CHANGELOG.md

Changelog

0.6.2-rc (2024-06-23)

Added

• ✨ add adjuster to webui [01f09b4]
• ✨ stream mode support adjuster [585d2dd]
• ✨ improve xtts_v2 api [fec66c7]
• ✨ improve normalize [d0da37e]
• ✨ improve normalize/spliter [163b649]
• ✨ add loudness equalization [bc8bda7]
• ✨ support `--use_cpu=chattts,enhancer,trainer,all` [23023bc]
• ✨ improve normalizetion.py [1a7c0ed]
• ✨ ignore_useless_warnings [4b9a32e]
• ✨ enhance logger, info => debug [73bc8e7]
• ✨ add playground.stream page [31377b0]
• ✨ tts api support stream #5 [15e0b2c]

Changed

• 🍱 add _p_en [56f1fbf]
• 🍱 update prompt [4f95b31]
• ⚡ Reduce popping sounds [2d0fd68]
• ⚡ improve `apply_character_map` [ea7399f]

Fixed

• 🐛 fix `apply_normalize` missing `sr` [2db6d65]
• 🐛 fix sentence spliter [5d8937c]
• 🐛 fix playground url_join [53e7cbc]
• 🐛 fix generate_audio args [a7a698c]
• 🐛 fix infer func [b0de527]
• 🐛 fix webui logging format [4adc29e]
• 🐛 fix webui speaker_tab missing progress [fafe242]

Miscellaneous

• 📝 添加整合包地址 [26122d4]
• 📝 details `.env` file and cli usage docs [ec3d36f]
• 📝 update changelog [22996e9]
• Windows not yet supported for torch.compile fix [74ac27d]
• fix: replace mispronounced words in TTS [de66e6b]
• feat: support stream mode [3da0f0c]
• optimize: mps audio quality by contiguous scores [1e4d79f]
• 📝 update changelog [ab55c14]

0.6.1 (2024-06-18)

Added

• ✨ add `--preload_models` [73a41e0]
• ✨ add webui progress [778802d]
• ✨ add merger error [51060bc]
• ✨ tts prompt => experimental [d3e6315]
• ✨ add 基本的 speaker finetune ui [5f68f19]
• ✨ add speaker finetune [5ce27ed]
• ✨ add `--ino_half` remove `--half` [5820e57]
• ✨ add webui podcast 默认值 [dd786a8]
• ✨ add webui 分割器配置 [589327b]
• ✨ add `eos` params to all api [79c994f]

Changed

• ⬆️ Bump urllib3 from 2.2.1 to 2.2.2 [097c15b]
• 🎨 run formatter [8c267e1]
• ⚡ Optimize `audio_data_to_segment` #57 [d33809c]
• ⚡ map_location="cpu" [0f58c10]
• ⚡ colab use default GPU [c7938ad]
• ⚡ improve hf calling [2dde612]
• 🍱 add `bob_ft10.pt` [9eee965]
• ⚡ enhance SynthesizeSegments [0bb4dd7]
• 🍱 add `bob_ft10.pt` [bef1b02]
• ♻️ refactor api [671fcc3]
• ⚡ improve xtts_v2 api [206fabc]
• ⚡ train text => just text [e2037e0]
• ⚡ improve TN [a0069ed]

Fixed

• 🐛 fix webui speaker_editor missing `describe` [2a2a36d]
• 💚 Dependabot alerts [f501860]
• 🐛 fix `numpy<2` #50 [e4fea4f]
• 🐛 fix Box() index #49 add testcase [d982e33]
• 🐛 fix Box() index #49 [1788318]
• 🐛 fix `--use_cpu` #47 update conftest [4095b08]
• 🐛 fix `--use_cpu` #47 [221962f]
• 🐛 fix webui speaker args [3b3c331]
• 🐛 fix speaker trainer [52d473f]
• 🐛 兼容 win32 [7ffa37f]
• 🐛 fix google api ssml synthesize #43 [1566f88]

Miscellaneous

• Merge pull request #58 from lenML/dependabot/pip/urllib3-2.2.2 [f259f18]
• 📝 update changelog [b9da7ec]
• 📝 update [8439437]
• 📝 update [ef97206]
• 📝 improve readme.md [7bf3de2]
• 📝 add bug report forms [091cf09]
• 📝 update changelog [3d519ec]
• 📝 update [66963f8]
• 📝 update [b7a63b5]

0.6.0 (2024-06-12)

Added

• ✨ add XTTSv2 api #42 [d1fc63c]
• ✨ google api 支持 enhancer [14fecdb]
• ✨ 修改 podcast 脚本默认 style [98186c2]
• ✨ playground google api [4109adb]
• ✨ 添加 unload api [ed9d61a]
• ✨ support api workers [babdada]
• ✨ add ffmpeg version to webui footer [e9241a1]
• ✨ support use internal ffmpeg [0e02ab0]
• ✨ 增加参数 debug_generate [94e876a]
• ✨ 支持 api 服务与 webui 并存 [4901491]
• ✨ refiner api support normalize [ef665da]
• ✨ add webui 音色编辑器 [fb4c7b3]
• ✨ add localization [c05035d]
• ✨ SSML 支持 enhancer [5c2788e]
• ✨ webui 增加 podcast 工具 tab [b0b169d]
• ✨ 完善 enhancer [205ebeb]

Changed

• ⚡ improve synthesize_audio [759adc2]
• ⚡ reduce enhancer chunk vram usage [3464b42]
• ⚡ 增加默认说话人 [d702ad5]
• 🍱 add `podcast` `podcast_p` style [2b9e5bf]
• 🎨 improve code [317951e]
• 🍱 update banner [dbc293e]
• ⚡ 增强 TN [092c1b9]
• ⚡ enhancer 支持 off_tqdm [94d34d6]
• ⚡ 增加 git env [43d9c65]
• ⚡ 修改 webui 保存文件格式 [2da41c9]

Breaking changes

• 💥 enhancer support --half [fef2ed6]

Removed

• 🔥 clear code [e8a1d7b]

Fixed

• 🐛 fix worker env loader [5b0bf4e]
• 🐛 fix colab default lang missing [d4e5190]
• 🐛 fix "reflection_pad1d" not implemented for 'Half' [536c19b]
• 🐛 fix #33 [76e0b58]
• 🐛 fix localization error [507dbe7]
• 🐛 block main thread #30 [3a7cbde]
• 🐛 fix webui skip no-translate [a8d595e]
• 🐛 fix hf.space force abort [f564536]
• 🐛 fix missing device #25 [07cf6c1]
• 🐛 fix Chat.refiner_prompt() [0839863]
• 🐛 fix --language type check [50d354c]
• 🐛 fix hparams config #22 [61d9809]
• 🐛 fix enhance 下载脚本 [d2e14b0]
• 🐛 fix 'trange' referenced [d1a8dae]
• 🐛 fix ssml to mp3 error `bad sample width` [564b7eb]
• 🐛 fix seed context exit behavior [d4e33c8]
• 🐛 fix colab script [687cc2c]

Miscellaneous

• 🐳 fix docker / 兼容 py 3.9 [ebb096f]
• 🐳 add .dockerignore [57262b8]
• 🧪 add tests [a807640]
• 🌐 fix [b34a0f8]
• 🌐 remove chat limit desc [3f81eca]
• 🧪 add tests [7a54225]
• 🔨 improve model downloader [79a0c59]
• 🌐 更新翻译文案 [f56caa7]

0.5.5 (2024-06-08)

Added

• ✨ add webui speaker creator [df26549]
• ✨ webui speaker tab and merger [7ad71fd]
• ✨ add enhance download script [37adec6]
• ✨ add audio enhance/denoise [00cbc8e]
• ✅ add speakers api test [fbe4304]
• ✅ add unit test [e7f9385]

Changed

• ♿ pin resemble-enhance [b8f41f9]
• ⚡ 调整 speaker 合并 step [906ecc3]
• ♻️ SSML refactor [6666082]
• ♻️ webui refactor [7585282]
• ⚡ add benchmark [ddb7670]

Removed

• 🔥 remove trainer [0c80c24]

Fixed

• 🐛 fix warning and hf.spaces error [f9700bb]
• 🐛 fix model thread competition [0ade6ac]
• 🐛 fix hf space error `ZeroGPU has not been initialized` [562e17c]
• 🐛 fix openai api [49088c5]

Miscellaneous

• 🐳 fix pip requirements [4256371]

0.5.2 (2024-06-06)

Changed

• ⚡ improve TN [6744323]

Fixed

• 🐛 fix window proxy env [d0f9760]

Miscellaneous

• 📝 add banchmark [3a72ba0]

0.5.1 (2024-06-06)

Fixed

• 🐛 fix SynthesizeSegments seed [83b63bd]

0.5.0 (2024-06-06)

Added

• ✨ add systeam versions info [ff94763]
• ✨ torch_gc [d8a8f35]
• ✨ normalize improve [c9db440]
• ✨ improve playground speaker manager [01ebda3]
• ✨ improve speaker manager [1b377d4]
• ✨ add speakers [6c4aa29]
• ✨ playground preact -> react [450a0f9]
• ✨ batch_size in api [616a262]
• ✨ add .env file [a0eddee]
• ✨ improve sentence spliter for markdown [7cac79a]
• ✨ support batch generate [cb9d9aa]
• ✨ add cli args, off_tqdm / half [dff2098]
• ✨ add cli args, no_playground no_docs [98629a2]
• ✨ colab script [5943fd9]
• ✨ emoji normalize [c88fc3f]
• ✨ improve webui [c48c227]
• ✨ add download models script [581f278]
• ✨ add spks [d2a7364]
• ✨ add dockerfile [fc0f4e7]
• ✨ add ssml example / fix ssml [a2c18b1]
• ✨ add styles [e106b1b]
• ✨ webui [3c959ad]
• 🎉 base code all in one [3051588]

Changed

• ⚡ docker [6b8eed1]
• ⚡ playground improve [f8da40b]
• ⚡ improve normalize for EN [3717ae3]
• ⚡ dockerfile [8ad7659]
• ⚡ improve webui [4ac24e8]
• ⚡ improve [153fa4f]
• ⚡ improve [2dbc76d]

Fixed

• 🐛 fix playground package url [cc33013]
• 🐛 fix webui TN [9f1e8f4]
• 🐛 improve rng #11 [af06646]
• 🐛 fix webui segment limit [a74034f]
• 🐛 fix speaker hash [bd3e532]
• 🐛 修复 openai api speed 验证 [124a430]
• 🐛 fix 接口地址调整 + 文档 #9 [3bf349e]
• 🐛 fix docker python env `unsupported operand type(s)` [c2c9658]
• 🐛 fix webui ssml [f2722e3]
• 🐛 fix batch window rolling [63434b3]
• 🐛 fix colab error [61121e9]
• 🐛 find speaker by name [25610b8]
• 🐛 fix dropdown miss [e5eab54]
• 🐛 fix env read [f9cb9d0]
• 🐛 fix env loader [d724659]
• 🐛 fix colab no half [dce20c8]
• 🐛 fix tqdm referenced [c29fd5c]
• 🐛 fix #6 [fc30977]
• 🐛 fix seed context error [faceb2b]
• 🐛 fix infer_seed range [2782182]
• 🐛 fix #2 改用境内 cdn [6ba27c3]
• 🐛 fix #3 [7ffe91f]
• 🐛 fix webui [b44156f]
• 🐛 fix webui [4ad0cee]
• 🐛 适配大写 % [5de4bf7]
• 🐛 webui normalize [d8113f8]
• 🐛 webui spk style fix [3319358]
• 🐛 speaker load [588848d]
• 🐛 speaker load [3f63aa7]
• 🐛 playground base_url [8b468fc]

Miscellaneous

• :zep: improve cache [eedc558]
• :zep: Improved device support [400afe6]
• :zep: improve TN [c22591f]
• add mps support [6983506]
• :zep: revert infer_utils [4a507cc]
• Optimize tqdm display [516eca6]
• Improve code [0e278ab]
• Add apple gpu mps backend [9764e23]
• Update issue templates [1d19417]
• LICENSE [0afb189]
• Initial commit [9fed2b6]