FunAudioLLM/SenseVoice issues and pull requests

#103 - 微调训练模型，训练loss收敛，验证loss不收敛

Issue - State: open - Opened by wuhongsheng 6 months ago - 2 comments
Labels: question

#102 - finetune model size 有2.8G 原版的模型只有1个G

Issue - State: closed - Opened by wuhongsheng 6 months ago - 1 comment
Labels: question

#101 - sensevoice 微调加载数据出错

Issue - State: open - Opened by yangppde 6 months ago - 3 comments
Labels: question

#100 - 微调训练报 cuda 同步超时 2x4090

Issue - State: closed - Opened by wuhongsheng 6 months ago - 1 comment
Labels: question

#99 - 模型支持输出文本对应的时间戳吗

Issue - State: open - Opened by kirayomato 6 months ago - 10 comments
Labels: question

#98 - 运行api报错

Issue - State: open - Opened by EINEZIO 6 months ago
Labels: bug

#97 - 能自定義热词嗎

Issue - State: open - Opened by TCWPanda 6 months ago - 5 comments
Labels: question

#96 - 无法识别wav文件

Issue - State: open - Opened by leslie2046 6 months ago
Labels: bug

#95 - 导出的onnx模型是否存在精度损失？如何解决？

Issue - State: closed - Opened by Etpoem 7 months ago
Labels: question

#94 - 修改哪些位置可以使其在torch2.4.0环境下正常运行

Issue - State: open - Opened by LuWu9 7 months ago
Labels: question

#93 - What languages does SenseVoice recognize?

Issue - State: open - Opened by brainhome 7 months ago
Labels: question

#92 - Is there a way to return the word timestamp of a sentence?

Issue - State: open - Opened by CuiRobert 7 months ago - 3 comments
Labels: question

#91 - AED 能力较差时训练数据不足吗

Issue - State: closed - Opened by wuhongsheng 7 months ago - 13 comments
Labels: question

#90 - 如何不输出【You are using the latest version of funasr-1.1.4】

Issue - State: closed - Opened by CHEFRA 7 months ago - 2 comments
Labels: question

#89 - Question related rtf_avg, time_speech, time_escape ?

Issue - State: open - Opened by gmumar788 7 months ago
Labels: question

#88 - 模块model无法pip下载

Issue - State: open - Opened by CHEFRA 7 months ago
Labels: question

#87 - IndexError: list index out of range

Issue - State: closed - Opened by wuhongsheng 7 months ago
Labels: bug

#86 - SenseVoiceSmall微调是否支撑增加事件/情绪/语言类型

Issue - State: open - Opened by Danyuhui 7 months ago - 4 comments
Labels: question

#85 - 模型能商用吗

Issue - State: open - Opened by KellyZhang2020 7 months ago
Labels: question

#84 - Pronunciation Correction

Issue - State: open - Opened by Drwaish 7 months ago
Labels: question

#83 - 在相关数据集上做benchmark的代码可以放出来吗

Issue - State: open - Opened by wuhongsheng 7 months ago
Labels: question

#82 - 报错OSError: Not found: "C:\Users\逸城\.cache\modelscope\hub\iic\SenseVoiceSmall\chn_jpn_yue_eng_ko_spectok.bpe.model": No such file or directory Error #2

Issue - State: open - Opened by YiCheng996 7 months ago
Labels: question

#81 - SenseVoiceSmall can not recognize English abbreviations in Japanese language speech

Issue - State: open - Opened by jason-ni 7 months ago
Labels: bug

#80 - model.generate的 input支持audio buffer输入吗？还是只支持wav/mp3文件格式

Issue - State: closed - Opened by jacobtang 7 months ago - 1 comment
Labels: question

#79 - sensevoice无法识别语音 FutureWarning: torch.cuda.amp.autocast(args...) is deprecated. Please use torch.amp.autocast('cuda', args...) instead.

Issue - State: closed - Opened by Ikaros-521 7 months ago - 1 comment
Labels: question

#78 - [about ONNX export]

Issue - State: open - Opened by LateLinux 7 months ago - 4 comments
Labels: question

#77 - 参照readme, 使用funasr推理，语音识别不准确

Issue - State: closed - Opened by lizhao-8202 7 months ago - 9 comments
Labels: question

#76 - onnxruntime-gpu inference is more slow then CPU

Issue - State: closed - Opened by aofengdaxia 7 months ago - 2 comments
Labels: bug

#75 - fix api torchaudio.load bug

Pull Request - State: closed - Opened by godxiaoji 7 months ago

#74 - 处理了一些中文 md 中的格式问题

Pull Request - State: closed - Opened by aboutmydreams 7 months ago

#73 - 能否解释一下案例当中的 remote_code 是如何用的

Issue - State: closed - Opened by aboutmydreams 7 months ago - 3 comments
Labels: question

#72 - 指定识别语音为中文，但是经常出现日文和英文，怎么解决

Issue - State: closed - Opened by crystone123 7 months ago - 1 comment
Labels: question

#71 - (new_env) PS D:\dm\TTS\SenseVoice-main> python webui.py Traceback (most recent call last): File "D:\dm\TTS\SenseVoice-main\webui.py", line 11, in <module> import torch File "D:\dm\TTS\SenseVoice-main\new_env\lib\site-packages\torch\init.py", line 148, in <module> raise err OSError: [WinError 126] 找不到指定的模块。 Error loading "D:\dm\TTS\SenseVoice-main\new_env\lib\site-packages\torch\lib\fbgemm.dll" or one of its dependencies. (new_env) PS D:\dm\TTS\SenseVoice-main>

Issue - State: open - Opened by 578691200 7 months ago
Labels: bug

#70 - why GPU inference is more slowly

Issue - State: closed - Opened by aofengdaxia 7 months ago - 1 comment
Labels: question

#69 - Add fastapi server

Pull Request - State: closed - Opened by iflamed 7 months ago

#68 - libtorch推理问题

Issue - State: closed - Opened by breath1124 7 months ago - 8 comments
Labels: question

#67 - How to reproduce the result on ESC50?

Issue - State: closed - Opened by mlxu995 7 months ago - 2 comments
Labels: question

#66 - sensevoice onnx batch推理时异常

Issue - State: closed - Opened by hbk1991 7 months ago - 7 comments
Labels: question

#65 - 运行webui.py，使用麦克风输入时，处理时间增加

Issue - State: closed - Opened by Hunter6324 7 months ago - 1 comment

#64 - update onnx codes

Pull Request - State: closed - Opened by R1ckShi 7 months ago

#64 - update onnx codes

Pull Request - State: closed - Opened by R1ckShi 7 months ago

#63 - Dev sx

Pull Request - State: closed - Opened by R1ckShi 7 months ago

#62 - branch merge

Pull Request - State: closed - Opened by R1ckShi 7 months ago

#61 - 支持维吾尔族语等少数民族语言吗？

Issue - State: closed - Opened by yzfly 7 months ago - 1 comment

#59 - 建议增加docker部署

Issue - State: open - Opened by Hunter6324 7 months ago - 2 comments

#58 - sensevoice微调

Issue - State: closed - Opened by songzhaohui12 7 months ago - 7 comments

#57 - Update in README file

Issue - State: closed - Opened by muthanii 7 months ago - 1 comment

#56 - sensevoice微调

Issue - State: closed - Opened by songzhaohui12 7 months ago - 3 comments

#55 - docs: add Japanese README file

Pull Request - State: closed - Opened by eltociear 7 months ago

#54 - 训练时AttributeError: 'str' object has no attribute 'size'

Issue - State: closed - Opened by badarrrr 7 months ago - 4 comments

#53 - add gradio in requirements

Pull Request - State: closed - Opened by loveQt 7 months ago

#52 - An error occurs in the demo1.py code. Please help.

Issue - State: closed - Opened by CHK-0000 7 months ago - 5 comments

#51 - 效果非常差

Issue - State: closed - Opened by MonolithFoundation 7 months ago - 4 comments

#50 - 无法得到输出的每一句话的时间戳，以及标点符号

Issue - State: closed - Opened by MonolithFoundation 7 months ago - 7 comments

#49 - expected Tensor as element 1 in argument 0, but got NoneType（webui上录制语音进行识别报错）

Issue - State: closed - Opened by badarrrr 7 months ago - 4 comments

#48 - about rich_transcription_postprocess

Pull Request - State: closed - Opened by v3ucn 7 months ago - 2 comments

#47 - SenseVoice 对于日文动漫语音识别得到的结果比 faster-whisper 差很多，完全不能用

Issue - State: closed - Opened by ShenHongFei 7 months ago - 1 comment

#46 - 有办法得到每个词的时间戳吗

Issue - State: closed - Opened by GrahLnn 7 months ago - 2 comments

#45 - 能指定两个语种吗？使用auto会有语种误识别的情况

Issue - State: closed - Opened by rookie0607 7 months ago - 1 comment

#44 - from funasr_onnx import SenseVoiceSmall ImportError: cannot import name 'SenseVoiceSmall' from 'funasr_onnx'

Issue - State: closed - Opened by lucasjinreal 7 months ago

#43 - ImportError: cannot import name 'rich_transcription_postprocess' from 'funasr.utils.postprocess_utils'

Issue - State: closed - Opened by lucasjinreal 7 months ago - 3 comments

#42 - sensevoice-small模型推理速度与报告不符

Issue - State: closed - Opened by yzslry 7 months ago - 2 comments

#41 - SenseVoice可以部署到手机本地运行吗？

Issue - State: open - Opened by qingfusheng 7 months ago - 3 comments

#40 - will release the mode SenseVoiceLarge?

Issue - State: open - Opened by friendmine 7 months ago - 4 comments

#39 - 执行 python .\demo_funasr.py 报错 ImportError: cannot import name 'rich_transcription_postprocess' from 'funasr.utils.postprocess_utils' 是怎么回事

Issue - State: closed - Opened by anstonjie 7 months ago - 11 comments

#38 - 与qwen-audio相比如何

Issue - State: open - Opened by andeyeluguo 7 months ago

#37 - 训练数据json中 source_len是指音频文件总的帧数么？

Issue - State: closed - Opened by zhoufan1978 8 months ago - 1 comment

#36 - with_or_wo_itn: 这个标签分类<|woitn|>和<|withitn|>分别代表什么意思呢？

Issue - State: closed - Opened by zhoufan1978 8 months ago - 1 comment

#35 - There has a bug for exporting the model to onnx and a question for funasr_onnx lib

Issue - State: closed - Opened by cx16528 8 months ago - 2 comments

#34 - Some questions about fine-tuning

Issue - State: closed - Opened by srdfjy 8 months ago - 2 comments

#33 - Does it support streaming inference?

Issue - State: closed - Opened by srdfjy 8 months ago

#32 - SenseVoice与LLM结合

Issue - State: open - Opened by lucasjinreal 8 months ago - 4 comments

#31 - Emotion output EMO_UNKNOWN

Issue - State: closed - Opened by startreker-shzy 8 months ago - 3 comments

#30 - AssertionError: choose a window size 400 that is [2, 0]

Issue - State: closed - Opened by Hy-1990 8 months ago - 12 comments

#29 - SenseVoice-Small支持商用吗

Issue - State: closed - Opened by dawenxi-only 8 months ago - 1 comment

#28 - OutOfMemoryError

Issue - State: closed - Opened by huacilang 8 months ago - 1 comment

#26 - chore: Add Docker configuration for SenseVoice OpenAI server

Pull Request - State: open - Opened by AndersonBY 8 months ago - 12 comments

#25 - webui.py中CosyVoice地址错误

Issue - State: closed - Opened by tiamohf 8 months ago - 1 comment

#24 - 是否能识别语音的停顿，自动加上标点符号

Issue - State: closed - Opened by leoterry-ulrica 8 months ago - 2 comments

#23 - 直接推理和funast推理都报错

Issue - State: closed - Opened by huacilang 8 months ago - 1 comment

#22 - could not parse ModelProto from iic/SenseVoiceSmall/chn_jpn_yue_eng_ko_spectok.bpe.model

Issue - State: closed - Opened by Rasin-wu 8 months ago - 1 comment

#21 - install requirements.txt fail

Issue - State: closed - Opened by fastfading 8 months ago - 5 comments

#20 - [Tips] The version of FunASR should be updated.

Issue - State: closed - Opened by SapphireLab 8 months ago - 1 comment

#19 - SenseVoice-Large有开源计划吗？

Issue - State: open - Opened by rookie0607 8 months ago - 4 comments

#18 - Does SenseVoice support word or sentence level timestamps?

Issue - State: open - Opened by bitiliu 8 months ago - 2 comments

#17 - 使用funasr推理时文字结果中带<|zh|><|NEUTRAL|><|Speech|><|woitn|>这样的标签，有没有办法去除掉？

Issue - State: closed - Opened by Coder0015 8 months ago - 3 comments

#16 - Convert pt to ggml model

Issue - State: open - Opened by xiaozong52 8 months ago - 1 comment

#15 - how to set skip_special_tokens and timestamp level?

Issue - State: open - Opened by MrRace 8 months ago - 8 comments

#14 - finetune时找不到iic/SenseVoiceCTC

Issue - State: closed - Opened by yangsuxia 8 months ago

#13 - Successfully Run On Macos 14.5 With Conda

Issue - State: open - Opened by xianlezheng 8 months ago - 1 comment

#12 - Error records begin to appear after uploading files in webui mode

Issue - State: closed - Opened by xjspace 8 months ago

#11 - 请问情绪识别功能怎么使用，有相应的推理脚本吗

Issue - State: closed - Opened by HeJinLing 8 months ago - 3 comments

#10 - http协议的webui无法接入麦克风（确认插入了麦克风），需要怎么设置浏览器或者改用https？Chrome浏览器

Issue - State: closed - Opened by yuanmaitian 8 months ago - 1 comment

#9 - 支持流式输出吗？

Issue - State: open - Opened by warmshao 8 months ago - 9 comments

#8 - ImportError: cannot import name 'SenseVoiceSmall' from 'funasr_onnx'

Issue - State: closed - Opened by YunzhaoLu 8 months ago - 3 comments

#7 - torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 214.66 GiB.

Issue - State: closed - Opened by jony4 8 months ago - 3 comments

#6 - fix(requirements): Numpy is not available

Pull Request - State: closed - Opened by QIN2DIM 8 months ago

GitHub / FunAudioLLM/SenseVoice issues and pull requests