Spaces:
Running
on
Zero
Running
on
Zero
zhzluke96
commited on
Commit
•
b473486
1
Parent(s):
0cb7b72
update
Browse files- CHANGELOG.md +172 -0
- modules/normalization.py +4 -0
- modules/repos_static/resemble_enhance/inference.py +14 -3
- modules/utils/zh_normalization/num.py +33 -14
- modules/webui/app.py +6 -1
- modules/webui/changelog_tab.py +13 -0
CHANGELOG.md
ADDED
@@ -0,0 +1,172 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Changelog
|
2 |
+
|
3 |
+
<a name="0.5.6-rc"></a>
|
4 |
+
## 0.5.6-rc (2024-06-09)
|
5 |
+
|
6 |
+
### Added
|
7 |
+
|
8 |
+
- ✨ 完善 enhancer [[205ebeb](https://github.com/lenML/ChatTTS-Forge/commit/205ebebeb7530c81fde7ea96c7e4c6a888a29835)]
|
9 |
+
|
10 |
+
### Changed
|
11 |
+
|
12 |
+
- ⚡ 增强 TN [[092c1b9](https://github.com/lenML/ChatTTS-Forge/commit/092c1b94147249880198fe2ad3dfe3b209099e19)]
|
13 |
+
- ⚡ enhancer 支持 off_tqdm [[94d34d6](https://github.com/lenML/ChatTTS-Forge/commit/94d34d657fa3433dae9ff61775e0c364a6f77aff)]
|
14 |
+
- ⚡ 增加 git env [[43d9c65](https://github.com/lenML/ChatTTS-Forge/commit/43d9c65877ff68ad94716bc2e505ccc7ae8869a8)]
|
15 |
+
- ⚡ 修改webui保存文件格式 [[2da41c9](https://github.com/lenML/ChatTTS-Forge/commit/2da41c90aa81bf87403598aefaea3e0ae2e83d79)]
|
16 |
+
|
17 |
+
### Fixed
|
18 |
+
|
19 |
+
- 🐛 fix seed context exit behavior [[d4e33c8](https://github.com/lenML/ChatTTS-Forge/commit/d4e33c8f0aabe253ce96756f907e979578c81b17)]
|
20 |
+
- 🐛 fix colab script [[687cc2c](https://github.com/lenML/ChatTTS-Forge/commit/687cc2cc97ff7e89328b747dbfcacbcd51bd5efc)]
|
21 |
+
|
22 |
+
### Miscellaneous
|
23 |
+
|
24 |
+
- 📝 update webui document [[7f2bb22](https://github.com/lenML/ChatTTS-Forge/commit/7f2bb227027cc0eff312c37758a20916c1ebade6)]
|
25 |
+
|
26 |
+
|
27 |
+
<a name="0.5.5"></a>
|
28 |
+
|
29 |
+
## 0.5.5 (2024-06-08)
|
30 |
+
|
31 |
+
### Added
|
32 |
+
|
33 |
+
- ✨ add webui speaker creator [[df26549](https://github.com/lenML/ChatTTS-Forge/commit/df265490f35b2b991c395455dd2f4ad563193cef)]
|
34 |
+
- ✨ webui speaker tab and merger [[7ad71fd](https://github.com/lenML/ChatTTS-Forge/commit/7ad71fddb61f3b41b3af66d201f6105ca09539d9)]
|
35 |
+
- ✨ add enhance download script [[37adec6](https://github.com/lenML/ChatTTS-Forge/commit/37adec6de3109b3829602c7c7be06fd7247f10eb)]
|
36 |
+
- ✨ add audio enhance/denoise [[00cbc8e](https://github.com/lenML/ChatTTS-Forge/commit/00cbc8e96833fbcaf6cc224dc330908fa647f317)]
|
37 |
+
- ✅ add speakers api test [[fbe4304](https://github.com/lenML/ChatTTS-Forge/commit/fbe4304c6716fb182442d356dbe3976982ca9d2b)]
|
38 |
+
- ✅ add unit test [[e7f9385](https://github.com/lenML/ChatTTS-Forge/commit/e7f938562c1173899cc4e7330d59a8e354cafea4)]
|
39 |
+
|
40 |
+
### Changed
|
41 |
+
|
42 |
+
- ♿ pin resemble-enhance [[b8f41f9](https://github.com/lenML/ChatTTS-Forge/commit/b8f41f90061c75ee3e09ddc6cae8d657bc67aad1)]
|
43 |
+
- ⚡ 调整 speaker 合并 step [[906ecc3](https://github.com/lenML/ChatTTS-Forge/commit/906ecc3d295d90459485cd131563ffd588914d52)]
|
44 |
+
- ♻️ SSML refactor [[6666082](https://github.com/lenML/ChatTTS-Forge/commit/6666082375c43b143d242bf425053e2ae661eb09)]
|
45 |
+
- ♻️ webui refactor [[7585282](https://github.com/lenML/ChatTTS-Forge/commit/75852822f7d9cd8b95b557e2870e5435a0932fa1)]
|
46 |
+
- ⚡ add benchmark [[ddb7670](https://github.com/lenML/ChatTTS-Forge/commit/ddb76704e5e6847bb0eeca2c0b50764a66783686)]
|
47 |
+
|
48 |
+
### Removed
|
49 |
+
|
50 |
+
- 🔥 remove trainer [[0c80c24](https://github.com/lenML/ChatTTS-Forge/commit/0c80c2437fb8e8b231ae770205089198f4ac1c13)]
|
51 |
+
|
52 |
+
### Fixed
|
53 |
+
|
54 |
+
- 🐛 fix warning and hf.spaces error [[f9700bb](https://github.com/lenML/ChatTTS-Forge/commit/f9700bbb1b057b2dfe4437de7cbd41a659be76c5)]
|
55 |
+
- 🐛 fix model thread competition [[0ade6ac](https://github.com/lenML/ChatTTS-Forge/commit/0ade6ac07a2c75eb1cdda1c3db8bdf9bc2665244)]
|
56 |
+
- 🐛 fix hf space error `ZeroGPU has not been initialized` [[562e17c](https://github.com/lenML/ChatTTS-Forge/commit/562e17c9372278c03705ec5a3ec77750854d5c7e)]
|
57 |
+
- 🐛 fix openai api [[49088c5](https://github.com/lenML/ChatTTS-Forge/commit/49088c5480043518bb9beda817f5e5b38d133fa8)]
|
58 |
+
|
59 |
+
### Miscellaneous
|
60 |
+
|
61 |
+
- 🐳 fix pip requirements [[4256371](https://github.com/lenML/ChatTTS-Forge/commit/4256371c9d3d8d290840a98fb6ac7bc19268a1e7)]
|
62 |
+
|
63 |
+
<a name="0.5.2"></a>
|
64 |
+
|
65 |
+
## 0.5.2 (2024-06-06)
|
66 |
+
|
67 |
+
### Changed
|
68 |
+
|
69 |
+
- ⚡ improve TN [[6744323](https://github.com/lenML/ChatTTS-Forge/commit/6744323df814430b2d92c3f16329ab8f09eb4ad3)]
|
70 |
+
|
71 |
+
### Fixed
|
72 |
+
|
73 |
+
- 🐛 fix window proxy env [[d0f9760](https://github.com/lenML/ChatTTS-Forge/commit/d0f97608cef2afdbeb803c906a71e05dbf2424a1)]
|
74 |
+
|
75 |
+
### Miscellaneous
|
76 |
+
|
77 |
+
- 📝 add banchmark [[3a72ba0](https://github.com/lenML/ChatTTS-Forge/commit/3a72ba0f97d5409502b9ff98e356f69affcce06b)]
|
78 |
+
|
79 |
+
<a name="0.5.1"></a>
|
80 |
+
|
81 |
+
## 0.5.1 (2024-06-06)
|
82 |
+
|
83 |
+
### Fixed
|
84 |
+
|
85 |
+
- 🐛 fix SynthesizeSegments seed [[83b63bd](https://github.com/lenML/ChatTTS-Forge/commit/83b63bdd0d92e115c9b6946f427343c48de1a313)]
|
86 |
+
|
87 |
+
<a name="0.5.0"></a>
|
88 |
+
|
89 |
+
## 0.5.0 (2024-06-06)
|
90 |
+
|
91 |
+
### Added
|
92 |
+
|
93 |
+
- ✨ add systeam versions info [[ff94763](https://github.com/lenML/ChatTTS-Forge/commit/ff947636c5e69d6bdf5111f95d8afb979d157fba)]
|
94 |
+
- ✨ torch_gc [[d8a8f35](https://github.com/lenML/ChatTTS-Forge/commit/d8a8f35958c25d931ce47b53730d388a71e86b2d)]
|
95 |
+
- ✨ normalize improve [[c9db440](https://github.com/lenML/ChatTTS-Forge/commit/c9db440b2719119285c6536c2c4658afdb20ff27)]
|
96 |
+
- ✨ improve playground speaker manager [[01ebda3](https://github.com/lenML/ChatTTS-Forge/commit/01ebda3a28cfc5f1e78f8a434b1077f01b22f399)]
|
97 |
+
- ✨ improve speaker manager [[1b377d4](https://github.com/lenML/ChatTTS-Forge/commit/1b377d448214e232d477bd828fca5eba6aa87e7b)]
|
98 |
+
- ✨ add speakers [[6c4aa29](https://github.com/lenML/ChatTTS-Forge/commit/6c4aa29f147942d1c2f3c095b4d832409e53e5cb)]
|
99 |
+
- ✨ playground preact -> react [[450a0f9](https://github.com/lenML/ChatTTS-Forge/commit/450a0f9d184c8c5d8df6cce3d8e4596c543dbfe1)]
|
100 |
+
- ✨ batch_size in api [[616a262](https://github.com/lenML/ChatTTS-Forge/commit/616a262012d1e23dd877bbfc4cde4f16f477d1a7)]
|
101 |
+
- ✨ add .env file [[a0eddee](https://github.com/lenML/ChatTTS-Forge/commit/a0eddeefc141630d43496881e73d5fcd90742828)]
|
102 |
+
- ✨ improve sentence spliter for markdown [[7cac79a](https://github.com/lenML/ChatTTS-Forge/commit/7cac79ad6720c98bdec0903dd473de70bdcef137)]
|
103 |
+
- ✨ support batch generate [[cb9d9aa](https://github.com/lenML/ChatTTS-Forge/commit/cb9d9aa5d55482a51018a4ebe95d84d95d803cc4)]
|
104 |
+
- ✨ add cli args, off_tqdm / half [[dff2098](https://github.com/lenML/ChatTTS-Forge/commit/dff2098b76deac30d738ce05ba99225914b781a3)]
|
105 |
+
- ✨ add cli args, no_playground no_docs [[98629a2](https://github.com/lenML/ChatTTS-Forge/commit/98629a2c46a376a3955ea8862c9b80361f44f8a3)]
|
106 |
+
- ✨ colab script [[5943fd9](https://github.com/lenML/ChatTTS-Forge/commit/5943fd9c168ad98c0d2caf6deffb1d63848d75c3)]
|
107 |
+
- ✨ emoji normalize [[c88fc3f](https://github.com/lenML/ChatTTS-Forge/commit/c88fc3f18f0ae512f477880cfea972376f825ab7)]
|
108 |
+
- ✨ improve webui [[c48c227](https://github.com/lenML/ChatTTS-Forge/commit/c48c2278d8a123694ecf9deeb36ea4d5e3a0499e)]
|
109 |
+
- ✨ add download models script [[581f278](https://github.com/lenML/ChatTTS-Forge/commit/581f27859dad4a6c3d1b4af9ebc854fc2f4829df)]
|
110 |
+
- ✨ add spks [[d2a7364](https://github.com/lenML/ChatTTS-Forge/commit/d2a736485907490e1dbf0db334ec2b6459d35dbe)]
|
111 |
+
- ✨ add dockerfile [[fc0f4e7](https://github.com/lenML/ChatTTS-Forge/commit/fc0f4e73f7f8e1698afd97065a84161f7807e655)]
|
112 |
+
- ✨ add ssml example / fix ssml [[a2c18b1](https://github.com/lenML/ChatTTS-Forge/commit/a2c18b149584137d323d14f0923a6c2f98969e89)]
|
113 |
+
- ✨ add styles [[e106b1b](https://github.com/lenML/ChatTTS-Forge/commit/e106b1bc3edffd71c8da8408e57444b50c0cc91a)]
|
114 |
+
- ✨ webui [[3c959ad](https://github.com/lenML/ChatTTS-Forge/commit/3c959ad341775e335233e6dd6797bca045ec106e)]
|
115 |
+
- 🎉 base code all in one [[3051588](https://github.com/lenML/ChatTTS-Forge/commit/30515881b31a5bb7b08be273f4cdb3c9a9854a6c)]
|
116 |
+
|
117 |
+
### Changed
|
118 |
+
|
119 |
+
- ⚡ docker [[6b8eed1](https://github.com/lenML/ChatTTS-Forge/commit/6b8eed1caa69af9787a019fb226d25a155ca58af)]
|
120 |
+
- ⚡ playground improve [[f8da40b](https://github.com/lenML/ChatTTS-Forge/commit/f8da40b773d4485f69dfd207fc33294dfac61b93)]
|
121 |
+
- ⚡ improve normalize for EN [[3717ae3](https://github.com/lenML/ChatTTS-Forge/commit/3717ae31478f561f9dd83fad8983282e7f9380c9)]
|
122 |
+
- ⚡ dockerfile [[8ad7659](https://github.com/lenML/ChatTTS-Forge/commit/8ad7659f62af5436545da239ee8f8f7cb3d6c103)]
|
123 |
+
- ⚡ improve webui [[4ac24e8](https://github.com/lenML/ChatTTS-Forge/commit/4ac24e835ca078ca09f81e41980757a84cc4fccd)]
|
124 |
+
- ⚡ improve [[153fa4f](https://github.com/lenML/ChatTTS-Forge/commit/153fa4f152b3ed084cb2b5e6f0b12d86f40f77f3)]
|
125 |
+
- ⚡ improve [[2dbc76d](https://github.com/lenML/ChatTTS-Forge/commit/2dbc76ddc9937d2df63106f77197fa28a5b0d23b)]
|
126 |
+
|
127 |
+
### Fixed
|
128 |
+
|
129 |
+
- 🐛 fix playground package url [[cc33013](https://github.com/lenML/ChatTTS-Forge/commit/cc33013a5a0a4c7f21011e44dbc0fb29228482fa)]
|
130 |
+
- 🐛 fix webui TN [[9f1e8f4](https://github.com/lenML/ChatTTS-Forge/commit/9f1e8f45dce8f519526896df7033fbc40690abeb)]
|
131 |
+
- 🐛 improve rng [#11](https://github.com/lenML/ChatTTS-Forge/issues/11) [[af06646](https://github.com/lenML/ChatTTS-Forge/commit/af06646877f56ea458a8b316dca4160d546bb4b5)]
|
132 |
+
- 🐛 fix webui segment limit [[a74034f](https://github.com/lenML/ChatTTS-Forge/commit/a74034f0fb86a997d621605967ceca59351f2627)]
|
133 |
+
- 🐛 fix speaker hash [[bd3e532](https://github.com/lenML/ChatTTS-Forge/commit/bd3e5324cf260ce75997a2b5d087af7d0c0106fe)]
|
134 |
+
- 🐛 修复 openai api speed 验证 [[124a430](https://github.com/lenML/ChatTTS-Forge/commit/124a4309e7be0e146379e175eb9cde6374ea207d)]
|
135 |
+
- 🐛 fix 接口地址调整 + 文档 [#9](https://github.com/lenML/ChatTTS-Forge/issues/9) [[3bf349e](https://github.com/lenML/ChatTTS-Forge/commit/3bf349e274d8ab343688d2d76bd59edd29c742ed)]
|
136 |
+
- 🐛 fix docker python env `unsupported operand type(s)` [[c2c9658](https://github.com/lenML/ChatTTS-Forge/commit/c2c965890334736ae0887c214fd731426239c3e8)]
|
137 |
+
- 🐛 fix webui ssml [[f2722e3](https://github.com/lenML/ChatTTS-Forge/commit/f2722e38a00cfd03cb44589d30148ecea6703cd5)]
|
138 |
+
- 🐛 fix batch window rolling [[63434b3](https://github.com/lenML/ChatTTS-Forge/commit/63434b34ab35d77dd91e6d6c244477459a8bf44f)]
|
139 |
+
- 🐛 fix colab error [[61121e9](https://github.com/lenML/ChatTTS-Forge/commit/61121e9c7150d64d5a3112e9e266eecc764c3e63)]
|
140 |
+
- 🐛 find speaker by name [[25610b8](https://github.com/lenML/ChatTTS-Forge/commit/25610b877c12bf68fd16064489c7d24354a1d900)]
|
141 |
+
- 🐛 fix dropdown miss [[e5eab54](https://github.com/lenML/ChatTTS-Forge/commit/e5eab54454069a4f7575ceee29a36960b916e015)]
|
142 |
+
- 🐛 fix env read [[f9cb9d0](https://github.com/lenML/ChatTTS-Forge/commit/f9cb9d099bd022635fe078155598411fdd7df00e)]
|
143 |
+
- 🐛 fix env loader [[d724659](https://github.com/lenML/ChatTTS-Forge/commit/d724659389197637f152b09079a36f6b1f26e79b)]
|
144 |
+
- 🐛 fix colab no half [[dce20c8](https://github.com/lenML/ChatTTS-Forge/commit/dce20c8e1592cce2871953a42f7ebfa7b0f0f54e)]
|
145 |
+
- 🐛 fix tqdm referenced [[c29fd5c](https://github.com/lenML/ChatTTS-Forge/commit/c29fd5ca82923fa1be0de70137f4ce0e2e32f4ae)]
|
146 |
+
- 🐛 fix [#6](https://github.com/lenML/ChatTTS-Forge/issues/6) [[fc30977](https://github.com/lenML/ChatTTS-Forge/commit/fc309774a73c3c9ae4c5c527c3aa9bef4a7a3c1f)]
|
147 |
+
- 🐛 fix seed context error [[faceb2b](https://github.com/lenML/ChatTTS-Forge/commit/faceb2b0af749a869df3656dee89b20f69b130b8)]
|
148 |
+
- 🐛 fix infer_seed range [[2782182](https://github.com/lenML/ChatTTS-Forge/commit/2782182367faef932429e2a7a012f7d867c0cb3a)]
|
149 |
+
- 🐛 fix [#2](https://github.com/lenML/ChatTTS-Forge/issues/2) 改用境内 cdn [[6ba27c3](https://github.com/lenML/ChatTTS-Forge/commit/6ba27c3c4d1ef4175cfe95de671599f64572eaeb)]
|
150 |
+
- 🐛 fix [#3](https://github.com/lenML/ChatTTS-Forge/issues/3) [[7ffe91f](https://github.com/lenML/ChatTTS-Forge/commit/7ffe91fa3028f88f0414786dd669b91cce409043)]
|
151 |
+
- 🐛 fix webui [[b44156f](https://github.com/lenML/ChatTTS-Forge/commit/b44156fbac7ea59765ced4bbfd303f3095166693)]
|
152 |
+
- 🐛 fix webui [[4ad0cee](https://github.com/lenML/ChatTTS-Forge/commit/4ad0ceef9d1251437cff57e12a0a82f1cd427e17)]
|
153 |
+
- 🐛 适配大写 % [[5de4bf7](https://github.com/lenML/ChatTTS-Forge/commit/5de4bf7e19d436c93bc584af859fc5e58596f0c9)]
|
154 |
+
- 🐛 webui normalize [[d8113f8](https://github.com/lenML/ChatTTS-Forge/commit/d8113f8c336d4a1b8e8fb0e82d0ba2ed02a36086)]
|
155 |
+
- 🐛 webui spk style fix [[3319358](https://github.com/lenML/ChatTTS-Forge/commit/3319358a0c51930cca6cc6c9326a694aff15c8c4)]
|
156 |
+
- 🐛 speaker load [[588848d](https://github.com/lenML/ChatTTS-Forge/commit/588848d30fa4e3fc669ced81d780517145be3e2e)]
|
157 |
+
- 🐛 speaker load [[3f63aa7](https://github.com/lenML/ChatTTS-Forge/commit/3f63aa76911274ddf62c47298d1758b0fd984e52)]
|
158 |
+
- 🐛 playground base_url [[8b468fc](https://github.com/lenML/ChatTTS-Forge/commit/8b468fc67e6c4b649844a419b8dda77bf01ea439)]
|
159 |
+
|
160 |
+
### Miscellaneous
|
161 |
+
|
162 |
+
- :zep: improve cache [[eedc558](https://github.com/lenML/ChatTTS-Forge/commit/eedc55880f4802ef9f23117d8266100be18cbe41)]
|
163 |
+
- :zep: Improved device support [[400afe6](https://github.com/lenML/ChatTTS-Forge/commit/400afe60c9261247998617b2943fa210230b7ad3)]
|
164 |
+
- :zep: improve TN [[c22591f](https://github.com/lenML/ChatTTS-Forge/commit/c22591ffb2c1ae9e14eef307d708925bcc7a67a7)]
|
165 |
+
- add mps support [[6983506](https://github.com/lenML/ChatTTS-Forge/commit/6983506d7237c05ea346fdbe7d042b69d3743a3b)]
|
166 |
+
- :zep: revert infer_utils [[4a507cc](https://github.com/lenML/ChatTTS-Forge/commit/4a507cccf23b94a11f470b67d96bbad6c9efbf74)]
|
167 |
+
- Optimize tqdm display [[516eca6](https://github.com/lenML/ChatTTS-Forge/commit/516eca6c99b7b8433d1cda8435a19214eb1cc678)]
|
168 |
+
- Improve code [[0e278ab](https://github.com/lenML/ChatTTS-Forge/commit/0e278ab40276e3e0610515bf0e4329c95a67e00c)]
|
169 |
+
- Add apple gpu mps backend [[9764e23](https://github.com/lenML/ChatTTS-Forge/commit/9764e233e1e82d8dbec607032744cb565b616525)]
|
170 |
+
- Update issue templates [[1d19417](https://github.com/lenML/ChatTTS-Forge/commit/1d194177ab6dd3b1055ca925bfd2befc6fe2f07c)]
|
171 |
+
- LICENSE [[0afb189](https://github.com/lenML/ChatTTS-Forge/commit/0afb189718797e0706ba64c1e8d2188c4fa0fe4c)]
|
172 |
+
- Initial commit [[9fed2b6](https://github.com/lenML/ChatTTS-Forge/commit/9fed2b60a90547286e6c06483167c397c7bbed78)]
|
modules/normalization.py
CHANGED
@@ -310,6 +310,10 @@ State-of-the-art Machine Learning for PyTorch, TensorFlow, and JAX.
|
|
310 |
🗣️ Audio: automatic speech recognition and audio classification.
|
311 |
🐙 Multimodal: table question answering, optical character recognition, information extraction from scanned documents, video classification, and visual question answering.
|
312 |
""",
|
|
|
|
|
|
|
|
|
313 |
]
|
314 |
|
315 |
for i, test_case in enumerate(test_cases):
|
|
|
310 |
🗣️ Audio: automatic speech recognition and audio classification.
|
311 |
🐙 Multimodal: table question answering, optical character recognition, information extraction from scanned documents, video classification, and visual question answering.
|
312 |
""",
|
313 |
+
"""
|
314 |
+
120米
|
315 |
+
有12%的概率会下雨
|
316 |
+
""",
|
317 |
]
|
318 |
|
319 |
for i, test_case in enumerate(test_cases):
|
modules/repos_static/resemble_enhance/inference.py
CHANGED
@@ -10,6 +10,8 @@ from tqdm import trange
|
|
10 |
|
11 |
from .hparams import HParams
|
12 |
|
|
|
|
|
13 |
logger = logging.getLogger(__name__)
|
14 |
|
15 |
|
@@ -120,7 +122,12 @@ def remove_weight_norm_recursively(module):
|
|
120 |
pass
|
121 |
|
122 |
|
123 |
-
def inference(
|
|
|
|
|
|
|
|
|
|
|
124 |
remove_weight_norm_recursively(model)
|
125 |
|
126 |
hp: HParams = model.hp
|
@@ -150,7 +157,9 @@ def inference(model, dwav, sr, device, chunk_seconds: float = 30.0, overlap_seco
|
|
150 |
|
151 |
chunks = []
|
152 |
for start in trange(0, dwav.shape[-1], hop_length):
|
153 |
-
chunks.append(
|
|
|
|
|
154 |
|
155 |
hwav = merge_chunks(chunks, chunk_length, hop_length, sr=sr, length=dwav.shape[-1])
|
156 |
|
@@ -158,6 +167,8 @@ def inference(model, dwav, sr, device, chunk_seconds: float = 30.0, overlap_seco
|
|
158 |
torch.cuda.synchronize()
|
159 |
|
160 |
elapsed_time = time.perf_counter() - start_time
|
161 |
-
logger.info(
|
|
|
|
|
162 |
|
163 |
return hwav, sr
|
|
|
10 |
|
11 |
from .hparams import HParams
|
12 |
|
13 |
+
from modules import config
|
14 |
+
|
15 |
logger = logging.getLogger(__name__)
|
16 |
|
17 |
|
|
|
122 |
pass
|
123 |
|
124 |
|
125 |
+
def inference(
|
126 |
+
model, dwav, sr, device, chunk_seconds: float = 30.0, overlap_seconds: float = 1.0
|
127 |
+
):
|
128 |
+
if config.runtime_env_vars.off_tqdm:
|
129 |
+
trange = range
|
130 |
+
|
131 |
remove_weight_norm_recursively(model)
|
132 |
|
133 |
hp: HParams = model.hp
|
|
|
157 |
|
158 |
chunks = []
|
159 |
for start in trange(0, dwav.shape[-1], hop_length):
|
160 |
+
chunks.append(
|
161 |
+
inference_chunk(model, dwav[start : start + chunk_length], sr, device)
|
162 |
+
)
|
163 |
|
164 |
hwav = merge_chunks(chunks, chunk_length, hop_length, sr=sr, length=dwav.shape[-1])
|
165 |
|
|
|
167 |
torch.cuda.synchronize()
|
168 |
|
169 |
elapsed_time = time.perf_counter() - start_time
|
170 |
+
logger.info(
|
171 |
+
f"Elapsed time: {elapsed_time:.3f} s, {hwav.shape[-1] / elapsed_time / 1000:.3f} kHz"
|
172 |
+
)
|
173 |
|
174 |
return hwav, sr
|
modules/utils/zh_normalization/num.py
CHANGED
@@ -46,10 +46,15 @@ def replace_frac(match) -> str:
|
|
46 |
sign = match.group(1)
|
47 |
nominator = match.group(2)
|
48 |
denominator = match.group(3)
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
|
|
|
|
|
|
|
|
|
|
53 |
return result
|
54 |
|
55 |
|
@@ -66,9 +71,14 @@ def replace_percentage(match) -> str:
|
|
66 |
"""
|
67 |
sign = match.group(1)
|
68 |
percent = match.group(2)
|
69 |
-
|
70 |
-
|
71 |
-
|
|
|
|
|
|
|
|
|
|
|
72 |
return result
|
73 |
|
74 |
|
@@ -86,9 +96,14 @@ def replace_negative_num(match) -> str:
|
|
86 |
"""
|
87 |
sign = match.group(1)
|
88 |
number = match.group(2)
|
89 |
-
|
90 |
-
|
91 |
-
|
|
|
|
|
|
|
|
|
|
|
92 |
return result
|
93 |
|
94 |
|
@@ -105,7 +120,12 @@ def replace_default_num(match):
|
|
105 |
str
|
106 |
"""
|
107 |
number = match.group(0)
|
108 |
-
|
|
|
|
|
|
|
|
|
|
|
109 |
|
110 |
|
111 |
# 数字表达式
|
@@ -144,9 +164,10 @@ def replace_number(match) -> str:
|
|
144 |
sign = match.group(1)
|
145 |
number = match.group(2)
|
146 |
pure_decimal = match.group(5)
|
147 |
-
|
148 |
# TODO 也许可以把 num2str 完全替换成 cn2an
|
149 |
import cn2an
|
|
|
150 |
text = pure_decimal if pure_decimal else f"{sign}{number}"
|
151 |
try:
|
152 |
result = cn2an.an2cn(text, "low")
|
@@ -158,8 +179,6 @@ def replace_number(match) -> str:
|
|
158 |
number: str = num2str(number)
|
159 |
result = f"{sign}{number}"
|
160 |
return result
|
161 |
-
|
162 |
-
|
163 |
|
164 |
|
165 |
# 范围表达式
|
|
|
46 |
sign = match.group(1)
|
47 |
nominator = match.group(2)
|
48 |
denominator = match.group(3)
|
49 |
+
try:
|
50 |
+
import cn2an
|
51 |
+
|
52 |
+
return cn2an.an2cn(f"{sign}{nominator}/{denominator}", "low")
|
53 |
+
except:
|
54 |
+
sign: str = "负" if sign else ""
|
55 |
+
nominator: str = num2str(nominator)
|
56 |
+
denominator: str = num2str(denominator)
|
57 |
+
result = f"{sign}{denominator}分之{nominator}"
|
58 |
return result
|
59 |
|
60 |
|
|
|
71 |
"""
|
72 |
sign = match.group(1)
|
73 |
percent = match.group(2)
|
74 |
+
try:
|
75 |
+
import cn2an
|
76 |
+
|
77 |
+
return cn2an.an2cn(f"{sign}{percent}%", "low")
|
78 |
+
except:
|
79 |
+
sign: str = "负" if sign else ""
|
80 |
+
percent: str = num2str(percent)
|
81 |
+
result = f"{sign}百分之{percent}"
|
82 |
return result
|
83 |
|
84 |
|
|
|
96 |
"""
|
97 |
sign = match.group(1)
|
98 |
number = match.group(2)
|
99 |
+
try:
|
100 |
+
import cn2an
|
101 |
+
|
102 |
+
return cn2an.an2cn(f"{sign}{number}", "low")
|
103 |
+
except:
|
104 |
+
sign: str = "负" if sign else ""
|
105 |
+
number: str = num2str(number)
|
106 |
+
result = f"{sign}{number}"
|
107 |
return result
|
108 |
|
109 |
|
|
|
120 |
str
|
121 |
"""
|
122 |
number = match.group(0)
|
123 |
+
try:
|
124 |
+
import cn2an
|
125 |
+
|
126 |
+
return cn2an.an2cn(number, "low")
|
127 |
+
except:
|
128 |
+
return verbalize_digit(number, alt_one=True)
|
129 |
|
130 |
|
131 |
# 数字表达式
|
|
|
164 |
sign = match.group(1)
|
165 |
number = match.group(2)
|
166 |
pure_decimal = match.group(5)
|
167 |
+
|
168 |
# TODO 也许可以把 num2str 完全替换成 cn2an
|
169 |
import cn2an
|
170 |
+
|
171 |
text = pure_decimal if pure_decimal else f"{sign}{number}"
|
172 |
try:
|
173 |
result = cn2an.an2cn(text, "low")
|
|
|
179 |
number: str = num2str(number)
|
180 |
result = f"{sign}{number}"
|
181 |
return result
|
|
|
|
|
182 |
|
183 |
|
184 |
# 范围表达式
|
modules/webui/app.py
CHANGED
@@ -7,6 +7,7 @@ import gradio as gr
|
|
7 |
from modules import config
|
8 |
from modules.webui import webui_config
|
9 |
|
|
|
10 |
from modules.webui.system_tab import create_system_tab
|
11 |
from modules.webui.tts_tab import create_tts_interface
|
12 |
from modules.webui.ssml_tab import create_ssml_interface
|
@@ -101,7 +102,11 @@ def create_interface():
|
|
101 |
create_system_tab()
|
102 |
|
103 |
with gr.TabItem("README"):
|
104 |
-
|
|
|
|
|
|
|
|
|
105 |
|
106 |
create_app_footer()
|
107 |
return demo
|
|
|
7 |
from modules import config
|
8 |
from modules.webui import webui_config
|
9 |
|
10 |
+
from modules.webui.changelog_tab import create_changelog_tab
|
11 |
from modules.webui.system_tab import create_system_tab
|
12 |
from modules.webui.tts_tab import create_tts_interface
|
13 |
from modules.webui.ssml_tab import create_ssml_interface
|
|
|
102 |
create_system_tab()
|
103 |
|
104 |
with gr.TabItem("README"):
|
105 |
+
with gr.Tabs():
|
106 |
+
with gr.TabItem("readme"):
|
107 |
+
create_readme_tab()
|
108 |
+
with gr.TabItem("changelog"):
|
109 |
+
create_changelog_tab()
|
110 |
|
111 |
create_app_footer()
|
112 |
return demo
|
modules/webui/changelog_tab.py
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
|
3 |
+
|
4 |
+
def read_local_changelog():
|
5 |
+
with open("CHANGELOG.md", "r", encoding="utf-8") as file:
|
6 |
+
content = file.read()
|
7 |
+
content = content[content.index("# ") :]
|
8 |
+
return content
|
9 |
+
|
10 |
+
|
11 |
+
def create_changelog_tab():
|
12 |
+
changelog_content = read_local_changelog()
|
13 |
+
gr.Markdown(changelog_content)
|