speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

docs

Browse Source
This commit is contained in:
游雁 2023-09-18 10:43:27 +08:00
parent b091828cea
commit 74aa12ee4b
5 changed files with 11 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
funasr/bin/asr_inference_launch.py
View File

@ -45,7 +45,7 @@ from funasr.utils.types import str2bool
from funasr.utils.types import str2triple_str
from funasr.utils.types import str_or_none
from funasr.utils.vad_utils import slice_padding_fbank
from tqdm import tqdm
def inference_asr(
maxlenratio: float,
@ -651,7 +651,8 @@ def inference_paraformer_vad_punc(
batch_size_token_ms_cum = 0
beg_idx = 0
for j, _ in enumerate(range(0, n)):
beg_asr_total = time.time()
for j, _ in tqdm(enumerate(range(0, n))):
batch_size_token_ms_cum += (sorted_data[j][0][1] - sorted_data[j][0][0])
if j < n - 1 and (batch_size_token_ms_cum + sorted_data[j + 1][0][1] - sorted_data[j + 1][0][0]) < batch_size_token_ms and (sorted_data[j + 1][0][1] - sorted_data[j + 1][0][0]) < batch_size_token_threshold_s:
continue
@ -661,16 +662,17 @@ def inference_paraformer_vad_punc(
beg_idx = end_idx
batch = {"speech": speech_j, "speech_lengths": speech_lengths_j}
batch = to_device(batch, device=device)
print("batch: ", speech_j.shape[0])
# print("batch: ", speech_j.shape[0])
beg_asr = time.time()
results = speech2text(**batch)
end_asr = time.time()
print("time cost asr: ", end_asr - beg_asr)
# print("time cost asr: ", end_asr - beg_asr)
if len(results) < 1:
results = [["", [], [], [], [], [], []]]
results_sorted.extend(results)
end_asr_total = time.time()
print("total time cost asr: ", end_asr_total-beg_asr_total)
restored_data = [0] * n
for j in range(n):
index = sorted_data[j][1]

2
funasr/quick_start.md
View File

@ -26,7 +26,7 @@ python funasr_wss_server.py --port 10095
python funasr_wss_client.py --host "127.0.0.1" --port 10095 --mode 2pass --chunk_size "5,10,5"
```
For more examples, please refer to [docs](https://alibaba-damo-academy.github.io/FunASR/en/runtime/websocket_python.html#id2).
For more examples, please refer to [docs](runtime/python/websocket/README.md).
### C++ version Example

2
funasr/quick_start_zh.md
View File

@ -26,7 +26,7 @@ python funasr_wss_server.py --port 10095
python funasr_wss_client.py --host "127.0.0.1" --port 10095 --mode 2pass --chunk_size "5,10,5"
#python funasr_wss_client.py --host "127.0.0.1" --port 10095 --mode 2pass --chunk_size "8,8,4" --audio_in "./data/wav.scp"
```
更多例子可以参考（[点击此处](runtime/websocket_python.html#id2)）
更多例子可以参考（[点击此处](runtime/python/websocket/README.md)）
<a name="cpp版本示例"></a>
#### c++版本示例

2
funasr/version.txt
View File

@ -1 +1 @@
0.7.6
0.7.7

1
setup.py
View File

@ -39,6 +39,7 @@ requirements = {
"edit-distance",
"textgrid",
"protobuf",
"tqdm",
],
# train: The modules invoked when training only.
"train": [