↓ 1 callersMethodcompute_freqs_mot(self, f, h, w, end: int = 1024, theta: float = 10000.0)
DiffSynth-Studio/diffsynth/models/wan_video_mot.py:146
↓ 1 callersMethodcompute_regularization(self, pipe: BasePipeline, trajectory_teacher, inputs_shared, inputs_posi, inputs_nega, num_inference_steps, c
DiffSynth-Studio/diffsynth/diffusion/loss.py:91
↓ 1 callersFunctionenable_vram_management_recursively(model: torch.nn.Module, module_map: dict, vram_config: dict, vram_limit=None, name_prefix="", disk_map=None,
DiffSynth-Studio/diffsynth/core/vram/layers.py:439
↓ 1 callersMethodencode_prompt(
self,
text_encoder,
tokenizer,
prompt: Union[str, List[str]],
dtype
DiffSynth-Studio/diffsynth/pipelines/flux2_image.py:243
↓ 1 callersMethodencode_prompt(
self,
tokenizer_1,
tokenizer_2,
text_encoder_1,
text_encoder_2,
DiffSynth-Studio/diffsynth/pipelines/flux_image.py:372
↓ 1 callersMethodencode_prompt(
self,
tokenizer_1,
tokenizer_2,
text_encoder_1,
text_encoder_2,
DiffSynth-Studio/diffsynth/pipelines/flux_image.py:553
↓ 1 callersMethodencode_prompt_using_clip(self, prompt, text_encoder, tokenizer, max_length, device)
DiffSynth-Studio/diffsynth/pipelines/flux_image.py:350
↓ 1 callersMethodencode_prompt_using_clip(self, prompt, text_encoder, tokenizer, max_length, device)
DiffSynth-Studio/diffsynth/pipelines/flux_image.py:531
↓ 1 callersMethodencode_prompt_using_t5(self, prompt, text_encoder, tokenizer, max_length, device)
DiffSynth-Studio/diffsynth/pipelines/flux_image.py:361
↓ 1 callersMethodencode_prompt_using_t5(self, prompt, text_encoder, tokenizer, max_length, device)
DiffSynth-Studio/diffsynth/pipelines/flux_image.py:542
↓ 1 callersFunctionexample(pipe, seeds, example_id, global_prompt, entity_prompts, height=784, width=1280)
DiffSynth-Studio/examples/qwen_image/model_inference/Qwen-Image-EliGen-Poster.py:66
↓ 1 callersFunctionexample(pipe, seeds, example_id, global_prompt, entity_prompts, height=784, width=1280)
DiffSynth-Studio/examples/qwen_image/model_inference_low_vram/Qwen-Image-EliGen-Poster.py:66
↓ 1 callersMethodextract_audio_feat(self, input_audio, sample_rate, processor, return_all_layers=False, dtype=torch.float32, device='cpu')
DiffSynth-Studio/diffsynth/models/wav2vec.py:102
↓ 1 callersMethodfetch_trajectory(self, pipe: BasePipeline, timesteps_student, inputs_shared, inputs_posi, inputs_nega, num_inference_steps, cf
DiffSynth-Studio/diffsynth/diffusion/loss.py:46
↓ 1 callersFunctionflash_attention_2(q: torch.Tensor, k: torch.Tensor, v: torch.Tensor, q_pattern="b n s d", k_pattern="b n s d", v_pattern="b n s
DiffSynth-Studio/diffsynth/core/attention/attention.py:84
↓ 1 callersFunctionflash_attention_3(q: torch.Tensor, k: torch.Tensor, v: torch.Tensor, q_pattern="b n s d", k_pattern="b n s d", v_pattern="b n s
DiffSynth-Studio/diffsynth/core/attention/attention.py:74
↓ 1 callersFunctionflow_to_color 将光流 (H, W, 2) 转换为颜色编码图。 参数: - flow: numpy.ndarray, 光流张量,形状为 (H, W, 2),值范围为 [-1, 1] - max_flow: float, 光流的最大值,用于归一化 (默认假设范围是 [-1,
VideoCus-Factory/track_to_map.py:95
↓ 1 callersMethodforward(self, sample, tiled=False, tile_size=64, tile_stride=32, **kwargs)
DiffSynth-Studio/diffsynth/models/flux_vae.py:344
↓ 1 callersMethodforward(self, sample, tiled=False, tile_size=64, tile_stride=32, **kwargs)
DiffSynth-Studio/diffsynth/models/flux_vae.py:412
↓ 1 callersMethodforward(
self, x, vace_context, context, t_mod, freqs,
use_gradient_checkpointing: bool = False,
DiffSynth-Studio/diffsynth/models/wan_video_vace.py:53
↓ 1 callersFunctiongenerate_camera_coordinates(
direction: Literal["Left", "Right", "Up", "Down", "LeftUp", "LeftDown", "RightUp", "RightDown", "In", "O
DiffSynth-Studio/diffsynth/models/wan_video_camera_controller.py:184
↓ 1 callersMethodget_audio_feats_per_inference(self, input_audio, sample_rate, processor, fps=16, batch_frames=80, m=0, dtype=torch.float32, device='cpu')
DiffSynth-Studio/diffsynth/models/wav2vec.py:186
↓ 1 callersFunctionget_cropped_flows trajs: (B, N, T, 2) in pixel coords returns: (C, n_frames_out, H, W) in [-1,1] We always generate exactly n_frames_out frames for output
VideoCus-Factory/track_to_map_local.py:61