fishaudio/fish-speech issues and pull requests

#540 - Context length reading configuration file

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

#539 - replace "sox" backend for reference_audio

Pull Request - State: closed - Opened by jmoney7823956789378 about 2 months ago - 1 comment

#538 - Fix BUG pytorch dependency resolver

Pull Request - State: closed - Opened by yorelog about 2 months ago - 1 comment

#537 - The expanded size of the tensor (472) must match the existing size (1023) at non-singleton dimension 1.

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

#535 - ModuleNotFoundError: No module named 'ormsgpack'

Issue - State: closed - Opened by one-pip about 2 months ago - 5 comments
Labels: bug

#534 - Fix breakdown infer

Pull Request - State: closed - Opened by AnyaCoder about 2 months ago

#533 - Tensor Size Mismatch and `setup_caches` Assertion Error in `generate.py`

Issue - State: closed - Opened by ppmzhang2 about 2 months ago - 3 comments
Labels: bug

#532 - Mac support?

Issue - State: closed - Opened by Pareak about 2 months ago - 1 comment
Labels: enhancement

#531 - License terms: what about using this in a commercial product?

Issue - State: open - Opened by juangea about 2 months ago - 10 comments
Labels: enhancement

#530 - Fix infer warmup

Pull Request - State: closed - Opened by AnyaCoder about 2 months ago

#529 - docs: add Japanese README

Pull Request - State: closed - Opened by eltociear about 2 months ago - 1 comment

#526 - Crash when launching new model

Issue - State: closed - Opened by ApatheticWrath about 2 months ago - 2 comments
Labels: bug

#525 - ModuleNotFoundError: No module named 'ormsgpack'

Issue - State: closed - Opened by one-pip about 2 months ago
Labels: bug

#524 - Update docs etc.

Pull Request - State: closed - Opened by AnyaCoder about 2 months ago

#523 - Upload V1.4 Demo Url

Pull Request - State: closed - Opened by PoTaTo-Mika about 2 months ago

#521 - Error occured while finetuning fish-speech-1.4: index 5 is out of bounds for dimension 1 with size 5

Issue - State: closed - Opened by wjddd about 2 months ago - 2 comments
Labels: bug

#520 - Update to 1.4

Pull Request - State: closed - Opened by leng-yue 2 months ago

#518 - fully support ormsgpack

Pull Request - State: closed - Opened by AnyaCoder 2 months ago

#516 - Update issue template

Pull Request - State: closed - Opened by Stardust-minus 2 months ago

#515 - Before you contribute, please read our 'Language Policy'.

Issue - State: open - Opened by Stardust-minus 2 months ago
Labels: enhancement

#514 - [BUG]The results of running the same model on different computers vary greatly.

Issue - State: open - Opened by tao1261060556 2 months ago - 2 comments
Labels: bug

#513 - keep up with official close-source api

Pull Request - State: closed - Opened by AnyaCoder 2 months ago

#512 - Why fish-speech implementation of transformer feedforward contains there linears?

Issue - State: closed - Opened by JohnHerry 2 months ago - 2 comments

#511 - [BUG]ASGI error occurred during WebUI operation

Issue - State: closed - Opened by ievenight 2 months ago - 3 comments
Labels: bug

#510 - [Feature] Request language support - add dog language clone

Issue - State: closed - Opened by lujiale621 2 months ago - 1 comment
Labels: enhancement

#509 - [BUG]The audio generated by the streaming request is noisy. How can I solve this problem?

Issue - State: open - Opened by ArvinC 2 months ago - 3 comments
Labels: bug

#508 - [Feature] The current streaming TTS seems to generate speech in the order of requests. Are there any plans to infer multiple requests simultaneously in the future?

Issue - State: closed - Opened by world1tree 2 months ago - 2 comments
Labels: enhancement

#507 - Fix reference to load_filelist

Pull Request - State: closed - Opened by ppmzhang2 2 months ago

#505 - [BUG] Unable to open WebUI

Issue - State: closed - Opened by ejiandan 2 months ago - 4 comments
Labels: bug

#504 - [BUG]Unable to continue last training session

Issue - State: closed - Opened by tao1261060556 2 months ago - 6 comments
Labels: bug

#503 - [BUG] centos, environment installed successfully, various dependency errors

Issue - State: closed - Opened by wukongbuku 2 months ago - 3 comments
Labels: bug

#502 - [BUG] I cannot install fish-speech. Is it related to the pip version?

Issue - State: closed - Opened by wukongbuku 2 months ago - 6 comments
Labels: bug

#501 - [BUG] model re-compiled each time, time consuming.

Issue - State: closed - Opened by didadida-r 3 months ago - 4 comments
Labels: bug

#500 - [BUG]TypeError: cannot unpack non-iterable NoneType object

Issue - State: closed - Opened by XJF2332 3 months ago - 1 comment
Labels: bug

#499 - Avoid cuda-dependent code for CPU-only inference

Pull Request - State: closed - Opened by dur-randir 3 months ago

#498 - Update README.md, add trending

Pull Request - State: closed - Opened by Stardust-minus 3 months ago

#497 - [Feature]How can I specify that characters are silent and that pinyin or English words are read automatically?

Issue - State: open - Opened by xeoshow 3 months ago - 4 comments
Labels: enhancement

#496 - [Bug] toos/api.py reports an error when half inference is enabled

Issue - State: closed - Opened by litao-zhx 3 months ago - 8 comments
Labels: bug

#495 - [Feature] Improved API interface, no need to provide reference audio and text

Issue - State: closed - Opened by fastfading 3 months ago - 6 comments
Labels: enhancement

#494 - [Feature] Can the online API consider submitting to create the model?

Issue - State: closed - Opened by 08090813 3 months ago - 4 comments
Labels: enhancement

#493 - After AutoDL is deployed, the 4090 graphics card uses GPU to infer the length of 1,000 Chinese characters, which takes 37 seconds. Is the audio length of 3 minutes and 10 seconds normal?

Issue - State: closed - Opened by wosuiyu 3 months ago - 2 comments
Labels: enhancement

#492 - Which will be better? with indices or with hiddens?

Issue - State: closed - Opened by JohnHerry 3 months ago - 1 comment

#491 - [BUG] Package name error

Issue - State: closed - Opened by shinoairisu 3 months ago - 2 comments
Labels: bug

#490 - [BUG] 按照linux方法安装报错

Issue - State: closed - Opened by shinoairisu 3 months ago - 10 comments
Labels: bug

#489 - [Feature] How to support new languages

Issue - State: closed - Opened by ILG2021 3 months ago - 1 comment
Labels: enhancement

#488 - [BUG] API接口调用报错，都是500错误 Internal Server Error

Issue - State: closed - Opened by zane8521 3 months ago - 1 comment
Labels: bug

#487 - Fix Import Path in tools/vqgan/inference.py

Pull Request - State: closed - Opened by octree 3 months ago

#486 - Fix(deps): remove audio-seperator

Pull Request - State: closed - Opened by Stardust-minus 3 months ago

#485 - [BUG]The API call does not seem to support wav and pcm.

Issue - State: closed - Opened by cqnimin 3 months ago - 3 comments
Labels: bug

#484 - [Feature]Tesla P40这类FP16性能稍差的显卡如何使用FP32精度进行推理？

Issue - State: closed - Opened by SakuraRK 3 months ago
Labels: enhancement

#483 - llama 训练速度

Issue - State: closed - Opened by dukGuo 3 months ago - 1 comment
Labels: enhancement

#482 - From whisper to sensevoice

Pull Request - State: closed - Opened by AnyaCoder 3 months ago

#481 - [BUG] Website reCaptcha error without reCaptcha

Issue - State: closed - Opened by MInecrafterPictures 3 months ago - 4 comments
Labels: bug

#480 - 【疑问】语音情感是否支持？

Issue - State: closed - Opened by l-dawei 3 months ago - 1 comment

#479 - [BUG] Normal LoRa finetune: Error(s) in loading state_dict for TextToSemantic

Issue - State: closed - Opened by SinanAkkoyun 3 months ago - 2 comments
Labels: bug

#478 - Fix win deps

Pull Request - State: closed - Opened by AnyaCoder 3 months ago

#477 - more stable streaming audio play and save

Pull Request - State: closed - Opened by Jalen-Zhong 3 months ago

#476 - [Feature] Save streaming audio to local path

Issue - State: closed - Opened by Jalen-Zhong 3 months ago - 1 comment
Labels: enhancement

#475 - [BUG] api推理不支持mp3格式的参考音频

Issue - State: closed - Opened by Jalen-Zhong 3 months ago - 2 comments
Labels: bug

#474 - [help wanted] faster whisper model的作用

Issue - State: closed - Opened by Jalen-Zhong 3 months ago - 6 comments

#473 - 后半段丟字严重， llm 过早预测了结束符

Issue - State: closed - Opened by kunyao2015 3 months ago - 3 comments
Labels: bug

#472 - [Bug]After fine-tuning, the sentence will end only after the first sentence is correct and there are many wrong words afterwards.

Issue - State: open - Opened by bobo-wmdigit 3 months ago - 4 comments
Labels: bug

#471 - [BUG] Incapable of speaking abbreviations

Issue - State: closed - Opened by SinanAkkoyun 3 months ago - 1 comment
Labels: bug

#470 - What are the improvements of fish-speech compared to TorToise-TTS or CosyVoice

Issue - State: closed - Opened by JohnHerry 3 months ago - 3 comments

#469 - 关于开源权重问题

Issue - State: closed - Opened by Jalen-Zhong 3 months ago - 4 comments
Labels: bug

#468 - 开启API时提示错误

Issue - State: closed - Opened by japssss019 3 months ago - 6 comments
Labels: enhancement

#467 - chore: retranslate ja/index.md

Pull Request - State: closed - Opened by Tps-F 3 months ago

#466 - [QUESTION]服务部署后，使用nginx进行反向代理，流式请求的结果播放都是杂音

Issue - State: closed - Opened by GGMGG 3 months ago - 1 comment
Labels: bug

#465 - [BUG] 速度很慢，加了compile后15tokens/s，一句话要十多秒

Issue - State: closed - Opened by Songjc0511 3 months ago - 1 comment
Labels: bug

#463 - [BUG]

Issue - State: closed - Opened by 222qwgjsb 3 months ago - 1 comment
Labels: bug

#462 - ，

Issue - State: closed - Opened by 222qwgjsb 3 months ago - 1 comment
Labels: enhancement

#461 - Support inference on mps device natively

Pull Request - State: closed - Opened by Tps-F 3 months ago - 9 comments

#460 - [BUG] --compile 加载之后推理更慢了，tokens 个位数

Issue - State: closed - Opened by josangmi 3 months ago - 13 comments
Labels: bug

#459 - [BUG] Error 在打开推理页面时

Issue - State: closed - Opened by Jason-JP-Yang 3 months ago - 2 comments
Labels: bug

#458 - [BUG] Nonetype when using --compile

Issue - State: closed - Opened by ifounda-bug 3 months ago - 4 comments
Labels: bug

#457 - [BUG] How many epoch should we finetune with lora?

Issue - State: closed - Opened by OswaldoBornemann 3 months ago - 3 comments
Labels: bug, stale

#456 - chore: speck

Pull Request - State: closed - Opened by hscspring 3 months ago - 1 comment

#455 - Website audio generation is giving constant Interval Server Error

Issue - State: closed - Opened by csjaugustus 3 months ago - 1 comment
Labels: bug

#454 - 训练数据集的speaker id命名问题

Issue - State: closed - Opened by ssccutyy 3 months ago - 1 comment
Labels: enhancement

#453 - Docs: Update docker configuration

Pull Request - State: closed - Opened by TsingShui 3 months ago - 1 comment

#452 - Training data for the pretrained model?

Issue - State: closed - Opened by awaisjafar 3 months ago
Labels: bug

#451 - [BUG] There are problems with the generated audio content when the number of words is small.

Issue - State: closed - Opened by Itsuka-Xx 3 months ago - 7 comments
Labels: bug

#450 - [Feature] GPU 支持情况 (more GPU support)

Issue - State: closed - Opened by secext2022 3 months ago - 3 comments
Labels: enhancement

#449 - 在微调llama前的准备数据集时，对.lab文件有什么要求？

Issue - State: closed - Opened by hizening 3 months ago - 1 comment
Labels: enhancement

#448 - 儿童音色去哪找？可以直接导入使用的那种

Issue - State: closed - Opened by asdfgnini 3 months ago

#447 - [pre-commit.ci] pre-commit autoupdate

Pull Request - State: closed - Opened by pre-commit-ci[bot] 3 months ago

#446 - [Benchmark] 希望能做一期跟同类项目的效果对比

Issue - State: closed - Opened by Pevernow 3 months ago
Labels: enhancement

#445 - [BUG]官网的API页面，充值了但是一直看不到额度，刷新页面提示：Internal Server Error

Issue - State: closed - Opened by YoungWiller 3 months ago - 1 comment
Labels: bug

#444 - [BUG] 3080 TI and it's still slow even with --compile

Issue - State: closed - Opened by ftk789 3 months ago - 13 comments
Labels: bug

#443 - Fix:(req) pinned torch version to 2.3.1, avoid inference speed abnorm…

Pull Request - State: closed - Opened by Stardust-minus 3 months ago

#442 - [Feature]想了解下推理速度与显存消耗

Issue - State: closed - Opened by Ming-Zhou0201 3 months ago - 3 comments
Labels: enhancement

#441 - No module named 'torch.nn.attention'这个是torch版本问题吗？该如何解决？[BUG]

Issue - State: closed - Opened by HAS-1314 3 months ago - 5 comments
Labels: bug

#440 - Use webui and enable model compilation, the progress bar will stuck for a few seconds before start every times[BUG]

Issue - State: closed - Opened by randoms 3 months ago - 7 comments
Labels: bug

#439 - [Bug] After deployment is complete, the 4090 graphics card uses GPU inference for 2 minutes and the audio duration is 40 seconds. Is this normal?

Issue - State: closed - Opened by 2776721779 3 months ago - 10 comments
Labels: bug, help wanted

#438 - [BUG] Some situations encountered after enabling multiple workers

Issue - State: closed - Opened by josangmi 3 months ago - 3 comments
Labels: bug

#437 - [Feature] Training on 24 kHz

Issue - State: closed - Opened by NoSavedDATA 3 months ago - 1 comment
Labels: enhancement

#436 - 从文本生成语义 token阶段加了--compile参数会出错，找不到原因

Issue - State: closed - Opened by CasonTsai 3 months ago - 5 comments

#435 - Fix：(req) add torchaudio req

Pull Request - State: closed - Opened by Stardust-minus 3 months ago

#433 - Add preprocessing and validation steps when loading the checkpoint.

Pull Request - State: closed - Opened by bfs18 3 months ago

#432 - windows install_env.bat安装后启动失败[BUG]

Issue - State: closed - Opened by Xiaodu233 3 months ago - 2 comments
Labels: bug

GitHub / fishaudio/fish-speech issues and pull requests