:135
Methodnum_timesteps
(self)
pipeline_minimax_remover_bmo.py:127
Methodprocess_video
Process video with BMO MiniMax-Remover with auto-download support Args: images: Input video frames [B, H, W,
minimax_mask_node_bmo.py:463
Methodvalidate_inputs
Validate input tensors and dimensions for compatibility with auto-resize
pipeline_minimax_remover_bmo.py:138