(device, dataset, T, S)

(self, audio, template, one_hot)

( shape: Tuple[int, int], mask_prob: float, mask_length: int, attention_mask: Optional[torch.T

(args)

(n)

(self, args)

(model)

(args,face_vert_mmap,data2array_verts)

(args,raw_audio)

(n)

(n_head, max_seq_len, period)

(features, input_fps, output_fps, output_len=None)

(args)

()

()

()

()

(args)

(args,mesh, t_center, rot=np.zeros(3), tex_img=None, z_offset=0)

(args,mesh, t_center, rot=np.zeros(3), tex_img=None, z_offset=0)

(args)

(args,sequence_vertices, template, out_path,predicted_vertices_path,vt, ft ,tex_img)

(args, model, test_loader,epoch)

(args)

(args, train_loader, dev_loader, model, optimizer, criterion, epoch=100)

(self, config)

(self, data,subjects_dict,data_type="train")

(self, d_model, dropout=0.1, period=25, max_seq_len=600)

(self)

( self, input_values, dataset, attention_mask=None, output_attentions=

(self, x)

(self, audio, template, vertice, one_hot, criterion,teacher_forcing=True)