From 6fa8ee48e117fa9c3bef450e02776e8c26b278e7 Mon Sep 17 00:00:00 2001 From: zhifu gao <zhifu.gzf@alibaba-inc.com> Date: 星期六, 30 三月 2024 10:13:42 +0800 Subject: [PATCH] Dev gzf new (#1567) --- funasr/datasets/large_datasets/datapipes/map.py | 22 ++++++++++++++++++++++ 1 files changed, 22 insertions(+), 0 deletions(-) diff --git a/funasr/datasets/large_datasets/datapipes/map.py b/funasr/datasets/large_datasets/datapipes/map.py new file mode 100644 index 0000000..dfcd6a0 --- /dev/null +++ b/funasr/datasets/large_datasets/datapipes/map.py @@ -0,0 +1,22 @@ +from torch.utils.data import IterableDataset + + +def default_fn(data): + return data + + +class MapperIterDataPipe(IterableDataset): + + def __init__(self, + datapipe, + fn=default_fn): + self.datapipe = datapipe + self.fn = fn + + def set_epoch(self, epoch): + self.datapipe.set_epoch(epoch) + + def __iter__(self): + assert callable(self.fn) + for data in self.datapipe: + yield self.fn(data) -- Gitblit v1.9.1