mrq/vall-e
5
9
Fork 0
Code Issues 8 Pull Requests Packages Projects Releases Wiki Activity

un 'experimental' the better target sequence preparation

Browse Source
This commit is contained in:
mrq 2023-10-22 09:06:59 -05:00
parent 9a6040383e
commit ed54f4ebec
1 changed files with 9 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
vall_e/models/ar_nar.py
View File

@ -127,15 +127,18 @@ class AR_NAR(Base):
targ_list = [r[..., l] for r, l in zip(resps_list, quant_levels)] # ensures we only have 1 RVQ-bin (our target)
resps_list = [r if l == 0 else r[..., :l] for r, l in zip(resps_list, quant_levels)] # r[..., 0] is technically correct, but only r[:, 0] gets passed through the embedding
"""
if cfg.experimental:
proms_list = [ r if l == 0 else trim(r, 75 * 3) for r, l in zip(proms_list, quant_levels) ] # trim input prompt to 3 seconds
# append stop tokens for AR
for i in range(batch_size):
if quant_levels[i] > 0:
continue
"""
# append stop tokens for AR
for i in range(batch_size):
if quant_levels[i] > 0:
continue
resps_list[i] = torch.cat([resps_list[i], torch.Tensor([[self.stop_token] * n_levels]).to(device=device, dtype=torch.int16) ])
targ_list[i] = torch.cat([targ_list[i], torch.Tensor([self.stop_token]).to(device=device, dtype=torch.int16) ])
resps_list[i] = torch.cat([resps_list[i], torch.Tensor([[self.stop_token] * n_levels]).to(device=device, dtype=torch.int16) ])
targ_list[i] = torch.cat([targ_list[i], torch.Tensor([self.stop_token]).to(device=device, dtype=torch.int16) ])
return super().forward(
text_list=text_list,