Replace load_manifest_lazy with load_manifest for MUSAN. (#412)

egs/aidatatang_200zh/ASR/pruned_transducer_stateless2/asr_datamodule.py

@@ -27,6 +27,7 @@ from lhotse import (
     CutSet,
     Fbank,
     FbankConfig,
+    load_manifest,
     load_manifest_lazy,
     set_caching_enabled,
 )
@@ -204,7 +205,7 @@ class Aidatatang_200zhAsrDataModule:
             The state dict for the training sampler.
         """
         logging.info("About to get Musan cuts")
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             self.args.manifest_dir / "musan_cuts.jsonl.gz"
         )
 

egs/aishell/ASR/tdnn_lstm_ctc/asr_datamodule.py

@@ -23,7 +23,7 @@ from functools import lru_cache
 from pathlib import Path
 from typing import List
 
-from lhotse import CutSet, Fbank, FbankConfig, load_manifest_lazy
+from lhotse import CutSet, Fbank, FbankConfig, load_manifest, load_manifest_lazy
 from lhotse.dataset import (
     CutConcatenate,
     CutMix,
@@ -183,7 +183,7 @@ class AishellAsrDataModule:
 
     def train_dataloaders(self, cuts_train: CutSet) -> DataLoader:
         logging.info("About to get Musan cuts")
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             self.args.manifest_dir / "musan_cuts.jsonl.gz"
         )
 

egs/aishell/ASR/transducer_stateless_modified-2/train.py

@@ -56,7 +56,7 @@ from asr_datamodule import AsrDataModule
 from conformer import Conformer
 from decoder import Decoder
 from joiner import Joiner
-from lhotse import CutSet, load_manifest_lazy
+from lhotse import CutSet, load_manifest
 from lhotse.cut import Cut
 from lhotse.utils import fix_random_seed
 from model import Transducer
@@ -735,7 +735,7 @@ def run(rank, world_size, args):
     train_datatang_cuts = train_datatang_cuts.repeat(times=None)
 
     if args.enable_musan:
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             Path(args.manifest_dir) / "musan_cuts.jsonl.gz"
         )
     else:

egs/alimeeting/ASR/pruned_transducer_stateless2/asr_datamodule.py

@@ -27,6 +27,7 @@ from lhotse import (
     CutSet,
     Fbank,
     FbankConfig,
+    load_manifest,
     load_manifest_lazy,
     set_caching_enabled,
 )
@@ -204,7 +205,7 @@ class AlimeetingAsrDataModule:
             The state dict for the training sampler.
         """
         logging.info("About to get Musan cuts")
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             self.args.manifest_dir / "musan_cuts.jsonl.gz"
         )
 

egs/gigaspeech/ASR/conformer_ctc/asr_datamodule.py

@@ -20,7 +20,7 @@ import logging
 from functools import lru_cache
 from pathlib import Path
 
-from lhotse import CutSet, Fbank, FbankConfig, load_manifest_lazy
+from lhotse import CutSet, Fbank, FbankConfig, load_manifest, load_manifest_lazy
 from lhotse.dataset import (
     CutConcatenate,
     CutMix,
@@ -189,7 +189,7 @@ class GigaSpeechAsrDataModule:
 
     def train_dataloaders(self, cuts_train: CutSet) -> DataLoader:
         logging.info("About to get Musan cuts")
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             self.args.manifest_dir / "musan_cuts.jsonl.gz"
         )
 

egs/gigaspeech/ASR/pruned_transducer_stateless2/asr_datamodule.py

@@ -23,7 +23,7 @@ from pathlib import Path
 from typing import Any, Dict, Optional
 
 import torch
-from lhotse import CutSet, Fbank, FbankConfig, load_manifest_lazy
+from lhotse import CutSet, Fbank, FbankConfig, load_manifest, load_manifest_lazy
 from lhotse.dataset import (
     CutConcatenate,
     CutMix,
@@ -216,7 +216,7 @@ class GigaSpeechAsrDataModule:
         if self.args.enable_musan:
             logging.info("Enable MUSAN")
             logging.info("About to get Musan cuts")
-            cuts_musan = load_manifest_lazy(
+            cuts_musan = load_manifest(
                 self.args.manifest_dir / "musan_cuts.jsonl.gz"
             )
             transforms.append(

egs/librispeech/ASR/pruned_transducer_stateless3/train.py

@@ -66,7 +66,7 @@ from conformer import Conformer
 from decoder import Decoder
 from gigaspeech import GigaSpeech
 from joiner import Joiner
-from lhotse import CutSet, load_manifest_lazy
+from lhotse import CutSet, load_manifest
 from lhotse.cut import Cut
 from lhotse.dataset.sampling.base import CutSampler
 from lhotse.utils import fix_random_seed
@@ -968,7 +968,7 @@ def run(rank, world_size, args):
     train_giga_cuts = train_giga_cuts.repeat(times=None)
 
     if args.enable_musan:
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             Path(args.manifest_dir) / "musan_cuts.jsonl.gz"
         )
     else:

egs/librispeech/ASR/tdnn_lstm_ctc/asr_datamodule.py

@@ -24,7 +24,7 @@ from pathlib import Path
 from typing import Any, Dict, Optional
 
 import torch
-from lhotse import CutSet, Fbank, FbankConfig, load_manifest_lazy
+from lhotse import CutSet, Fbank, FbankConfig, load_manifest, load_manifest_lazy
 from lhotse.dataset import (  # noqa F401 for PrecomputedFeatures
     CutConcatenate,
     CutMix,
@@ -224,7 +224,7 @@ class LibriSpeechAsrDataModule:
         if self.args.enable_musan:
             logging.info("Enable MUSAN")
             logging.info("About to get Musan cuts")
-            cuts_musan = load_manifest_lazy(
+            cuts_musan = load_manifest(
                 self.args.manifest_dir / "musan_cuts.jsonl.gz"
             )
             transforms.append(

egs/librispeech/ASR/transducer_stateless_multi_datasets/test_asr_datamodule.py

@@ -28,7 +28,7 @@ from pathlib import Path
 
 from asr_datamodule import AsrDataModule
 from gigaspeech import GigaSpeech
-from lhotse import load_manifest_lazy
+from lhotse import load_manifest
 from librispeech import LibriSpeech
 
 
@@ -41,7 +41,7 @@ def test_dataset():
     print(args)
 
     if args.enable_musan:
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             Path(args.manifest_dir) / "musan_cuts.jsonl.gz"
         )
     else:

egs/librispeech/ASR/transducer_stateless_multi_datasets/train.py

@@ -73,7 +73,7 @@ from conformer import Conformer
 from decoder import Decoder
 from gigaspeech import GigaSpeech
 from joiner import Joiner
-from lhotse import CutSet, load_manifest_lazy
+from lhotse import CutSet, load_manifest
 from lhotse.cut import Cut
 from lhotse.utils import fix_random_seed
 from librispeech import LibriSpeech
@@ -775,7 +775,7 @@ def run(rank, world_size, args):
     train_giga_cuts = train_giga_cuts.repeat(times=None)
 
     if args.enable_musan:
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             Path(args.manifest_dir) / "musan_cuts.jsonl.gz"
         )
     else:

egs/spgispeech/ASR/pruned_transducer_stateless2/asr_datamodule.py

@@ -22,7 +22,7 @@ from pathlib import Path
 from typing import Any, Dict, Optional
 
 import torch
-from lhotse import CutSet, Fbank, FbankConfig, load_manifest_lazy
+from lhotse import CutSet, Fbank, FbankConfig, load_manifest, load_manifest_lazy
 from lhotse.dataset import (
     CutConcatenate,
     CutMix,
@@ -176,7 +176,7 @@ class SPGISpeechAsrDataModule:
             The state dict for the training sampler.
         """
         logging.info("About to get Musan cuts")
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             self.args.manifest_dir / "cuts_musan.jsonl.gz"
         )
 

egs/tedlium3/ASR/transducer_stateless/asr_datamodule.py

@@ -22,7 +22,7 @@ import logging
 from functools import lru_cache
 from pathlib import Path
 
-from lhotse import CutSet, Fbank, FbankConfig, load_manifest_lazy
+from lhotse import CutSet, Fbank, FbankConfig, load_manifest, load_manifest_lazy
 from lhotse.dataset import (
     CutConcatenate,
     CutMix,
@@ -179,7 +179,7 @@ class TedLiumAsrDataModule:
         transforms = []
         if self.args.enable_musan:
             logging.info("Enable MUSAN")
-            cuts_musan = load_manifest_lazy(
+            cuts_musan = load_manifest(
                 self.args.manifest_dir / "musan_cuts.jsonl.gz"
             )
             transforms.append(

egs/timit/ASR/tdnn_lstm_ctc/asr_datamodule.py

@@ -23,7 +23,7 @@ from functools import lru_cache
 from pathlib import Path
 from typing import List, Union
 
-from lhotse import CutSet, Fbank, FbankConfig, load_manifest_lazy
+from lhotse import CutSet, Fbank, FbankConfig, load_manifest, load_manifest_lazy
 from lhotse.dataset import (
     CutConcatenate,
     CutMix,
@@ -154,7 +154,7 @@ class TimitAsrDataModule(DataModule):
         cuts_train = self.train_cuts()
 
         logging.info("About to get Musan cuts")
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             self.args.feature_dir / "cuts_musan.jsonl.gz"
         )
 

egs/wenetspeech/ASR/pruned_transducer_stateless2/asr_datamodule.py

@@ -27,6 +27,7 @@ from lhotse import (
     CutSet,
     Fbank,
     FbankConfig,
+    load_manifest,
     load_manifest_lazy,
     set_caching_enabled,
 )
@@ -218,7 +219,7 @@ class WenetSpeechAsrDataModule:
             The state dict for the training sampler.
         """
         logging.info("About to get Musan cuts")
-        cuts_musan = load_manifest_lazy(
+        cuts_musan = load_manifest(
             self.args.manifest_dir / "musan_cuts.jsonl.gz"
         )