Functions — github.com/Xiaojiu-z/EasyControl

CodeHub MCPcopy Create free account

hub / github.com/Xiaojiu-z/EasyControl / functions

Functions139 in github.com/Xiaojiu-z/EasyControl

⨍ Functions 139◇ Types & classes 15

↓ 8 callersMethod_pack_latents

(latents, batch_size, num_channels_latents, height, width)

train/src/pipeline.py:459

↓ 6 callersMethodset_attn_processor

r""" Sets the attention processor to use to compute attention. Parameters: processor (`dict` of `AttentionProcessor` or o

train/src/transformer_flux.py:321

↓ 5 callersMethod_pack_latents

(latents, batch_size, num_channels_latents, height, width)

src/pipeline.py:376

↓ 5 callersMethodset_attn_processor

r""" Sets the attention processor to use to compute attention. Parameters: processor (`dict` of `AttentionProcessor` or o

src/transformer_flux.py:321

↓ 4 callersFunctionmultiple_16

train/src/jsonl_datasets.py:9

↓ 3 callersFunctionload_checkpoint

train/src/lora_helper.py:21

↓ 3 callersFunctionload_image_safely

(image_path, size)

train/src/jsonl_datasets.py:16

↓ 3 callersFunctionset_single_lora

(transformer, local_path, lora_weights=[], cond_size=512)

src/lora_helper.py:176

↓ 2 callersMethod__init__

( self, dim: int, num_attention_heads: int, attention_head_dim: int, qk_norm: str = "rms_norm", eps: f

src/transformer_flux.py:108

↓ 2 callersMethod__init__

( self, in_features: int, out_features: int, rank: int = 4, network_al

src/layers_cache.py:12

↓ 2 callersMethod__init__

( self, dim: int, num_attention_heads: int, attention_head_dim: int, qk_norm: str = "rms_norm", eps: f

train/src/transformer_flux.py:108

↓ 2 callersMethod__init__

( self, in_features: int, out_features: int, rank: int = 4, network_al

train/src/layers.py:12

↓ 2 callersMethod_encode_vae_image

(self, image: torch.Tensor, generator: torch.Generator)

src/pipeline.py:309

↓ 2 callersMethod_encode_vae_image

(self, image: torch.Tensor, generator: torch.Generator)

train/src/pipeline.py:385

↓ 2 callersMethod_unpack_latents

(latents, height, width, vae_scale_factor)

train/src/pipeline.py:466

↓ 2 callersFunctionget_lora_rank

src/lora_helper.py:16

↓ 2 callersFunctionget_lora_rank

train/src/lora_helper.py:16

↓ 2 callersFunctionimport_model_class_from_model_name_or_path

( pretrained_model_name_or_path: str, revision: str, subfolder: str = "text_encoder" )

train/train.py:101

↓ 2 callersFunctionload_checkpoint

src/lora_helper.py:21

↓ 2 callersFunctionprepare_latent_subject_ids

(height, width, device, dtype)

train/src/pipeline.py:52

↓ 2 callersMethodprocess_image

(self, prompt='', subject_imgs=[], spatial_imgs=[], height=768, width=768, output_path=None, seed=42)

↓ 2 callersFunctionresize_position_encoding

(batch_size, original_height, original_width, target_height, target_width, device, dtype)

train/src/pipeline.py:62

↓ 2 callersFunctionretrieve_latents

( encoder_output: torch.Tensor, generator: Optional[torch.Generator] = None, sample_mode: str = "sampl

src/pipeline.py:82

↓ 2 callersFunctionretrieve_latents

( encoder_output: torch.Tensor, generator: Optional[torch.Generator] = None, sample_mode: str = "sampl

train/src/pipeline.py:84

↓ 2 callersFunctionset_multi_lora

(transformer, local_paths, lora_weights=[[]], cond_size=512)

src/lora_helper.py:180

↓ 1 callersFunction_encode_prompt_with_clip

( text_encoder, tokenizer, prompt: str, device=None, text_input_ids=No

train/src/prompt_helper.py:97

↓ 1 callersFunction_encode_prompt_with_t5

( text_encoder, tokenizer, max_sequence_length=512, prompt=None, num_i

train/src/prompt_helper.py:56

↓ 1 callersMethod_get_clip_prompt_embeds

( self, prompt: Union[str, List[str]], num_images_per_prompt: int = 1,

src/pipeline.py:209

↓ 1 callersMethod_get_clip_prompt_embeds

( self, prompt: Union[str, List[str]], num_images_per_prompt: int = 1,

train/src/pipeline.py:264

↓ 1 callersMethod_get_t5_prompt_embeds

( self, prompt: Union[str, List[str]] = None, num_images_per_prompt: int =

src/pipeline.py:163

↓ 1 callersMethod_get_t5_prompt_embeds

( self, prompt: Union[str, List[str]] = None, num_images_per_prompt: int =

train/src/pipeline.py:218

↓ 1 callersMethod_unpack_latents

(latents, height, width, vae_scale_factor)

src/pipeline.py:383

↓ 1 callersFunctioncalculate_shift

( image_seq_len, base_seq_len: int = 256, max_seq_len: int = 4096, base_shift:

src/pipeline.py:34

↓ 1 callersFunctioncalculate_shift

( image_seq_len, base_seq_len: int = 256, max_seq_len: int = 4096, base_shift:

train/src/pipeline.py:34

↓ 1 callersMethodcheck_inputs

( self, prompt, prompt_2, height, width,

src/pipeline.py:323

↓ 1 callersMethodcheck_inputs

( self, prompt, prompt_2, height, width,

train/src/pipeline.py:399

↓ 1 callersFunctionclear_cache

↓ 1 callersFunctionclear_cache

infer_multi.py:12

↓ 1 callersMethodclear_cache

(self, transformer)

↓ 1 callersMethodencode_prompt

( self, prompt: Union[str, List[str]], prompt_2: Union[str, List[str]],

src/pipeline.py:250

↓ 1 callersMethodencode_prompt

r""" Args: prompt (`str` or `List[str]`, *optional*): prompt to be encoded prompt_2 (`str` or `List[s

train/src/pipeline.py:305

↓ 1 callersFunctionencode_token_ids

(text_encoders, tokens, accelerator, num_images_per_prompt=1, device=None)

train/src/prompt_helper.py:173

↓ 1 callersFunctionget_random_resolution

(min_size=512, max_size=1280, multiple=16)

train/src/jsonl_datasets.py:12

↓ 1 callersFunctionget_sigmas

(timesteps, n_dim=4, dtype=torch.float32)

train/train.py:797

↓ 1 callersFunctionload_and_transform_cond_images

train/src/jsonl_datasets.py:87

↓ 1 callersFunctionload_and_transform_subject_images

train/src/jsonl_datasets.py:92

↓ 1 callersFunctionload_safetensors

src/lora_helper.py:9

↓ 1 callersFunctionload_safetensors

train/src/lora_helper.py:9

↓ 1 callersFunctionload_text_encoders

(args, class_one, class_two)

train/src/prompt_helper.py:4

↓ 1 callersFunctionlog_validation

( pipeline, args, accelerator, pipeline_args, step, torch_dtyp

train/train.py:57

↓ 1 callersFunctionmain

train/train.py:474

↓ 1 callersFunctionmake_train_dataset

(args, tokenizer, accelerator=None)

train/src/jsonl_datasets.py:26

↓ 1 callersFunctionparse_args

(input_args=None)

train/train.py:120

↓ 1 callersFunctionprepare_latent_image_ids_

(height, width, device, dtype)

src/pipeline.py:46

↓ 1 callersFunctionprepare_latent_image_ids_2

(height, width, device, dtype)

train/src/pipeline.py:46

↓ 1 callersFunctionprepare_latent_subject_ids

(height, width, device, dtype)

src/pipeline.py:52

↓ 1 callersMethodprepare_latents

( self, batch_size, num_channels_latents, height,

src/pipeline.py:425

↓ 1 callersMethodprepare_latents

( self, batch_size, num_channels_latents, height,

train/src/pipeline.py:508

↓ 1 callersFunctionresize_position_encoding

(batch_size, original_height, original_width, target_height, target_width, device, dtype)

src/pipeline.py:62

↓ 1 callersFunctionretrieve_timesteps

( scheduler, num_inference_steps: Optional[int] = None, device: Optional[Union[str, to

src/pipeline.py:96

↓ 1 callersFunctionretrieve_timesteps

Calls the scheduler's `set_timesteps` method and retrieves timesteps from the scheduler after the call. Handles custom timesteps. Any kwargs

train/src/pipeline.py:98

↓ 1 callersFunctiontokenize_prompt_clip_t5

train/src/jsonl_datasets.py:100

↓ 1 callersFunctiontrain_transforms

(image, noise_size)

train/src/jsonl_datasets.py:73

↓ 1 callersFunctionupdate_model_with_lora

(checkpoint, lora_weights, transformer, cond_size)

src/lora_helper.py:31

↓ 1 callersFunctionupdate_model_with_lora

(checkpoint, lora_weights, transformer, cond_size)

train/src/lora_helper.py:31

↓ 1 callersFunctionupdate_model_with_multi_lora

(checkpoints, lora_weights, transformer, cond_size)

src/lora_helper.py:96

↓ 1 callersFunctionupdate_model_with_multi_lora

(checkpoints, lora_weights, transformer, cond_size)

train/src/lora_helper.py:96

( self, prompt: Union[str, List[str]] = None, prompt_2: Optional[Union[str

src/pipeline.py:509

(self, attn: Attention, hidden_states: torch.FloatTensor, encoder_hidden_states: torch

src/layers_cache.py:92

(self, attn: Attention, hidden_states: torch.FloatTensor, encoder_hidden_states: torch

src/layers_cache.py:223

(self, attn: Attention, hidden_states: torch.FloatTensor, encoder_hidden_states: torch

train/src/layers.py:90

(self, attn: Attention, hidden_states: torch.FloatTensor, encoder_hidden_states: torch

train/src/layers.py:180

( self, prompt: Union[str, List[str]] = None, prompt_2: Optional[Union[str

train/src/pipeline.py:592

(self, dim, num_attention_heads, attention_head_dim, mlp_ratio=4.0)

src/transformer_flux.py:31

( self, patch_size: int = 1, in_channels: int = 64, out_channels: Optional[int

src/transformer_flux.py:238

( self, scheduler: FlowMatchEulerDiscreteScheduler, vae: AutoencoderKL,

src/pipeline.py:133

(self, dim: int, ranks=[], lora_weights=[], network_alphas=[], device=None, dtype=None, cond_width=512, cond_h

src/layers_cache.py:68

(self, dim: int, ranks=[], lora_weights=[], network_alphas=[], device=None, dtype=None, cond_width=512, cond_h

src/layers_cache.py:195

(self, dim, num_attention_heads, attention_head_dim, mlp_ratio=4.0)

train/src/transformer_flux.py:31

( self, patch_size: int = 1, in_channels: int = 64, out_channels: Optional[int

train/src/transformer_flux.py:238

(self, dim: int, ranks=[], lora_weights=[], network_alphas=[], device=None, dtype=None, cond_width=512, cond_h

train/src/layers.py:68

(self, dim: int, ranks=[], lora_weights=[], network_alphas=[], device=None, dtype=None, cond_width=512, cond_h

train/src/layers.py:154

( self, scheduler: FlowMatchEulerDiscreteScheduler, vae: AutoencoderKL,

train/src/pipeline.py:188

Method_prepare_latent_image_ids

(batch_size, height, width, device, dtype)

src/pipeline.py:365

Method_prepare_latent_image_ids

(batch_size, height, width, device, dtype)

train/src/pipeline.py:448

Method_set_gradient_checkpointing

(self, module, value=False)

src/transformer_flux.py:395

Method_set_gradient_checkpointing

(self, module, value=False)

train/src/transformer_flux.py:395

Methodattn_processors

r""" Returns: `dict` of attention processors: A dictionary containing all attention processors used in the model with

src/transformer_flux.py:297

Methodattn_processors

r""" Returns: `dict` of attention processors: A dictionary containing all attention processors used in the model with

train/src/transformer_flux.py:297

Functioncollate_fn

train/src/jsonl_datasets.py:163

Methodcreate_custom_forward

(module, return_dict=None)

src/transformer_flux.py:484

Methodcreate_custom_forward

(module, return_dict=None)

train/src/transformer_flux.py:484

Methodcustom_forward

src/transformer_flux.py:485

Methodcustom_forward

train/src/transformer_flux.py:485

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

src/pipeline.py:403

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

train/src/pipeline.py:486

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

src/pipeline.py:418

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

train/src/pipeline.py:501

Methodenable_vae_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

src/pipeline.py:396

next →1–100 of 139, ranked by callers