From fce4e1d1b48f23cd8332e60afce3df8d6209a6a7 Mon Sep 17 00:00:00 2001
From: gaochangfeng <54253717+gaochangfeng@users.noreply.github.com>
Date: 星期四, 11 四月 2024 14:59:22 +0800
Subject: [PATCH] SenseVoice对富文本解码的参数 (#1608)
---
funasr/models/seaco_paraformer/model.py | 26 ++++++++++++++++++--------
1 files changed, 18 insertions(+), 8 deletions(-)
diff --git a/funasr/models/seaco_paraformer/model.py b/funasr/models/seaco_paraformer/model.py
index 5d0f602..21b6aba 100644
--- a/funasr/models/seaco_paraformer/model.py
+++ b/funasr/models/seaco_paraformer/model.py
@@ -117,6 +117,8 @@
text: (Batch, Length)
text_lengths: (Batch,)
"""
+ text_lengths = text_lengths.squeeze()
+ speech_lengths = speech_lengths.squeeze()
assert text_lengths.dim() == 1, text_lengths.shape
# Check that batch_size is unified
assert (
@@ -128,7 +130,7 @@
hotword_pad = kwargs.get("hotword_pad")
hotword_lengths = kwargs.get("hotword_lengths")
- dha_pad = kwargs.get("dha_pad")
+ seaco_label_pad = kwargs.get("seaco_label_pad")
batch_size = speech.shape[0]
# for data-parallel
@@ -148,7 +150,7 @@
ys_lengths,
hotword_pad,
hotword_lengths,
- dha_pad,
+ seaco_label_pad,
)
if self.train_decoder:
loss_att, acc_att = self._calc_att_loss(
@@ -164,7 +166,7 @@
# force_gatherable: to-device and to-tensor if scalar for DataParallel
if self.length_normalized_loss:
- batch_size = (text_lengths + self.predictor_bias).sum().type_as(batch_size)
+ batch_size = (text_lengths + self.predictor_bias).sum()
loss, stats, weight = force_gatherable((loss, stats, batch_size), loss.device)
return loss, stats, weight
@@ -185,13 +187,12 @@
ys_lengths: torch.Tensor,
hotword_pad: torch.Tensor,
hotword_lengths: torch.Tensor,
- dha_pad: torch.Tensor,
+ seaco_label_pad: torch.Tensor,
):
# predictor forward
encoder_out_mask = (~make_pad_mask(encoder_out_lens, maxlen=encoder_out.size(1))[:, None, :]).to(
encoder_out.device)
- pre_acoustic_embeds, _, _, _ = self.predictor(encoder_out, ys_pad, encoder_out_mask,
- ignore_id=self.ignore_id)
+ pre_acoustic_embeds = self.predictor(encoder_out, ys_pad, encoder_out_mask, ignore_id=self.ignore_id)[0]
# decoder forward
decoder_out, _ = self.decoder(encoder_out, encoder_out_lens, pre_acoustic_embeds, ys_lengths, return_hidden=True)
selected = self._hotword_representation(hotword_pad,
@@ -204,7 +205,7 @@
dec_attended, _ = self.seaco_decoder(contextual_info, _contextual_length, decoder_out, ys_lengths)
merged = self._merge(cif_attended, dec_attended)
dha_output = self.hotword_output_layer(merged[:, :-1]) # remove the last token in loss calculation
- loss_att = self.criterion_seaco(dha_output, dha_pad)
+ loss_att = self.criterion_seaco(dha_output, seaco_label_pad)
return loss_att
def _seaco_decode_with_ASF(self,
@@ -430,7 +431,6 @@
return results, meta_data
-
def generate_hotwords_list(self, hotword_list_or_file, tokenizer=None, frontend=None):
def load_seg_dict(seg_dict_file):
seg_dict = {}
@@ -532,3 +532,13 @@
hotword_list = None
return hotword_list
+ def export(
+ self,
+ **kwargs,
+ ):
+ if 'max_seq_len' not in kwargs:
+ kwargs['max_seq_len'] = 512
+ from .export_meta import export_rebuild_model
+ models = export_rebuild_model(model=self, **kwargs)
+ return models
+
--
Gitblit v1.9.1