search

modelscope / FunASR

A Fundamental End-to-End Speech Recognition Toolkit and Open Source SOTA Pretrained Models, Supporting Speech Recognition, Voice Activity Detection, Text Post-processing etc.
https://www.funasr.com
Other
6.45k stars 687 forks source link

是不是没有同时支持热词和时间戳的长音频模型？ #1210

Closed JVfisher closed 9 months ago

JVfisher commented 9 months ago

https://github.com/alibaba-damo-academy/FunASR/tree/dev_gzf_funasr2/examples/industrial_data_pretraining 里面的例子中没有同时支持热词和时间戳的长音频模型

R1ckShi commented 9 months ago

seaco-paraformer现在同时支持时间戳和热词功能，在funasr1.0分支中可以提前体验：https://github.com/alibaba-damo-academy/FunASR/blob/funasr1.0/examples/industrial_data_pretraining/seaco_paraformer/demo.py

buaibieshanghai commented 1 month ago

seaco-paraformer现在同时支持时间戳和热词功能，在funasr1.0分支中可以提前体验：https://github.com/alibaba-damo-academy/FunASR/blob/funasr1.0/examples/industrial_data_pretraining/seaco_paraformer/demo.py

你好，链接失效了，我使用命令：bash run_server.sh --download-model-dir /workspace/models --vad-dir damo/speech_fsmn_vad_zh-cn-16k-common-pytorch --model-dir damo/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch ，时间戳和热词仍然无法同时支持