Merge branch 'multi_ja_en_mls_english_clean' into musan-mls-clean-final

egs/mls_english/ASR/prepare.sh

@@ -40,7 +40,7 @@ log() {
 log "Starting MLS English data preparation"
 
 if [ $stage -le 0 ] && [ $stop_stage -ge 0 ]; then
-	log "Stage 0: Download data"
+log "Stage 0: Download data"
 	# Check if huggingface_hub is installed
 	if ! python -c "import huggingface_hub" &> /dev/null; then
 		log "huggingface_hub Python library not found. Installing it now..."

egs/mls_english/ASR/zipformer/train.py

@@ -1219,7 +1219,6 @@ def run(rank, world_size, args):
     train_cuts = mls_english_corpus.train_cuts()
     # mls_english_corpus.load_dataset(args.dataset_path)
 
-
     if params.start_batch > 0 and checkpoints and "sampler" in checkpoints:
         # We only load the sampler's state dict when it loads a checkpoint
         # saved in the middle of an epoch
@@ -1241,7 +1240,6 @@ def run(rank, world_size, args):
     train_dl = mls_english_corpus.train_dataloaders(
             train_cuts, sampler_state_dict=sampler_state_dict
     )
-
     valid_dl = mls_english_corpus.valid_dataloader()
 
     if not params.print_diagnostics:

egs/multi_ja_en/ASR/zipformer/train.py

@@ -791,7 +791,7 @@ def compute_loss(
     warm_step = params.warm_step
 
     texts = batch["supervisions"]["text"]
-        y = sentencepiece_processor.encode(texts, out_type=int)
+    y = sentencepiece_processor.encode(texts, out_type=int)
     y = k2.RaggedTensor(y)
 
     with torch.set_grad_enabled(is_training):
@@ -1120,7 +1120,7 @@ def run(rank, world_size, args):
 
     # <blk> is defined in local/prepare_lang_char.py
     params.blank_id = sentencepiece_processor.piece_to_id("<blk>")
-    arams.vocab_size = sentencepiece_processor.get_piece_size()
+    params.vocab_size = sentencepiece_processor.get_piece_size()
 
     if not params.use_transducer:
         params.ctc_loss_scale = 1.0
@@ -1178,22 +1178,20 @@ def run(rank, world_size, args):
     if params.inf_check:
         register_inf_check_hooks(model)
 
-    reazonspeech_corpus = ReazonSpeechAsrDataModule(args)
+    multidataset_datamodule = MultiDatasetAsrDataModule(args)
 
     multi_dataset = MultiDataset(args)
 
     train_cuts = multi_dataset.train_cuts()
 
     def remove_short_and_long_utt(c: Cut):
-        # Keep only utterances with duration between 1 second and 30 seconds
-        #
-        # Caution: There is a reason to select 30.0 here. Please see
-        # ../local/display_manifest_statistics.py
+
+        # Keep only utterances greater than 1 second
         #
         # You should use ../local/display_manifest_statistics.py to get
         # an utterance duration distribution for your dataset to select
-        # the threshold
-        if c.duration < 1.0 or c.duration > 30.0:
+        # the threshold as this is dependent on which datasets you choose
+        if c.duration < 1.0:
             logging.warning(
                 f"Exclude cut with ID {c.id} from training. Duration: {c.duration}"
             )
@@ -1244,7 +1242,8 @@ def run(rank, world_size, args):
     )
 
     valid_cuts = multi_dataset.dev_cuts()
-    valid_dl = reazonspeech_corpus.valid_dataloaders(valid_cuts)
+
+    valid_dl = multidataset_datamodule.valid_dataloaders(valid_cuts)
 
     if not params.print_diagnostics:
         scan_pessimistic_batches_for_oom(
@@ -1386,7 +1385,7 @@ def scan_pessimistic_batches_for_oom(
 
 def main():
     parser = get_parser()
-    ReazonSpeechAsrDataModule.add_arguments(parser)
+    MultiDatasetAsrDataModule.add_arguments(parser)
     args = parser.parse_args()
     args.exp_dir = Path(args.exp_dir)