phoonnx 0.2.4a1__py3-none-any.whl → 0.2.6__py3-none-any.whl

phoonnx/version.py

@@ -1,8 +1,8 @@
 # START_VERSION_BLOCK
 VERSION_MAJOR = 0
 VERSION_MINOR = 2
-VERSION_BUILD = 4
-VERSION_ALPHA = 1
+VERSION_BUILD = 6
+VERSION_ALPHA = 0
 # END_VERSION_BLOCK
 
 VERSION_STR = f"{VERSION_MAJOR}.{VERSION_MINOR}.{VERSION_BUILD}"

phoonnx/voice.py

@@ -203,7 +203,6 @@ class TTSVoice:
             phonemes = self.phonemizer.phonemize(
                 text_part, self.config.lang_code
             )
-            phonemes.extend(phonemes)
 
         if phonemes and (not phonemes[-1]):
             # Remove empty phonemes

{phoonnx-0.2.4a1.dist-info → phoonnx-0.2.6.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: phoonnx
-Version: 0.2.4a1
+Version: 0.2.6
 Home-page: https://github.com/TigreGotico/phoonnx
 Author: JarbasAi
 Author-email: jarbasai@mailfence.com

{phoonnx-0.2.4a1.dist-info → phoonnx-0.2.6.dist-info}/RECORD

@@ -2,8 +2,8 @@ phoonnx/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 phoonnx/config.py,sha256=DKgsU03g8jrAuMcVqbu-w3MWPXOUihFtRnavg6WGQ1Y,19983
 phoonnx/phoneme_ids.py,sha256=FiNgZwV6naEsBh6XwFLh3_FyOgPiCsK9qo7S0v-CmI4,13667
 phoonnx/util.py,sha256=XSjFEoqSFcujFTHxednacgC9GrSYyF-Il5L6Utmxmu4,25909
-phoonnx/version.py,sha256=VZdcd-pCi-vCStncJnB1n_gha4f3TNjR-W_dSyoACOw,237
-phoonnx/voice.py,sha256=JXjmbrhJd4mmTiLgz4O_Pa5_rKGUC9xzuBfqxYDw3Mg,19420
+phoonnx/version.py,sha256=Ed7aHpgOwiCu6_EBeRxVmQI4Pm7CgBOFmHep_sKNUO0,237
+phoonnx/voice.py,sha256=Fh60o87-_sIw4zTP8NEGvEV7XzTC0FhW9Jkq7Fr60lQ,19382
 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
@@ -62,8 +62,8 @@ phoonnx/thirdparty/tashkeel/hint_id_map.json,sha256=gJMdtTsfEDFgmmbyO2Shw315rkqK
 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/export_onnx.py,sha256=CPfgNEm0hnXPSlgme0R9jr-6jZ5fKFpG5DZJFMkC-h4,12820
-phoonnx_train/preprocess.py,sha256=T1YcM89bizARKMaqjDtr3JzidoNvAbm-3n3eClbzMsI,21880
+phoonnx_train/export_onnx.py,sha256=UFn1id5-p9tIXxK9FvHOry59WkotNqXtupinkeKcSDs,12881
+phoonnx_train/preprocess.py,sha256=VYyU8XaCroa6tnLhQkQJ8wq1w1Gf9PPnIIM1N0PIcNE,22042
 phoonnx_train/train.py,sha256=nsINvDQ3dYvBne5UWPgLZ0a4qZFdSsOKk8HzZHGTLY4,8757
 phoonnx_train/norm_audio/__init__.py,sha256=Al_YwqMnENXRWp0c79cDZqbdd7pFYARXKxCfBaedr1c,3030
 phoonnx_train/norm_audio/trim.py,sha256=_ZsE3SYhahQSdEdBLeSwyFJGcvEbt-5E_lnWwTT4tcY,1698
@@ -83,7 +83,7 @@ phoonnx_train/vits/utils.py,sha256=exiyrtPHbnnGvcHWSbaH9-gR6srH5ZPHlKiqV2IHUrQ,4
 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-0.2.4a1.dist-info/METADATA,sha256=p_Hn1MLi2aPfctGgz_APHUmfwxl8nBkjozwqdxbU4b4,8250
-phoonnx-0.2.4a1.dist-info/WHEEL,sha256=tZoeGjtWxWRfdplE7E3d45VPlLNQnvbKiYnx7gwAy8A,92
-phoonnx-0.2.4a1.dist-info/top_level.txt,sha256=ZrnHXe-4HqbOSX6fbdY-JiP7YEu2Bok9T0ji351MrmM,22
-phoonnx-0.2.4a1.dist-info/RECORD,,
+phoonnx-0.2.6.dist-info/METADATA,sha256=Lg0o0_aPp_pb-Dhvcx3IMgw5LiDHWwbfyXIKF8blFWE,8248
+phoonnx-0.2.6.dist-info/WHEEL,sha256=tZoeGjtWxWRfdplE7E3d45VPlLNQnvbKiYnx7gwAy8A,92
+phoonnx-0.2.6.dist-info/top_level.txt,sha256=ZrnHXe-4HqbOSX6fbdY-JiP7YEu2Bok9T0ji351MrmM,22
+phoonnx-0.2.6.dist-info/RECORD,,

phoonnx_train/export_onnx.py

@@ -113,7 +113,8 @@ def convert_to_piper(config_path: Path, output_path: Path = Path("piper.json"))
         "phoneme_map": {},
         "audio": config.get("audio", {}),
         "inference": config.get("inference", {}),
-        "phoneme_id_map": {k: [v] for k, v in config.get("phoneme_id_map", {}).items()},
+        "phoneme_id_map": {k: [v] if not isinstance(v, list) else v
+                           for k, v in config.get("phoneme_id_map", {}).items()},
         "espeak": {
             "voice": config.get("lang_code", "")
         },

phoonnx_train/preprocess.py

@@ -518,7 +518,10 @@ def cli(
     if prev_config:
         with open(prev_config) as f:
             cfg = json.load(f)
-        prev_phoneme_id_map = cfg["phoneme_id_map"]
+        # flatten list, same models (eg. piper) use a list of ids
+        prev_phoneme_id_map = {k: v if not isinstance(v, list) else v[0]
+                               for k, v in cfg["phoneme_id_map"].items()}
+
         prev_num_symbols = cfg.get("num_symbols", MAX_PHONEMES)
         _LOGGER.info(f"Loaded phoneme map from previous config: '{prev_config}'")
         all_phonemes.update(prev_phoneme_id_map.keys())