small fix

egs/tal_csasr/ASR/lstm_transducer_stateless3/train.py

@@ -108,6 +108,27 @@ def add_model_arguments(parser: argparse.ArgumentParser):
         default=512,
         help="Encoder output dimesion.",
     )
+    
+    parser.add_argument(
+        "--decoder-dim",
+        type=int,
+        default=512,
+        help="Decoder output dimension.",
+    )
+
+    parser.add_argument(
+        "--joiner-dim",
+        type=int,
+        default=512,
+        help="Joiner output dimension.",
+    )
+
+    parser.add_argument(
+        "--dim-feedforward",
+        type=int,
+        default=2048,
+        help="Dimension of feed forward.",
+    )
 
     parser.add_argument(
         "--rnn-hidden-size",
@@ -402,11 +423,6 @@ def get_params() -> AttributeDict:
             # parameters for conformer
             "feature_dim": 80,
             "subsampling_factor": 4,
-            "dim_feedforward": 2048,
-            # parameters for decoder
-            "decoder_dim": 512,
-            # parameters for joiner
-            "joiner_dim": 512,
             # parameters for Noam
             "model_warm_step": 3000,  # arg given to model, not for lrate
             "env_info": get_env_info(),
@@ -426,6 +442,7 @@ def get_encoder_model(params: AttributeDict) -> nn.Module:
         dim_feedforward=params.dim_feedforward,
         num_encoder_layers=params.num_encoder_layers,
         aux_layer_period=params.aux_layer_period,
+        is_pnnx=params.is_pnnx,
     )
     return encoder
 
@@ -619,6 +636,7 @@ def compute_loss(
     feature_lens = supervisions["num_frames"].to(device)
 
     texts = batch["supervisions"]["text"]
+    #import pdb; pdb.set_trace()
     y = graph_compiler.texts_to_ids_with_bpe(texts)
     if type(y) == list:
         y = k2.RaggedTensor(y).to(device)
@@ -910,8 +928,6 @@ def run(rank, world_size, args):
     """
     params = get_params()
     params.update(vars(args))
-    if params.full_libri is False:
-        params.valid_interval = 800
 
     fix_random_seed(params.seed)
     if world_size > 1:
@@ -930,6 +946,12 @@ def run(rank, world_size, args):
         device = torch.device("cuda", rank)
     logging.info(f"Device: {device}")
 
+    bpe_model = params.lang_dir + "/bpe.model"
+    import sentencepiece as spm
+
+    sp = spm.SentencePieceProcessor()
+    sp.load(bpe_model)
+
     lexicon = Lexicon(params.lang_dir)
     graph_compiler = CharCtcTrainingGraphCompiler(
         lexicon=lexicon,
@@ -1001,33 +1023,7 @@ def run(rank, world_size, args):
         # 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 > 20.0:
-            logging.warning(
-                f"Exclude cut with ID {c.id} from training. Duration: {c.duration}"
-            )
-            return False
-
-        # In pruned RNN-T, we require that T >= S
-        # where T is the number of feature frames after subsampling
-        # and S is the number of tokens in the utterance
-
-        # In ./lstm.py, the conv module uses the following expression
-        # for subsampling
-        T = ((c.num_frames - 3) // 2 - 1) // 2
-        tokens = sp.encode(c.supervisions[0].text, out_type=str)
-
-        if T < len(tokens):
-            logging.warning(
-                f"Exclude cut with ID {c.id} from training. "
-                f"Number of frames (before subsampling): {c.num_frames}. "
-                f"Number of frames (after subsampling): {T}. "
-                f"Text: {c.supervisions[0].text}. "
-                f"Tokens: {tokens}. "
-                f"Number of tokens: {len(tokens)}"
-            )
-            return False
-
-        return True
+        return 1.0 <= c.duration <= 20.0
     
     def text_normalize_for_cut(c: Cut):
         # Text normalize for each sample
@@ -1056,15 +1052,15 @@ def run(rank, world_size, args):
     valid_cuts = valid_cuts.map(text_normalize_for_cut)
     valid_dl = tal_csasr.valid_dataloaders(valid_cuts)
 
-    if not params.print_diagnostics:
-        scan_pessimistic_batches_for_oom(
-            model=model,
-            train_dl=train_dl,
-            optimizer=optimizer,
-            graph_compiler=graph_compiler,
-            params=params,
-            warmup=0.0 if params.start_epoch == 1 else 1.0,
-        )
+    # if not params.print_diagnostics:
+    #     scan_pessimistic_batches_for_oom(
+    #         model=model,
+    #         train_dl=train_dl,
+    #         optimizer=optimizer,
+    #         graph_compiler=graph_compiler,
+    #         params=params,
+    #         warmup=0.0 if params.start_epoch == 1 else 1.0,
+    #     )
 
     scaler = GradScaler(enabled=params.use_fp16)
     if checkpoints and "grad_scaler" in checkpoints: