MCPcopy Create free account
hub / github.com/FunAudioLLM/ThinkSound / forward

Method forward

ThinkSound/models/mmdit.py:380–469  ·  view source on GitHub ↗

latent: (B, N, C) vf: (B, T, C_V) t: (B,)

(self, latent: torch.Tensor, t: torch.Tensor, clip_f: torch.Tensor, sync_f: torch.Tensor,
                text_f: torch.Tensor, inpaint_masked_input, t5_features, metaclip_global_text_features, cfg_scale:float,cfg_dropout_prob:float,scale_phi:float)

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

Callers

nothing calls this directly

Calls 7

get_empty_t5_sequenceMethod · 0.95
preprocess_conditionsMethod · 0.95
predict_flowMethod · 0.95
safe_catFunction · 0.85

Tested by

no test coverage detected