MCPcopy Create free account

hub / github.com/OpenImagingLab/FlashVSR / functions

Functions2,110 in github.com/OpenImagingLab/FlashVSR

⨍ Functions 2,110◇ Types & classes 521
↓ 2 callersMethod_calculate_score
Calculate the reward score for a single image and prompt. Args: image (torch.Tensor): The processed image tensor. pro
diffsynth/extensions/ImageQualityMetric/mps.py:45
↓ 2 callersMethod_calculate_score
Calculate the HPS score for a single image and prompt. Args: image (torch.Tensor): The processed image tensor. prompt
diffsynth/extensions/ImageQualityMetric/hps.py:62
↓ 2 callersMethod_calculate_score
Calculate the score for a single image and prompt. Args: prompt (str): The prompt text. image (torch.Tensor): The pro
diffsynth/extensions/ImageQualityMetric/imagereward.py:125
↓ 2 callersMethod_calculate_score
Calculate the score for a single image and prompt. Args: image (torch.Tensor): The processed image tensor. prompt (st
diffsynth/extensions/ImageQualityMetric/pickscore.py:22
↓ 2 callersMethod_encode_text
(self, text, normalize=True, embed_cls=True)
diffsynth/extensions/ImageQualityMetric/open_clip/coca_model.py:136
↓ 2 callersMethod_filter2d
(self, input, kernel)
diffsynth/pipelines/svd_video.py:248
↓ 2 callersMethod_gaussian
(self, window_size: int, sigma)
diffsynth/pipelines/svd_video.py:271
↓ 2 callersMethod_global_pool
(self, x: torch.Tensor)
diffsynth/extensions/ImageQualityMetric/open_clip/transformer.py:452
↓ 2 callersFunction_rescan_model_configs
()
diffsynth/extensions/ImageQualityMetric/open_clip/factory.py:33
↓ 2 callersMethod_unflatten_time
(y: torch.Tensor, B: int, f: int)
diffsynth/pipelines/flashvsr_tiny.py:100
↓ 2 callersMethod_unflatten_time
(y: torch.Tensor, B: int, f: int)
diffsynth/pipelines/flashvsr_full.py:100
↓ 2 callersMethod_unflatten_time
(y: torch.Tensor, B: int, f: int)
diffsynth/pipelines/flashvsr_tiny_long.py:100
↓ 2 callersFunction_wavelet_decompose
(x: torch.Tensor, levels: int = 5)
diffsynth/pipelines/flashvsr_tiny.py:66
↓ 2 callersFunction_wavelet_decompose
(x: torch.Tensor, levels: int = 5)
diffsynth/pipelines/flashvsr_full.py:66
↓ 2 callersFunction_wavelet_decompose
(x: torch.Tensor, levels: int = 5)
diffsynth/pipelines/flashvsr_tiny_long.py:66
↓ 2 callersFunction_wavelet_reconstruct
(content: torch.Tensor, style: torch.Tensor, levels: int = 5)
diffsynth/pipelines/flashvsr_tiny.py:78
↓ 2 callersFunction_wavelet_reconstruct
(content: torch.Tensor, style: torch.Tensor, levels: int = 5)
diffsynth/pipelines/flashvsr_full.py:78
↓ 2 callersFunction_wavelet_reconstruct
(content: torch.Tensor, style: torch.Tensor, levels: int = 5)
diffsynth/pipelines/flashvsr_tiny_long.py:78
↓ 2 callersMethodalign_res_stack_to_original_blocks
(self, res_stack, num_blocks, hidden_states)
diffsynth/models/flux_controlnet.py:51
↓ 2 callersMethodapply_rope1d
(self, tokens, pos1d, cos, sin)
diffsynth/models/stepvideo_dit.py:389
↓ 2 callersMethodapply_rope3d
(self, x, fhw_positions, rope_ch_split, parallel=True)
diffsynth/models/stepvideo_dit.py:473
↓ 2 callersMethodapply_rotary_emb
(self, xq, xk, freqs_cis)
diffsynth/models/hunyuan_dit.py:25
↓ 2 callersMethodapply_rotary_emb
(self, x, freqs_cis)
diffsynth/models/cog_dit.py:61
↓ 2 callersFunctionapply_rotary_pos_emb
(x: torch.Tensor, rope_cache: torch.Tensor)
diffsynth/models/kolors_text_encoder.py:419
↓ 2 callersFunctionattention
(q, k, v, attn_mask, mode="torch")
diffsynth/models/step1x_connector.py:12
↓ 2 callersMethodattention
( self, q_x: torch.Tensor, k_x: Optional[torch.Tensor] = None,
diffsynth/extensions/ImageQualityMetric/open_clip/transformer.py:217
↓ 2 callersMethodattn_processor
(self, attn_type)
diffsynth/models/stepvideo_dit.py:330
↓ 2 callersFunctionbase_conv2d
(x, conv_layer, channel_last=False, residual=None)
diffsynth/models/stepvideo_vae.py:62
↓ 2 callersFunctionbase_conv3d_channel_last
(x, conv_layer, residual=None)
diffsynth/models/stepvideo_vae.py:120
↓ 2 callersFunctionbasic_clean
(text)
diffsynth/extensions/ImageQualityMetric/open_clip/tokenizer.py:63
↓ 2 callersMethodbuild_1d_mask
(self, length, left_bound, right_bound, border_width)
diffsynth/models/stepvideo_vae.py:1032
↓ 2 callersMethodbuild_1d_mask
(self, length, left_bound, right_bound, border_width)
diffsynth/models/wan_video_vae.py:651
↓ 2 callersMethodbuild_mask
(self, data, is_bound, border_width)
diffsynth/models/wan_video_vae.py:660
↓ 2 callersFunctionbuild_model_from_openai_state_dict
( state_dict: dict, quick_gelu=True, cast_dtype=torch.float16, )
diffsynth/extensions/ImageQualityMetric/open_clip/model.py:354
↓ 2 callersMethodbuild_single_message
(self, role, metadata, message)
diffsynth/prompters/kolors_prompter.py:196
↓ 2 callersFunctionbytes_to_unicode
Returns list of utf-8 byte and a corresponding list of unicode strings. The reversible bpe codes work on unicode strings. This means you
diffsynth/extensions/ImageQualityMetric/open_clip/tokenizer.py:29
↓ 2 callersMethodcheck
(self, dit: HunyuanVideoDiT, img, vec)
diffsynth/pipelines/hunyuan_video.py:261
↓ 2 callersFunctioncompute_scaled_and_target_dims
(w0: int, h0: int, scale: float = 4.0, multiple: int = 128)
examples/WanVSR/infer_flashvsr_v1.1_tiny_long_video.py:49
↓ 2 callersFunctioncompute_scaled_and_target_dims
(w0: int, h0: int, scale: float = 4.0, multiple: int = 128)
examples/WanVSR/infer_flashvsr_v1.1_tiny.py:49
↓ 2 callersFunctioncompute_scaled_and_target_dims
(w0: int, h0: int, scale: int = 4, multiple: int = 128)
examples/WanVSR/infer_flashvsr_full.py:48
↓ 2 callersFunctioncompute_scaled_and_target_dims
(w0: int, h0: int, scale: float = 4.0, multiple: int = 128)
examples/WanVSR/infer_flashvsr_tiny_long_video.py:49
↓ 2 callersFunctioncompute_scaled_and_target_dims
(w0: int, h0: int, scale: int = 4, multiple: int = 128)
examples/WanVSR/infer_flashvsr_v1.1_full.py:48
↓ 2 callersFunctioncompute_scaled_and_target_dims
(w0: int, h0: int, scale: float = 4.0, multiple: int = 128)
examples/WanVSR/infer_flashvsr_tiny.py:49
↓ 2 callersMethodconvert_state_dict
(self, state_dict, lora_prefix="lora_unet_", alpha=1.0)
diffsynth/models/lora.py:23
↓ 2 callersFunctionconvert_weights_to_lp
Convert applicable model parameters to low-precision (bf16 or fp16)
diffsynth/extensions/ImageQualityMetric/open_clip/model.py:308
↓ 2 callersFunctioncount_conv3d
(model)
diffsynth/models/wan_video_vae.py:485
↓ 2 callersFunctioncreate_model
( model_name: str, pretrained: Optional[str] = None, precision: str = 'fp32',
diffsynth/extensions/ImageQualityMetric/open_clip/factory.py:108
↓ 2 callersFunctioncreate_model_and_transforms
( model_name: str, pretrained: Optional[str] = None, precision: str = 'fp32',
diffsynth/extensions/ImageQualityMetric/open_clip/factory.py:309
↓ 2 callersMethodcropped_pos_embed
Crops positional embeddings for SD3 compatibility.
diffsynth/models/omnigen.py:426
↓ 2 callersMethoddecode_images
(self, images)
diffsynth/extensions/RIFE/__init__.py:144
↓ 2 callersMethoddecode_images
(self, images)
diffsynth/processors/RIFE.py:38
↓ 2 callersMethoddecode_small_video
(self, sample)
diffsynth/models/cog_vae.py:254
↓ 2 callersMethoddecode_video
(self, latents, tiled=False, tile_size=64, tile_stride=32)
diffsynth/pipelines/sdxl_video.py:78
↓ 2 callersMethoddecode_video
Decode a sequence of frames from latents. x: NTCHW latent tensor; returns NTCHW RGB in ~[0, 1].
examples/WanVSR/utils/TCDecoder.py:254
↓ 2 callersMethoddenoising_model
(self)
diffsynth/pipelines/flashvsr_tiny.py:223
↓ 2 callersMethoddenoising_model
(self)
diffsynth/pipelines/flashvsr_full.py:244
↓ 2 callersMethoddenoising_model
(self)
diffsynth/pipelines/flashvsr_tiny_long.py:223
↓ 2 callersMethoddist_loss
(self, teacher_logits, student_logits)
diffsynth/extensions/ImageQualityMetric/open_clip/loss.py:236
↓ 2 callersFunctiondownload_pretrained_from_url
( url: str, cache_dir: Union[str, None] = None, )
diffsynth/extensions/ImageQualityMetric/open_clip/pretrained.py:283
↓ 2 callersMethodenable_auto_offload
(self, dtype=torch.bfloat16, device="cuda")
diffsynth/models/hunyuan_video_dit.py:664
↓ 2 callersMethodencode_image
(self, image, tiled=False, tile_size=64, tile_stride=32)
diffsynth/pipelines/omnigen_image.py:157
↓ 2 callersMethodencode_prompt
(self, prompt, positive=True)
diffsynth/pipelines/step_video.py:135
↓ 2 callersMethodencode_prompt
(self, prompt, positive=True, clip_sequence_length=77, llm_sequence_length=256, input_images=None)
diffsynth/pipelines/hunyuan_video.py:105
↓ 2 callersMethodencode_prompt
(self, prompt, positive=True)
diffsynth/pipelines/cog_video.py:52
↓ 2 callersMethodencode_prompt
(self, prompt, positive=True)
diffsynth/pipelines/wan_video.py:201
↓ 2 callersMethodencode_prompt_using_clip
(self, prompt, text_encoder, tokenizer, max_length, device)
diffsynth/prompters/sd3_prompter.py:39
↓ 2 callersMethodencode_prompt_using_signle_model
(self, prompt, text_encoder, tokenizer, max_length, clip_skip, device)
diffsynth/prompters/hunyuan_dit_prompter.py:34
↓ 2 callersMethodencode_small_video
(self, sample)
diffsynth/models/cog_vae.py:339
↓ 2 callersMethodencode_video
(self, processed_images, tiled=False, tile_size=64, tile_stride=32)
diffsynth/pipelines/sdxl_video.py:86
↓ 2 callersMethodencode_video
(self, frames, tile_size=(17, 30, 30), tile_stride=(12, 20, 20))
diffsynth/pipelines/hunyuan_video.py:125
↓ 2 callersMethodencode_video_with_vae
(self, video)
diffsynth/pipelines/svd_video.py:60
↓ 2 callersMethodevict_previous_layer
Moves the previous layer cache to the CPU
diffsynth/pipelines/omnigen_image.py:38
↓ 2 callersMethodextract_number_from_filenames
(self, file_names)
diffsynth/extensions/FastBlend/api.py:113
↓ 2 callersFunctionextract_weight_to_half
(weight: torch.Tensor, scale_list: torch.Tensor, source_bit_width: int)
diffsynth/models/kolors_text_encoder.py:122
↓ 2 callersMethodfetch_tokenizer
(self, tokenizer_path=None)
diffsynth/prompters/wan_prompter.py:92
↓ 2 callersMethodforward
diffsynth/models/omnigen.py:489
↓ 2 callersMethodforward
imgs: [B, 3, H, W] of torch.float32. - mean: [0.48145466, 0.4578275, 0.40821073] - std: [0.26862954, 0.2613025
diffsynth/models/wan_video_image_encoder.py:571
↓ 2 callersMethodforward
( self, input_ids=None, attention_mask=None, )
diffsynth/models/stepvideo_text_encoder.py:508
↓ 2 callersMethodforward
(self, x)
diffsynth/models/wan_video_vae.py:68
↓ 2 callersFunctionfp16_clamp
(x)
diffsynth/models/wan_video_text_encoder.py:8
↓ 2 callersMethodfrom_civitai
(self, state_dict)
diffsynth/models/sd3_dit.py:472
↓ 2 callersMethodfrom_diffusers
(self, state_dict)
diffsynth/models/sd3_dit.py:408
↓ 2 callersMethodfrom_diffusers
(self, state_dict)
diffsynth/models/cog_dit.py:347
↓ 2 callersMethodfrom_model_manager
(model_manager: ModelManager, controlnet_config_units: List[ControlNetConfigUnit]=[], prompt_refiner_classes=[
diffsynth/pipelines/sd_video.py:116
↓ 2 callersMethodfrom_pretrained
(path, torch_dtype=torch.bfloat16, device="cuda")
diffsynth/models/qwenvl.py:35
↓ 2 callersFunctiongate
(x, gate)
diffsynth/models/stepvideo_dit.py:615
↓ 2 callersMethodget_1d_rotary_pos_embed
(self, dim: int, pos, theta: float = 10000.0, use_real=False)
diffsynth/pipelines/hunyuan_image.py:97
↓ 2 callersFunctionget_1d_sincos_pos_embed_from_grid
embed_dim: output dimension for each position pos: a list of positions to be encoded: size (M,) out: (M, D)
diffsynth/models/omnigen.py:286
↓ 2 callersFunctionget_activation
Helper function to get activation function from string. Args: act_fn (str): Name of activation function. Returns: nn.Module:
diffsynth/models/stepvideo_dit.py:86
↓ 2 callersMethodget_block_name
(self, names)
diffsynth/models/svd_unet.py:423
↓ 2 callersFunctionget_cast_dtype
(precision: str)
diffsynth/extensions/ImageQualityMetric/open_clip/model.py:66
↓ 2 callersMethodget_cast_dtype
(self)
diffsynth/extensions/ImageQualityMetric/open_clip/transformer.py:310
↓ 2 callersMethodget_cos_sin
(self, D, seq_len, device, dtype)
diffsynth/models/stepvideo_dit.py:373
↓ 2 callersMethodget_error
(self, source_guide, target_guide, source_style, target_style, nnf)
diffsynth/extensions/FastBlend/patch_match.py:76
↓ 2 callersMethodget_extended_attention_mask
Makes broadcastable attention and causal masks so that future and masked tokens are ignored. Arguments: attention_mask (
diffsynth/extensions/ImageQualityMetric/BLIP/med.py:601
↓ 2 callersMethodget_ground_truth
(self, device, num_logits)
diffsynth/extensions/ImageQualityMetric/open_clip/loss.py:90
↓ 2 callersMethodget_name_dict
(self, lora_state_dict)
diffsynth/models/lora.py:205
↓ 2 callersFunctionget_norm_layer
Get the normalization layer. Args: norm_layer (str): The type of normalization layer. Returns: norm_layer (nn.Module):
diffsynth/models/step1x_connector.py:246
↓ 2 callersFunctionget_pairs
Return set of symbol pairs in a word. Word is represented as tuple of symbols (symbols being variable-length strings).
diffsynth/extensions/ImageQualityMetric/open_clip/tokenizer.py:51
↓ 2 callersMethodget_patch_error
(self, source, nnf, target)
diffsynth/extensions/FastBlend/patch_match.py:53
← previousnext →201–300 of 2,110, ranked by callers