phoonnx 0.0.0__py3-none-any.whl

phoonnx-0.0.0.dist-info/METADATA

@@ -0,0 +1,255 @@
+Metadata-Version: 2.1
+Name: phoonnx
+Version: 0.0.0
+Home-page: https://github.com/TigreGotico/phoonnx
+Author: JarbasAi
+Author-email: jarbasai@mailfence.com
+Requires-Dist: numpy
+Requires-Dist: onnxruntime
+Requires-Dist: quebra-frases
+Requires-Dist: langcodes
+Requires-Dist: ovos-number-parser>=0.3.3a1
+Requires-Dist: ovos-date-parser>=0.6.4a1
+Provides-Extra: aa
+Requires-Dist: epitran; extra == "aa"
+Provides-Extra: aii
+Requires-Dist: epitran; extra == "aii"
+Provides-Extra: am
+Requires-Dist: epitran; extra == "am"
+Provides-Extra: ar
+Requires-Dist: gruut[ar]<3.0,>=2.3.0; extra == "ar"
+Requires-Dist: epitran; extra == "ar"
+Provides-Extra: av
+Requires-Dist: epitran; extra == "av"
+Provides-Extra: az
+Requires-Dist: epitran; extra == "az"
+Provides-Extra: bn
+Requires-Dist: epitran; extra == "bn"
+Provides-Extra: ca
+Requires-Dist: gruut[ca]<3.0,>=2.3.0; extra == "ca"
+Requires-Dist: epitran; extra == "ca"
+Provides-Extra: ceb
+Requires-Dist: epitran; extra == "ceb"
+Provides-Extra: cjy
+Requires-Dist: epitran; extra == "cjy"
+Provides-Extra: ckb
+Requires-Dist: epitran; extra == "ckb"
+Provides-Extra: cs
+Requires-Dist: gruut[cs]<3.0,>=2.3.0; extra == "cs"
+Requires-Dist: epitran; extra == "cs"
+Provides-Extra: csb
+Requires-Dist: epitran; extra == "csb"
+Provides-Extra: de
+Requires-Dist: gruut[de]<3.0,>=2.3.0; extra == "de"
+Requires-Dist: epitran; extra == "de"
+Provides-Extra: en
+Requires-Dist: epitran; extra == "en"
+Requires-Dist: gruut[en]<3.0,>=2.3.0; extra == "en"
+Requires-Dist: misaki[en]; extra == "en"
+Provides-Extra: eo
+Requires-Dist: epitran; extra == "eo"
+Provides-Extra: es
+Requires-Dist: gruut[es]<3.0,>=2.3.0; extra == "es"
+Requires-Dist: epitran; extra == "es"
+Provides-Extra: et
+Requires-Dist: epitran; extra == "et"
+Provides-Extra: eu
+Requires-Dist: epitran; extra == "eu"
+Provides-Extra: fa
+Requires-Dist: persian-phonemizer; extra == "fa"
+Requires-Dist: gruut[fa]<3.0,>=2.3.0; extra == "fa"
+Requires-Dist: epitran; extra == "fa"
+Provides-Extra: ff
+Requires-Dist: epitran; extra == "ff"
+Provides-Extra: fi
+Requires-Dist: epitran; extra == "fi"
+Provides-Extra: fil
+Requires-Dist: epitran; extra == "fil"
+Provides-Extra: fr
+Requires-Dist: gruut[fr]<3.0,>=2.3.0; extra == "fr"
+Requires-Dist: epitran; extra == "fr"
+Provides-Extra: gan
+Requires-Dist: epitran; extra == "gan"
+Provides-Extra: gl
+Requires-Dist: epitran; extra == "gl"
+Provides-Extra: got
+Requires-Dist: epitran; extra == "got"
+Provides-Extra: ha
+Requires-Dist: epitran; extra == "ha"
+Provides-Extra: hak
+Requires-Dist: epitran; extra == "hak"
+Provides-Extra: he
+Requires-Dist: phonikud-onnx; extra == "he"
+Requires-Dist: phonikud; extra == "he"
+Provides-Extra: hi
+Requires-Dist: epitran; extra == "hi"
+Provides-Extra: hmn
+Requires-Dist: epitran; extra == "hmn"
+Provides-Extra: hr
+Requires-Dist: epitran; extra == "hr"
+Provides-Extra: hsn
+Requires-Dist: epitran; extra == "hsn"
+Provides-Extra: hu
+Requires-Dist: epitran; extra == "hu"
+Provides-Extra: id
+Requires-Dist: epitran; extra == "id"
+Provides-Extra: ilo
+Requires-Dist: epitran; extra == "ilo"
+Provides-Extra: it
+Requires-Dist: gruut[it]<3.0,>=2.3.0; extra == "it"
+Requires-Dist: epitran; extra == "it"
+Provides-Extra: ja
+Requires-Dist: pyopenjtalk; extra == "ja"
+Requires-Dist: pykakasi; extra == "ja"
+Requires-Dist: cutlet; extra == "ja"
+Requires-Dist: unidic-lite; extra == "ja"
+Requires-Dist: misaki[ja]; extra == "ja"
+Provides-Extra: jam
+Requires-Dist: epitran; extra == "jam"
+Provides-Extra: jv
+Requires-Dist: epitran; extra == "jv"
+Provides-Extra: ka
+Requires-Dist: epitran; extra == "ka"
+Provides-Extra: kbd
+Requires-Dist: epitran; extra == "kbd"
+Provides-Extra: kk
+Requires-Dist: epitran; extra == "kk"
+Provides-Extra: km
+Requires-Dist: epitran; extra == "km"
+Provides-Extra: ko
+Requires-Dist: g2pk; extra == "ko"
+Requires-Dist: epitran; extra == "ko"
+Requires-Dist: misaki[ko]; extra == "ko"
+Provides-Extra: ku
+Requires-Dist: epitran; extra == "ku"
+Provides-Extra: ky
+Requires-Dist: epitran; extra == "ky"
+Provides-Extra: lb
+Requires-Dist: gruut[lb]<3.0,>=2.3.0; extra == "lb"
+Provides-Extra: lez
+Requires-Dist: epitran; extra == "lez"
+Provides-Extra: lij
+Requires-Dist: epitran; extra == "lij"
+Provides-Extra: lo
+Requires-Dist: epitran; extra == "lo"
+Provides-Extra: lsm
+Requires-Dist: epitran; extra == "lsm"
+Provides-Extra: lt
+Requires-Dist: epitran; extra == "lt"
+Provides-Extra: luy
+Requires-Dist: epitran; extra == "luy"
+Provides-Extra: lv
+Requires-Dist: epitran; extra == "lv"
+Provides-Extra: mi
+Requires-Dist: epitran; extra == "mi"
+Provides-Extra: ml
+Requires-Dist: epitran; extra == "ml"
+Provides-Extra: mr
+Requires-Dist: epitran; extra == "mr"
+Provides-Extra: ms
+Requires-Dist: epitran; extra == "ms"
+Provides-Extra: mt
+Requires-Dist: epitran; extra == "mt"
+Provides-Extra: my
+Requires-Dist: epitran; extra == "my"
+Provides-Extra: nan
+Requires-Dist: epitran; extra == "nan"
+Provides-Extra: nl
+Requires-Dist: gruut[nl]<3.0,>=2.3.0; extra == "nl"
+Provides-Extra: ny
+Requires-Dist: epitran; extra == "ny"
+Provides-Extra: om
+Requires-Dist: epitran; extra == "om"
+Provides-Extra: or
+Requires-Dist: epitran; extra == "or"
+Provides-Extra: pa
+Requires-Dist: epitran; extra == "pa"
+Provides-Extra: pl
+Requires-Dist: epitran; extra == "pl"
+Provides-Extra: pt
+Requires-Dist: gruut[pt]<3.0,>=2.3.0; extra == "pt"
+Requires-Dist: epitran; extra == "pt"
+Provides-Extra: quy
+Requires-Dist: epitran; extra == "quy"
+Provides-Extra: rn
+Requires-Dist: epitran; extra == "rn"
+Provides-Extra: ro
+Requires-Dist: epitran; extra == "ro"
+Provides-Extra: ru
+Requires-Dist: gruut[ru]<3.0,>=2.3.0; extra == "ru"
+Requires-Dist: epitran; extra == "ru"
+Provides-Extra: rw
+Requires-Dist: epitran; extra == "rw"
+Provides-Extra: sg
+Requires-Dist: epitran; extra == "sg"
+Provides-Extra: si
+Requires-Dist: epitran; extra == "si"
+Provides-Extra: sn
+Requires-Dist: epitran; extra == "sn"
+Provides-Extra: so
+Requires-Dist: epitran; extra == "so"
+Provides-Extra: sq
+Requires-Dist: epitran; extra == "sq"
+Provides-Extra: sr
+Requires-Dist: epitran; extra == "sr"
+Provides-Extra: sv
+Requires-Dist: gruut[sv]<3.0,>=2.3.0; extra == "sv"
+Requires-Dist: epitran; extra == "sv"
+Provides-Extra: sw
+Requires-Dist: gruut[sw]<3.0,>=2.3.0; extra == "sw"
+Requires-Dist: epitran; extra == "sw"
+Provides-Extra: ta
+Requires-Dist: epitran; extra == "ta"
+Provides-Extra: te
+Requires-Dist: epitran; extra == "te"
+Provides-Extra: tg
+Requires-Dist: epitran; extra == "tg"
+Provides-Extra: th
+Requires-Dist: epitran; extra == "th"
+Provides-Extra: ti
+Requires-Dist: epitran; extra == "ti"
+Provides-Extra: tk
+Requires-Dist: epitran; extra == "tk"
+Provides-Extra: tpi
+Requires-Dist: epitran; extra == "tpi"
+Provides-Extra: tr
+Requires-Dist: epitran; extra == "tr"
+Provides-Extra: uew
+Requires-Dist: epitran; extra == "uew"
+Provides-Extra: ug
+Requires-Dist: epitran; extra == "ug"
+Provides-Extra: uk
+Requires-Dist: epitran; extra == "uk"
+Provides-Extra: ur
+Requires-Dist: epitran; extra == "ur"
+Provides-Extra: uz
+Requires-Dist: epitran; extra == "uz"
+Provides-Extra: vi
+Requires-Dist: misaki[vi]; extra == "vi"
+Requires-Dist: epitran; extra == "vi"
+Requires-Dist: viphoneme; extra == "vi"
+Requires-Dist: underthesea; extra == "vi"
+Requires-Dist: eng-to-ipa; extra == "vi"
+Requires-Dist: vinorm; extra == "vi"
+Provides-Extra: wuu
+Requires-Dist: epitran; extra == "wuu"
+Provides-Extra: xh
+Requires-Dist: epitran; extra == "xh"
+Provides-Extra: yo
+Requires-Dist: epitran; extra == "yo"
+Provides-Extra: yue
+Requires-Dist: epitran; extra == "yue"
+Provides-Extra: za
+Requires-Dist: epitran; extra == "za"
+Provides-Extra: zh
+Requires-Dist: g2pM; extra == "zh"
+Requires-Dist: misaki[zh]; extra == "zh"
+Requires-Dist: epitran; extra == "zh"
+Requires-Dist: pypinyin; extra == "zh"
+Requires-Dist: xpinyin; extra == "zh"
+Requires-Dist: pinyin-to-ipa; extra == "zh"
+Requires-Dist: jieba; extra == "zh"
+Requires-Dist: cn2an; extra == "zh"
+Provides-Extra: zu
+Requires-Dist: epitran; extra == "zu"
+

phoonnx-0.0.0.dist-info/RECORD

@@ -0,0 +1,86 @@
+phoonnx/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+phoonnx/config.py,sha256=bO7dx2tfLotkohict3UKlCEVm-BRFB1feYYR1HarUkk,19382
+phoonnx/phoneme_ids.py,sha256=FiNgZwV6naEsBh6XwFLh3_FyOgPiCsK9qo7S0v-CmI4,13667
+phoonnx/util.py,sha256=3DqSRwsF498oP_uFVZnriP9PyjXhCfEy97Ey0ePTEqA,25042
+phoonnx/version.py,sha256=peYHqopxd1W3l3PTIN2GzerAUX8Su-_tfbPZbZ4og3o,114
+phoonnx/voice.py,sha256=FR_LafK1vSi_anPERJjZBuH3Bb9vUIof0MAW6TnALlA,20024
+phoonnx/locale/ca/phonetic_spellings.txt,sha256=igv3t7jxLSRE5GHsdn57HOpxiWNcEmECPql6m02wbO0,47
+phoonnx/locale/en/phonetic_spellings.txt,sha256=xGQlWOABLzbttpQvopl9CU-NnwEJRqKx8iuylsdUoQA,27
+phoonnx/locale/gl/phonetic_spellings.txt,sha256=igv3t7jxLSRE5GHsdn57HOpxiWNcEmECPql6m02wbO0,47
+phoonnx/locale/pt/phonetic_spellings.txt,sha256=KntS8QMynEJ5A3Clvcjq4qlmL-ThSbhfD6v0nKSrlqs,49
+phoonnx/phonemizers/__init__.py,sha256=QGBZk0QUgJdg2MwUWY9Kpk6ucwrEJYtHb07YcNvXCV4,1647
+phoonnx/phonemizers/ar.py,sha256=W3wJ_GI8gV3qb9fng5xeazX8ZlY-aNSywATTdxryiZg,1257
+phoonnx/phonemizers/base.py,sha256=yPg6-dvscYpl3rR3JEULG1PRF-i8DWC_C3HAZGLbxOo,7648
+phoonnx/phonemizers/en.py,sha256=N2SVoVhplQao7Ej5TXbxJU-YkAgkY0Fr9iYBFnsjFSE,9271
+phoonnx/phonemizers/fa.py,sha256=d_DZM2wqomf4gcRH_rFcNA3VkQWKHru8vwBwaNG8Ll8,1452
+phoonnx/phonemizers/gl.py,sha256=05DX_9lQa3Uws3RPcWPtw0nXX6Vo3bybrZkpMErguTo,5529
+phoonnx/phonemizers/he.py,sha256=KbRI3XRZa8UtJdNWmn_fd-t5lmFSIp4Mw8UgcO5l-Po,2211
+phoonnx/phonemizers/ja.py,sha256=Xojsrt715ihnIiEk9K6giYqDo9Iykw-SHfIidrHtHSU,3834
+phoonnx/phonemizers/ko.py,sha256=kwWoOFqanCB8kv2JRx17A0hP78P1wbXlX6e8VBn1ezQ,2989
+phoonnx/phonemizers/mul.py,sha256=37G_G58aGnVpdEm9vZEAOdGEHJ9TLBE17bU1HFvQ2rU,27291
+phoonnx/phonemizers/vi.py,sha256=_XJc-Xeawr1Lxr7o8mE_hJao1aGcj4g01XYAOxC_Scg,1311
+phoonnx/phonemizers/zh.py,sha256=88Ywq8h9LDanlyz8RHjRSCY_PRK_Dq808tBADyrgaP8,9657
+phoonnx/thirdparty/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+phoonnx/thirdparty/arpa2ipa.py,sha256=Uj1G5NgP5oBBfSm26LGB8QoumdT-NqCLQTZHT165-_o,5850
+phoonnx/thirdparty/hangul2ipa.py,sha256=e2c0WOy5lFMcf6GS7pNqIbauMKBX07S84lCczZAZJGA,27518
+phoonnx/thirdparty/zh_num.py,sha256=SESA6gvSJW3LZ0FLoybXn2SpbxqhQTi9Tg_U2IZ5JYY,7147
+phoonnx/thirdparty/cotovia/cotovia_aarch64,sha256=BsAWZN452Lm9kDU4i6rQGHFSlmxP3GfHRKhbJMUQrfA,6764592
+phoonnx/thirdparty/cotovia/cotovia_x86_64,sha256=-6BNx_cd49nnDreOAsGtVtePs_X76esrqcNAfmksN1o,1379832
+phoonnx/thirdparty/ko_tables/aspiration.csv,sha256=Dg3W1rr3qe9DDEMSFx5tIIg-WbqOZw9vO1Skg7q2Pmo,219
+phoonnx/thirdparty/ko_tables/assimilation.csv,sha256=WJYLvUitjzAptAw06X620t0Rd25-GU_GRVf8IeaQ_tM,430
+phoonnx/thirdparty/ko_tables/double_coda.csv,sha256=BNr5coYgzNN_P3DmV1Dtj6-UUqlFoe1oC-IiATHBi48,253
+phoonnx/thirdparty/ko_tables/hanja.tsv,sha256=Spwa87HqgCh_sD6xskJLq6A3Sw9AVBFtTloqk7udC6M,93835
+phoonnx/thirdparty/ko_tables/ipa.csv,sha256=y_MFDCmCsJJ-Szu-3xjqXUvydBCVEr4muWsPzRN8aTU,316
+phoonnx/thirdparty/ko_tables/neutralization.csv,sha256=XKgdPy5G7wHXflnThUimDYX-7RwGuJfMZesWl4w44sA,90
+phoonnx/thirdparty/ko_tables/tensification.csv,sha256=V4Xf3A1G1iMBzwZevBKQuk_lPa-U7a7Ra__mz3d3VpQ,780
+phoonnx/thirdparty/ko_tables/yale.csv,sha256=UhtDbPXRAAyAKoQMXmwhVBwJ5pfZQ_Duk28qBtRUdsU,297
+phoonnx/thirdparty/kog2p/__init__.py,sha256=yLizadg7RXM-3dQyftD4XSk8r2jb0QOlHQ6as9uUa4U,10267
+phoonnx/thirdparty/kog2p/rulebook.txt,sha256=FQE3nej8wojl6ilVUBYo7f8bIk0Hjci-B7HPXhM-xNc,9303
+phoonnx/thirdparty/mantoq/__init__.py,sha256=4kZuZ3RA5ZhQwTOQGkHF9jQYSvetNTn9uWi5Dsx101k,2106
+phoonnx/thirdparty/mantoq/num2words.py,sha256=9-ncMtxV1FusD9rNur1lu7l2DWhwUwI1mFiqiPSMH_Q,1264
+phoonnx/thirdparty/mantoq/unicode_symbol2label.py,sha256=CeZNv7qWeQS4Ejvz-sKgK--5eNYdVVv04WHPaOeK4gk,259409
+phoonnx/thirdparty/mantoq/buck/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+phoonnx/thirdparty/mantoq/buck/phonetise_buckwalter.py,sha256=QE2ADBH0HdnGpS6LY5zMqMyobgrVtG3ZDGs1HvbZ9XA,20253
+phoonnx/thirdparty/mantoq/buck/symbols.py,sha256=AoH1kbTVE-0t16CfAEKopIEU0FVAYV3ZRxRJ_i20CFA,1096
+phoonnx/thirdparty/mantoq/buck/tokenization.py,sha256=RyMzu5TLAYy_GybL9GjltwEOPh5iEvoNPpE5fviYbQU,2276
+phoonnx/thirdparty/mantoq/pyarabic/__init__.py,sha256=zTb5JNfQuVye-WlcEmBET25d-Ub4N3Ijk9mCHAD3CLE,226
+phoonnx/thirdparty/mantoq/pyarabic/arabrepr.py,sha256=kKKZyElV9dtWKeYqwIWsUovbK5x8J7kbFKnv2QIwhtk,1757
+phoonnx/thirdparty/mantoq/pyarabic/araby.py,sha256=8kDPpnMKDMOI63h83V7jyCHvn4K2TJjMAhfv4o3NZr4,44797
+phoonnx/thirdparty/mantoq/pyarabic/named_const.py,sha256=CtT5PGSwBYUombw8MkdyW6tu8G3-ghuIK2sd-4y37X4,4455
+phoonnx/thirdparty/mantoq/pyarabic/normalize.py,sha256=iaTyTOWHNc38Ov2hgOkD0zDJmWM-kUoEba21es6CpJ8,4845
+phoonnx/thirdparty/mantoq/pyarabic/number.py,sha256=NjFZPWRu-9dZDLgxfv9oDjmh-kWYkaUGntPQC7iT5mQ,27041
+phoonnx/thirdparty/mantoq/pyarabic/number_const.py,sha256=vAvRVENxTrl9gWPllSXF-yqK9fAW6htuA2d041btC_A,42361
+phoonnx/thirdparty/mantoq/pyarabic/stack.py,sha256=aJeSzQxVNdomDTWXuxIXWXVOc2BW_3iRWnwmBLkB8jM,1022
+phoonnx/thirdparty/mantoq/pyarabic/trans.py,sha256=cusyHk9Y01iuvMLJXxgCnIiGyAORzEdSosDKX4cAhPc,13713
+phoonnx/thirdparty/tashkeel/LICENSE,sha256=mQjTJ6MGAXzmYkO7x4O2VuEeSwCMx7lncbc26TnrVjw,1067
+phoonnx/thirdparty/tashkeel/SOURCE,sha256=SmnRz-Am5EXv-n2-RokJVEhnn8zeF1QZJVvMQDA_Qds,38
+phoonnx/thirdparty/tashkeel/__init__.py,sha256=FRdGNCTQaai9X077vlNh4tFOvWgm1U2lIUgnQKO5q0s,7119
+phoonnx/thirdparty/tashkeel/hint_id_map.json,sha256=gJMdtTsfEDFgmmbyO2Shw315rkqKTHhQI3KGNqf6k-U,193
+phoonnx/thirdparty/tashkeel/input_id_map.json,sha256=cnpJqjx-k53AbzKyfC4GxMS771ltzkv1EnYmHKc2w8M,628
+phoonnx/thirdparty/tashkeel/model.onnx,sha256=UsQNQsoJT_n_B6CR0KHq_XuqXPI4jmCpzIm6zY5elV8,4788213
+phoonnx/thirdparty/tashkeel/target_id_map.json,sha256=baNAJL_UwP9U91mLt01aAEBRRNdGr-csFB_O6roh7TA,181
+phoonnx_train/__main__.py,sha256=FUAIsbQ-w2i_hoNiBuriQFk4uoryhL4ydyVY-hVjw1U,5086
+phoonnx_train/export_onnx.py,sha256=dcFJRZl4YvBk_Dj3j0aNAQVEqKfBHTzV22pzvQwSETQ,2909
+phoonnx_train/preprocess.py,sha256=0kto9Holywby6lnoQucBXq2wYEKDItRvdkvYbQnLJeo,14447
+phoonnx_train/norm_audio/__init__.py,sha256=Al_YwqMnENXRWp0c79cDZqbdd7pFYARXKxCfBaedr1c,3030
+phoonnx_train/norm_audio/trim.py,sha256=_ZsE3SYhahQSdEdBLeSwyFJGcvEbt-5E_lnWwTT4tcY,1698
+phoonnx_train/norm_audio/vad.py,sha256=DXHfRD0qqFJ52FjPvrL5LlN6keJWuc9Nf6TNhxpwC_4,1600
+phoonnx_train/vits/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+phoonnx_train/vits/attentions.py,sha256=yc_ViF8zR8z68DzphmVVVn27f9xK_5wi8S4ITLXVQL0,15134
+phoonnx_train/vits/commons.py,sha256=JsD8CdZ3ZcYYubYhw8So5hICBziFlCrKLrv1lMDRCDM,4645
+phoonnx_train/vits/config.py,sha256=oSuUIhw9Am7BQ5JwDgtCO-P1zRyN7nPgR-U1XuncJls,10789
+phoonnx_train/vits/dataset.py,sha256=DLLGSCkn3GF9uktoTprH1ERblZ18GO6-QsClQKWa98o,6804
+phoonnx_train/vits/lightning.py,sha256=iJJNAT1-jm0wYvTfBy9T_eZapDN62Pny9YdHEpnApso,12298
+phoonnx_train/vits/losses.py,sha256=j-uINhBcYxVXFvFutiewQpTuw-qF-J6M6hdJVeOKqNE,1401
+phoonnx_train/vits/mel_processing.py,sha256=huIjbQgewSmM39hdzRZvZUCI7fTNSMmLcAv3f8zYb8k,3956
+phoonnx_train/vits/models.py,sha256=9PziprRtkdBQ6AowFe1vG4QTCk02By-LDS9W8EtZGvE,24303
+phoonnx_train/vits/modules.py,sha256=DMNev5g23gkSq1EP0GR-jDI30Dvoy335k4q4zYKO1Y0,16880
+phoonnx_train/vits/transforms.py,sha256=On8hI8WPD6noP_X0RKrGv5FzJ7bnmFVGlC-SJ_UuAY0,7457
+phoonnx_train/vits/utils.py,sha256=exiyrtPHbnnGvcHWSbaH9-gR6srH5ZPHlKiqV2IHUrQ,491
+phoonnx_train/vits/wavfile.py,sha256=oQZiTIrdw0oLTbcVwKfGXye1WtKte6qK_52qVwiMvfc,26396
+phoonnx_train/vits/monotonic_align/__init__.py,sha256=5IdAOD1Z7UloMb6d_9NRFsXoNIjEQ3h9mvOSh_AtO3k,636
+phoonnx_train/vits/monotonic_align/setup.py,sha256=0K5iJJ2mKIklx6ncEfCQS34skm5hHPiz9vRlQEvevvY,266
+phoonnx-0.0.0.dist-info/METADATA,sha256=403yc0XWgga5G22oBJb9xTb2kEfYVKRcI1O7rCnYOls,7866
+phoonnx-0.0.0.dist-info/WHEEL,sha256=tZoeGjtWxWRfdplE7E3d45VPlLNQnvbKiYnx7gwAy8A,92
+phoonnx-0.0.0.dist-info/top_level.txt,sha256=ZrnHXe-4HqbOSX6fbdY-JiP7YEu2Bok9T0ji351MrmM,22
+phoonnx-0.0.0.dist-info/RECORD,,

phoonnx-0.0.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: bdist_wheel (0.45.1)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

phoonnx-0.0.0.dist-info/top_level.txt

@@ -0,0 +1,2 @@
+phoonnx
+phoonnx_train

phoonnx_train/__main__.py

@@ -0,0 +1,151 @@
+import argparse
+import json
+import logging
+from pathlib import Path
+
+import torch
+from pytorch_lightning import Trainer
+from pytorch_lightning.callbacks import ModelCheckpoint
+
+from phoonnx_train.vits.lightning import VitsModel
+
+_LOGGER = logging.getLogger(__package__)
+
+
+def main():
+    logging.basicConfig(level=logging.DEBUG)
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--dataset-dir", required=True, help="Path to pre-processed dataset directory"
+    )
+    parser.add_argument(
+        "--checkpoint-epochs",
+        type=int,
+        help="Save checkpoint every N epochs (default: 1)",
+    )
+    parser.add_argument(
+        "--quality",
+        default="medium",
+        choices=("x-low", "medium", "high"),
+        help="Quality/size of model (default: medium)",
+    )
+    parser.add_argument(
+        "--resume_from_single_speaker_checkpoint",
+        help="For multi-speaker models only. Converts a single-speaker checkpoint to multi-speaker and resumes training",
+    )
+    Trainer.add_argparse_args(parser)
+    VitsModel.add_model_specific_args(parser)
+    parser.add_argument("--seed", type=int, default=1234)
+    args = parser.parse_args()
+    _LOGGER.debug(args)
+
+    args.dataset_dir = Path(args.dataset_dir)
+    if not args.default_root_dir:
+        args.default_root_dir = args.dataset_dir
+
+    torch.backends.cudnn.benchmark = True
+    torch.manual_seed(args.seed)
+
+    config_path = args.dataset_dir / "config.json"
+    dataset_path = args.dataset_dir / "dataset.jsonl"
+
+    print(f"INFO - config_path: '{config_path}'")
+    print(f"INFO - dataset_path: '{dataset_path}'")
+    
+    with open(config_path, "r", encoding="utf-8") as config_file:
+        # See preprocess.py for format
+        config = json.load(config_file)
+        num_symbols = int(config["num_symbols"])
+        num_speakers = int(config["num_speakers"])
+        sample_rate = int(config["audio"]["sample_rate"])
+
+    trainer = Trainer.from_argparse_args(args)
+    if args.checkpoint_epochs is not None:
+        trainer.callbacks = [ModelCheckpoint(every_n_epochs=args.checkpoint_epochs)]
+        _LOGGER.info(
+            "Checkpoints will be saved every %s epoch(s)", args.checkpoint_epochs
+        )
+
+    dict_args = vars(args)
+    if args.quality == "x-low":
+        dict_args["hidden_channels"] = 96
+        dict_args["inter_channels"] = 96
+        dict_args["filter_channels"] = 384
+    elif args.quality == "high":
+        dict_args["resblock"] = "1"
+        dict_args["resblock_kernel_sizes"] = (3, 7, 11)
+        dict_args["resblock_dilation_sizes"] = (
+            (1, 3, 5),
+            (1, 3, 5),
+            (1, 3, 5),
+        )
+        dict_args["upsample_rates"] = (8, 8, 2, 2)
+        dict_args["upsample_initial_channel"] = 512
+        dict_args["upsample_kernel_sizes"] = (16, 16, 4, 4)
+
+    print(f"VitsModel params: num_symbols={num_symbols} num_speakers={num_speakers} sample_rate={sample_rate}")
+    model = VitsModel(
+        num_symbols=num_symbols,
+        num_speakers=num_speakers,
+        sample_rate=sample_rate,
+        dataset=[dataset_path],
+        **dict_args,
+    )
+
+    if args.resume_from_single_speaker_checkpoint:
+        assert (
+            num_speakers > 1
+        ), "--resume_from_single_speaker_checkpoint is only for multi-speaker models. Use --resume_from_checkpoint for single-speaker models."
+
+        # Load single-speaker checkpoint
+        _LOGGER.info(
+            "Resuming from single-speaker checkpoint: %s",
+            args.resume_from_single_speaker_checkpoint,
+        )
+        model_single = VitsModel.load_from_checkpoint(
+            args.resume_from_single_speaker_checkpoint,
+            dataset=None,
+        )
+        g_dict = model_single.model_g.state_dict()
+        for key in list(g_dict.keys()):
+            # Remove keys that can't be copied over due to missing speaker embedding
+            if (
+                key.startswith("dec.cond")
+                or key.startswith("dp.cond")
+                or ("enc.cond_layer" in key)
+            ):
+                g_dict.pop(key, None)
+
+        # Copy over the multi-speaker model, excluding keys related to the
+        # speaker embedding (which is missing from the single-speaker model).
+        load_state_dict(model.model_g, g_dict)
+        load_state_dict(model.model_d, model_single.model_d.state_dict())
+        _LOGGER.info(
+            "Successfully converted single-speaker checkpoint to multi-speaker"
+        )
+    print("training started!!")
+    trainer.fit(model)
+
+
+def load_state_dict(model, saved_state_dict):
+    state_dict = model.state_dict()
+    new_state_dict = {}
+
+    for k, v in state_dict.items():
+        if k in saved_state_dict:
+            # Use saved value
+            new_state_dict[k] = saved_state_dict[k]
+        else:
+            # Use initialized value
+            _LOGGER.debug("%s is not in the checkpoint", k)
+            new_state_dict[k] = v
+
+    model.load_state_dict(new_state_dict)
+
+
+# -----------------------------------------------------------------------------
+
+
+if __name__ == "__main__":
+    main()

phoonnx_train/export_onnx.py

@@ -0,0 +1,109 @@
+#!/usr/bin/env python3
+import argparse
+import logging
+from pathlib import Path
+from typing import Optional
+
+import torch
+
+from phoonnx_train.vits.lightning import VitsModel
+
+_LOGGER = logging.getLogger("piper_train.export_onnx")
+
+OPSET_VERSION = 15
+
+
+def main() -> None:
+    """Main entry point"""
+    torch.manual_seed(1234)
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint", help="Path to model checkpoint (.ckpt)")
+    parser.add_argument("output", help="Path to output model (.onnx)")
+
+    parser.add_argument(
+        "--debug", action="store_true", help="Print DEBUG messages to the console"
+    )
+    args = parser.parse_args()
+
+    if args.debug:
+        logging.basicConfig(level=logging.DEBUG)
+    else:
+        logging.basicConfig(level=logging.INFO)
+
+    _LOGGER.debug(args)
+
+    # -------------------------------------------------------------------------
+
+    args.checkpoint = Path(args.checkpoint)
+    args.output = Path(args.output)
+    args.output.parent.mkdir(parents=True, exist_ok=True)
+
+    model = VitsModel.load_from_checkpoint(args.checkpoint, dataset=None)
+    model_g = model.model_g
+
+    num_symbols = model_g.n_vocab
+    num_speakers = model_g.n_speakers
+
+    # Inference only
+    model_g.eval()
+
+    with torch.no_grad():
+        model_g.dec.remove_weight_norm()
+
+    # old_forward = model_g.infer
+
+    def infer_forward(text, text_lengths, scales, sid=None):
+        noise_scale = scales[0]
+        length_scale = scales[1]
+        noise_scale_w = scales[2]
+        audio = model_g.infer(
+            text,
+            text_lengths,
+            noise_scale=noise_scale,
+            length_scale=length_scale,
+            noise_scale_w=noise_scale_w,
+            sid=sid,
+        )[0].unsqueeze(1)
+
+        return audio
+
+    model_g.forward = infer_forward
+
+    dummy_input_length = 50
+    sequences = torch.randint(
+        low=0, high=num_symbols, size=(1, dummy_input_length), dtype=torch.long
+    )
+    sequence_lengths = torch.LongTensor([sequences.size(1)])
+
+    sid: Optional[torch.LongTensor] = None
+    if num_speakers > 1:
+        sid = torch.LongTensor([0])
+
+    # noise, noise_w, length
+    scales = torch.FloatTensor([0.667, 1.0, 0.8])
+    dummy_input = (sequences, sequence_lengths, scales, sid)
+
+    # Export
+    torch.onnx.export(
+        model=model_g,
+        args=dummy_input,
+        f=str(args.output),
+        verbose=False,
+        opset_version=OPSET_VERSION,
+        input_names=["input", "input_lengths", "scales", "sid"],
+        output_names=["output"],
+        dynamic_axes={
+            "input": {0: "batch_size", 1: "phonemes"},
+            "input_lengths": {0: "batch_size"},
+            "output": {0: "batch_size", 1: "time"},
+        },
+    )
+
+    _LOGGER.info("Exported model to %s", args.output)
+
+
+# -----------------------------------------------------------------------------
+
+if __name__ == "__main__":
+    main()