From d60306e7a435053a1ed626213f9fa6fe12af2b3e Mon Sep 17 00:00:00 2001
From: 语帆 <yf352572@alibaba-inc.com>
Date: 星期五, 23 二月 2024 16:47:15 +0800
Subject: [PATCH] test
---
funasr/models/lcbnet/model.py | 11 ++++++++---
1 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/funasr/models/lcbnet/model.py b/funasr/models/lcbnet/model.py
index 9646e1e..555d4e6 100644
--- a/funasr/models/lcbnet/model.py
+++ b/funasr/models/lcbnet/model.py
@@ -21,7 +21,7 @@
from funasr.utils import postprocess_utils
from funasr.utils.datadir_writer import DatadirWriter
from funasr.register import tables
-
+import pdb
@tables.register("model_classes", "LCBNet")
class LCBNet(nn.Module):
"""
@@ -89,8 +89,8 @@
text_encoder = text_encoder_class(input_size=vocab_size, **text_encoder_conf)
fusion_encoder_class = tables.encoder_classes.get(fusion_encoder)
fusion_encoder = fusion_encoder_class(**fusion_encoder_conf)
- bias_predictor_class = tables.encoder_classes.get_class(bias_predictor)
- bias_predictor = bias_predictor_class(args.bias_predictor_conf)
+ bias_predictor_class = tables.encoder_classes.get(bias_predictor)
+ bias_predictor = bias_predictor_class(**bias_predictor_conf)
if decoder is not None:
decoder_class = tables.decoder_classes.get(decoder)
@@ -117,9 +117,13 @@
self.specaug = specaug
self.normalize = normalize
self.encoder = encoder
+ # lcbnet
self.text_encoder = text_encoder
self.fusion_encoder = fusion_encoder
self.bias_predictor = bias_predictor
+ self.select_num = select_num
+ self.select_length = select_length
+ self.insert_blank = insert_blank
if not hasattr(self.encoder, "interctc_use_conditioning"):
self.encoder.interctc_use_conditioning = False
@@ -409,6 +413,7 @@
logging.info("enable beam_search")
self.init_beam_search(**kwargs)
self.nbest = kwargs.get("nbest", 1)
+ pdb.set_trace()
meta_data = {}
if isinstance(data_in, torch.Tensor) and kwargs.get("data_type", "sound") == "fbank": # fbank
--
Gitblit v1.9.1