From 4ba1011b42e041ee1d71448eefd7ef2e7bd61bb6 Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期五, 31 三月 2023 15:31:26 +0800
Subject: [PATCH] export
---
funasr/bin/tp_inference.py | 81 +++++++++-------------------------------
1 files changed, 18 insertions(+), 63 deletions(-)
diff --git a/funasr/bin/tp_inference.py b/funasr/bin/tp_inference.py
index a2472cf..6360b17 100644
--- a/funasr/bin/tp_inference.py
+++ b/funasr/bin/tp_inference.py
@@ -28,6 +28,8 @@
from funasr.utils.types import str_or_none
from funasr.models.frontend.wav_frontend import WavFrontend
from funasr.text.token_id_converter import TokenIDConverter
+from funasr.utils.timestamp_tools import ts_prediction_lfr6_standard
+
header_colors = '\033[95m'
end_colors = '\033[0m'
@@ -37,61 +39,6 @@
'audio_fs': 16000,
'model_fs': 16000
}
-
-def time_stamp_lfr6_advance(us_alphas, us_cif_peak, char_list):
- START_END_THRESHOLD = 5
- MAX_TOKEN_DURATION = 12
- TIME_RATE = 10.0 * 6 / 1000 / 3 # 3 times upsampled
- if len(us_cif_peak.shape) == 2:
- alphas, cif_peak = us_alphas[0], us_cif_peak[0] # support inference batch_size=1 only
- else:
- alphas, cif_peak = us_alphas, us_cif_peak
- num_frames = cif_peak.shape[0]
- if char_list[-1] == '</s>':
- char_list = char_list[:-1]
- # char_list = [i for i in text]
- timestamp_list = []
- new_char_list = []
- # for bicif model trained with large data, cif2 actually fires when a character starts
- # so treat the frames between two peaks as the duration of the former token
- fire_place = torch.where(cif_peak>1.0-1e-4)[0].cpu().numpy() - 3.2 # total offset
- num_peak = len(fire_place)
- assert num_peak == len(char_list) + 1 # number of peaks is supposed to be number of tokens + 1
- # begin silence
- if fire_place[0] > START_END_THRESHOLD:
- # char_list.insert(0, '<sil>')
- timestamp_list.append([0.0, fire_place[0]*TIME_RATE])
- new_char_list.append('<sil>')
- # tokens timestamp
- for i in range(len(fire_place)-1):
- new_char_list.append(char_list[i])
- if MAX_TOKEN_DURATION < 0 or fire_place[i+1] - fire_place[i] < MAX_TOKEN_DURATION:
- timestamp_list.append([fire_place[i]*TIME_RATE, fire_place[i+1]*TIME_RATE])
- else:
- # cut the duration to token and sil of the 0-weight frames last long
- _split = fire_place[i] + MAX_TOKEN_DURATION
- timestamp_list.append([fire_place[i]*TIME_RATE, _split*TIME_RATE])
- timestamp_list.append([_split*TIME_RATE, fire_place[i+1]*TIME_RATE])
- new_char_list.append('<sil>')
- # tail token and end silence
- # new_char_list.append(char_list[-1])
- if num_frames - fire_place[-1] > START_END_THRESHOLD:
- _end = (num_frames + fire_place[-1]) * 0.5
- # _end = fire_place[-1]
- timestamp_list[-1][1] = _end*TIME_RATE
- timestamp_list.append([_end*TIME_RATE, num_frames*TIME_RATE])
- new_char_list.append("<sil>")
- else:
- timestamp_list[-1][1] = num_frames*TIME_RATE
- assert len(new_char_list) == len(timestamp_list)
- res_str = ""
- for char, timestamp in zip(new_char_list, timestamp_list):
- res_str += "{} {} {};".format(char, str(timestamp[0]+0.0005)[:5], str(timestamp[1]+0.0005)[:5])
- res = []
- for char, timestamp in zip(new_char_list, timestamp_list):
- if char != '<sil>':
- res.append([int(timestamp[0] * 1000), int(timestamp[1] * 1000)])
- return res_str, res
class SpeechText2Timestamp:
@@ -110,7 +57,7 @@
timestamp_infer_config, timestamp_model_file, device
)
if 'cuda' in device:
- tp_model = tp_model.cuda()
+ tp_model = tp_model.cuda() # force model to cuda
frontend = None
if tp_train_args.frontend is not None:
@@ -169,8 +116,8 @@
enc = enc[0]
# c. Forward Predictor
- _, _, us_alphas, us_cif_peak = self.tp_model.calc_predictor_timestamp(enc, enc_len, text_lengths.to(self.device)+1)
- return us_alphas, us_cif_peak
+ _, _, us_alphas, us_peaks = self.tp_model.calc_predictor_timestamp(enc, enc_len, text_lengths.to(self.device)+1)
+ return us_alphas, us_peaks
def inference(
@@ -299,9 +246,6 @@
inference=True,
)
- finish_count = 0
- file_count = 1
-
tp_result_list = []
for keys, batch in loader:
assert isinstance(batch, dict), type(batch)
@@ -318,10 +262,9 @@
for batch_id in range(_bs):
key = keys[batch_id]
token = speechtext2timestamp.converter.ids2tokens(batch['text'][batch_id])
- ts_str, ts_list = time_stamp_lfr6_advance(us_alphas[batch_id], us_cif_peak[batch_id], token)
+ ts_str, ts_list = ts_prediction_lfr6_standard(us_alphas[batch_id], us_cif_peak[batch_id], token, force_time_shift=-3.0)
logging.warning(ts_str)
item = {'key': key, 'value': ts_str, 'timestamp':ts_list}
- # tp_result_list.append({'text':"".join([i for i in token if i != '<sil>']), 'timestamp': ts_list})
tp_result_list.append(item)
return tp_result_list
@@ -407,6 +350,18 @@
default=1,
help="The batch size for inference",
)
+ group.add_argument(
+ "--seg_dict_file",
+ type=str,
+ default=None,
+ help="The batch size for inference",
+ )
+ group.add_argument(
+ "--split_with_space",
+ type=bool,
+ default=False,
+ help="The batch size for inference",
+ )
return parser
--
Gitblit v1.9.1