-
准备根据examples/aishell/paraformer/run.sh文件测试得到CER指标,但是发现代码中似乎都是写的计算WER,对于中文语音一般测试的指标都是CER,请问在中文语音识别中,在数据集上测试得出的WER指标是不是就是等同于CER指标。
![Uploading {62B2EB8B-C27C-4900-9848-806136E880F7}.png…]()
-
## 🐛 Bug
funasr=1.0.14 : 'AutoModel' object has no attribute 'export'.
使用最新main分支源码安装版本,能够正常导出onnx模型。
最近是否会发布一个新版本到pypi仓库。
-
[
![微信图片_20240201155933](https://github.com/alibaba-damo-academy/FunASR/assets/31496638/490cec33-4239-4f65-916e-248bc0ccec49)
](url)
-
我在多卡训练的时候报了这个错误,单卡训练是没问题的
Task related config: error: unrecognized arguments: --local-rank=0
ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 2) local_rank: 0 (pid: 185479) of b…
-
https://github.com/alibaba-damo-academy/FunASR/blob/main/examples/industrial_data_pretraining/paraformer_streaming/demo.py
-
from funasr import AutoModel
# paraformer-zh is a multi-functional asr model
# use vad, punc, spk or not as you need
model = AutoModel(model="paraformer-zh", model_revision="v2.0.4",
…
-
## ❓ Questions and Help
我以iic/speech_paraformer_asr_nat-zh-cn-8k-common-vocab8358-tensorflow1模型为基础,进行finetune训练。训练到model.pt.ep0.10000,此时所有的音频均可推理,并且可以得到比较准确的结果。
但是从model.pt.ep0之后,逐渐开始对于一些包含wifi,KD…
-
Ubuntu 环境,docker 24.0.5,最新2.0.1镜像,识别大型音频(300m以上)之后,内存不会释放,导致多次识别之后内存溢出
juvu updated
4 months ago
-
Hi, all. I'm collecting the required features which will be considered implementing in Jan. 2024. Please let me know your concern and feel free to comment below. Thanks. To make FunCodec better!
-
项目主页上的代码:
```
from funasr import AutoModel
# paraformer-zh is a multi-functional asr model
# use vad, punc, spk or not as you need
model = AutoModel(model="paraformer-zh", model_revision="v2.0.2"…