OpenSTBench 1.1.0__tar.gz → 1.2.0__tar.gz

{openstbench-1.1.0/src/OpenSTBench.egg-info → openstbench-1.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: OpenSTBench
-Version: 1.1.0
+Version: 1.2.0
 Summary: Unified multidimensional evaluation toolkit for S2TT and S2ST systems in offline and streaming speech translation settings
 Author-email: Yanjie An <691476922@qq.com>
 License: MIT
@@ -17,6 +17,7 @@ Classifier: Programming Language :: Python :: 3
 Requires-Python: <3.11,>=3.9
 Description-Content-Type: text/markdown
 License-File: LICENSE
+Requires-Dist: setuptools==80.9.0
 Requires-Dist: torch<2.7,>=2.6
 Requires-Dist: torchaudio<2.7,>=2.6
 Requires-Dist: numpy
@@ -36,6 +37,7 @@ Provides-Extra: whisper
 Requires-Dist: openai-whisper; extra == "whisper"
 Provides-Extra: speech-quality
 Requires-Dist: jiwer; extra == "speech-quality"
+Requires-Dist: opencc-python-reimplemented; extra == "speech-quality"
 Provides-Extra: emotion
 Requires-Dist: funasr; extra == "emotion"
 Requires-Dist: modelscope; extra == "emotion"
@@ -45,6 +47,7 @@ Provides-Extra: all
 Requires-Dist: unbabel-comet>=2.2.2; extra == "all"
 Requires-Dist: openai-whisper; extra == "all"
 Requires-Dist: jiwer; extra == "all"
+Requires-Dist: opencc-python-reimplemented; extra == "all"
 Requires-Dist: funasr; extra == "all"
 Requires-Dist: modelscope; extra == "all"
 Requires-Dist: transformers; extra == "all"

{openstbench-1.1.0 → openstbench-1.2.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "OpenSTBench"
-version = "1.1.0"  
+version = "1.2.0"  
 description = "Unified multidimensional evaluation toolkit for S2TT and S2ST systems in offline and streaming speech translation settings"
 readme = {file = "README.md", content-type = "text/markdown"}
 requires-python = ">=3.9,<3.11"
@@ -29,6 +29,7 @@ classifiers = [
 ]
 
 dependencies = [
+    "setuptools==80.9.0",
     "torch>=2.6,<2.7",
     "torchaudio>=2.6,<2.7",
     "numpy",
@@ -53,10 +54,10 @@ Documentation = "https://github.com/sjtuayj/OpenSTBench#readme"
 [project.optional-dependencies]
 comet = ["unbabel-comet>=2.2.2"]
 whisper = ["openai-whisper"]
-speech_quality = ["jiwer"]
+speech_quality = ["jiwer", "opencc-python-reimplemented"]
 emotion = ["funasr", "modelscope"]  
 paralinguistics = ["transformers"]
-all = ["unbabel-comet>=2.2.2", "openai-whisper", "jiwer", "funasr", "modelscope", "transformers"]
+all = ["unbabel-comet>=2.2.2", "openai-whisper", "jiwer", "opencc-python-reimplemented", "funasr", "modelscope", "transformers"]
 
 [tool.setuptools]
 package-dir = {"" = "src"}

{openstbench-1.1.0 → openstbench-1.2.0/src/OpenSTBench.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: OpenSTBench
-Version: 1.1.0
+Version: 1.2.0
 Summary: Unified multidimensional evaluation toolkit for S2TT and S2ST systems in offline and streaming speech translation settings
 Author-email: Yanjie An <691476922@qq.com>
 License: MIT
@@ -17,6 +17,7 @@ Classifier: Programming Language :: Python :: 3
 Requires-Python: <3.11,>=3.9
 Description-Content-Type: text/markdown
 License-File: LICENSE
+Requires-Dist: setuptools==80.9.0
 Requires-Dist: torch<2.7,>=2.6
 Requires-Dist: torchaudio<2.7,>=2.6
 Requires-Dist: numpy
@@ -36,6 +37,7 @@ Provides-Extra: whisper
 Requires-Dist: openai-whisper; extra == "whisper"
 Provides-Extra: speech-quality
 Requires-Dist: jiwer; extra == "speech-quality"
+Requires-Dist: opencc-python-reimplemented; extra == "speech-quality"
 Provides-Extra: emotion
 Requires-Dist: funasr; extra == "emotion"
 Requires-Dist: modelscope; extra == "emotion"
@@ -45,6 +47,7 @@ Provides-Extra: all
 Requires-Dist: unbabel-comet>=2.2.2; extra == "all"
 Requires-Dist: openai-whisper; extra == "all"
 Requires-Dist: jiwer; extra == "all"
+Requires-Dist: opencc-python-reimplemented; extra == "all"
 Requires-Dist: funasr; extra == "all"
 Requires-Dist: modelscope; extra == "all"
 Requires-Dist: transformers; extra == "all"

{openstbench-1.1.0 → openstbench-1.2.0}/src/OpenSTBench.egg-info/requires.txt

@@ -1,3 +1,4 @@
+setuptools==80.9.0
 torch<2.7,>=2.6
 torchaudio<2.7,>=2.6
 numpy
@@ -16,6 +17,7 @@ resemblyzer
 unbabel-comet>=2.2.2
 openai-whisper
 jiwer
+opencc-python-reimplemented
 funasr
 modelscope
 transformers
@@ -32,6 +34,7 @@ transformers
 
 [speech_quality]
 jiwer
+opencc-python-reimplemented
 
 [whisper]
 openai-whisper

{openstbench-1.1.0 → openstbench-1.2.0}/src/openstbench/__init__.py

@@ -9,7 +9,7 @@ actually accessed.
 from importlib import import_module
 from typing import Dict, Tuple
 
-__version__ = "1.1.0"
+__version__ = "1.2.0"
 
 __all__ = [
     "TranslationEvaluator",

{openstbench-1.1.0 → openstbench-1.2.0}/src/openstbench/speech_quality_evaluator.py

@@ -15,6 +15,11 @@ try:
 except ImportError:
     whisper = None
 
+try:
+    from opencc import OpenCC
+except ImportError:
+    OpenCC = None
+
 class SpeechQualityEvaluator:
     DEFAULT_WHISPER_MODEL = "medium"
     """
@@ -40,6 +45,8 @@ class SpeechQualityEvaluator:
         self.whisper_language = self._normalize_whisper_language(whisper_language)
         self.utmos_path = utmos_model_path
         self.utmos_ckpt = utmos_ckpt_path
+        self.zh_converter = OpenCC("t2s") if OpenCC is not None else None
+        self._warned_missing_opencc = False
         
         self.whisper_model = None
         self.utmos_model = None
@@ -148,8 +155,15 @@ class SpeechQualityEvaluator:
         return sum(scores) / len(scores) if scores else 0.0
 
     def _preprocess_for_wer(self, text: str, lang: str) -> str:
+        normalized_lang = self._normalize_whisper_language(lang) or str(lang).strip().lower()
         text = self.wer_transform(text)
-        if lang in ["zh", "ja", "ko"]:
+        if normalized_lang == "zh":
+            if self.zh_converter is not None:
+                text = self.zh_converter.convert(text)
+            elif not self._warned_missing_opencc:
+                print("Warning: OpenCC is not installed; Chinese CER will not normalize traditional/simplified variants.")
+                self._warned_missing_opencc = True
+        if normalized_lang in ["zh", "ja", "ko"]:
             text = text.replace(" ", "")
             return " ".join(list(text))
         return text
@@ -194,7 +208,8 @@ class SpeechQualityEvaluator:
                 
                 if clean_refs:
                     error_rate = jiwer.wer(clean_refs, clean_hyps)
-                    metric_name = "CER_Consistency" if target_lang in ['zh', 'ja', 'ko'] else "WER_Consistency"
+                    normalized_target_lang = self._normalize_whisper_language(target_lang) or str(target_lang).strip().lower()
+                    metric_name = "CER_Consistency" if normalized_target_lang in ['zh', 'ja', 'ko'] else "WER_Consistency"
                     results[metric_name] = round(error_rate, 4)
                 else:
                     print("   ⚠️ The provided generated text is empty after cleaning, cannot compute distribution error rate.")