speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

resume from step

Browse Source
This commit is contained in:
游雁 2024-04-26 11:31:09 +08:00
parent 1cdb3cc28d
commit ab6391da4f
1 changed files with 1 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
funasr/datasets/sense_voice_datasets/datasets.py
View File

@ -159,7 +159,7 @@ class SenseVoiceDataset(torch.utils.data.Dataset):
def _filter_badcase(self, outputs, i=0):
b, t, _ = outputs["speech"].shape
if b * t > self.batch_size * 1.25:
beg = torch.randint(0, 2, ()).item()
if b < 2:
@ -170,7 +170,6 @@ class SenseVoiceDataset(torch.utils.data.Dataset):
for key, data_list in outputs.items():
outputs[key] = outputs[key][beg : beg + b : 2]
speech_lengths_max = outputs["speech_lengths"].max().item()
outputs["speech"] = outputs["speech"][:, :speech_lengths_max, :]
text_lengths_max = outputs["text_lengths"].max().item()