From 343a281ca14809153e2ab1df49ca0c5ffdb01abd Mon Sep 17 00:00:00 2001
From: 语帆 <yf352572@alibaba-inc.com>
Date: 星期三, 28 二月 2024 13:56:32 +0800
Subject: [PATCH] test
---
funasr/models/lcbnet/model.py | 12 +++++++++---
1 files changed, 9 insertions(+), 3 deletions(-)
diff --git a/funasr/models/lcbnet/model.py b/funasr/models/lcbnet/model.py
index 9646e1e..d1ebc5c 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
@@ -420,11 +424,13 @@
else:
# extract fbank feats
time1 = time.perf_counter()
+ pdb.set_trace()
audio_sample_list = load_audio_text_image_video(data_in, fs=frontend.fs, audio_fs=kwargs.get("fs", 16000),
data_type=kwargs.get("data_type", "sound"),
tokenizer=tokenizer)
time2 = time.perf_counter()
meta_data["load_data"] = f"{time2 - time1:0.3f}"
+ pdb.set_trace()
speech, speech_lengths = extract_fbank(audio_sample_list, data_type=kwargs.get("data_type", "sound"),
frontend=frontend)
time3 = time.perf_counter()
--
Gitblit v1.9.1