Refactor build_json_result and stream_builder functions

requirements_version.txt

@@ -1,48 +1,48 @@
-annotated-types==0.6.0
+annotated-types==0.7.0
-anyio==3.7.1
+anyio==4.4.0
-av==10.0.0
+av==12.2.0
-certifi==2023.7.22
+certifi==2024.7.4
 cffi==1.16.0
 charset-normalizer==3.3.2
 click==8.1.7
 coloredlogs==15.0.1
-ctranslate2==3.21.0
+ctranslate2==4.3.1
-fastapi==0.104.1
+fastapi==0.111.0
-faster-whisper==0.9.0
+faster-whisper==1.0.3
-filelock==3.13.1
+filelock==3.15.4
-flatbuffers==23.5.26
+flatbuffers==24.3.25
-fsspec==2023.10.0
+fsspec==2024.6.1
 h11==0.14.0
 httptools==0.6.1
-huggingface-hub==0.17.3
+huggingface-hub==0.23.4
 humanfriendly==10.0
-idna==3.4
+idna==3.7
 mpmath==1.3.0
-numpy==1.26.2
+numpy==1.26.4
-onnxruntime==1.16.2
+onnxruntime==1.18.1
 OpenCC==1.1.7
-packaging==23.2
+packaging==24.1
 prometheus-client==0.18.0
-prometheus-fastapi-instrumentator==6.1.0
+prometheus-fastapi-instrumentator==7.0.0
-protobuf==4.25.0
+protobuf==5.27.2
-pycparser==2.21
+pycparser==2.22
-pydantic==2.5.0
+pydantic==2.8.2
-pydantic_core==2.14.1
+pydantic_core==2.20.1
 pydub==0.25.1
-python-dotenv==1.0.0
+python-dotenv==1.0.1
-python-multipart==0.0.6
+python-multipart==0.0.9
 PyYAML==6.0.1
-requests==2.31.0
+requests==2.32.3
-sniffio==1.3.0
+sniffio==1.3.1
-sounddevice==0.4.6
+sounddevice==0.4.7
-starlette==0.27.0
+starlette==0.37.2
-sympy==1.12
+sympy==1.12.1
-tokenizers==0.14.1
+tokenizers==0.19.1
-tqdm==4.66.1
+tqdm==4.66.4
-typing_extensions==4.8.0
+typing_extensions==4.12.2
-urllib3==2.1.0
+urllib3==2.2.2
-uvicorn==0.24.0.post1
+uvicorn==0.30.1
 uvloop==0.19.0
-watchfiles==0.21.0
+watchfiles==0.22.0
 websockets==12.0
-whisper-ctranslate2==0.3.2
+whisper-ctranslate2==0.4.5

whisper_fastapi.py

@@ -1,4 +1,3 @@
-from faster_whisper import vad
 import tqdm
 import json
 from fastapi.responses import StreamingResponse
@@ -8,7 +7,7 @@ import io
 import hashlib
 import argparse
 import uvicorn
-from typing import Annotated, Any, BinaryIO, Literal, Generator
+from typing import Annotated, Any, BinaryIO, Literal, Generator, Tuple, Iterable
 from fastapi import (
     File,
     HTTPException,
@@ -22,20 +21,21 @@ from fastapi import (
 from fastapi.middleware.cors import CORSMiddleware
 from src.whisper_ctranslate2.whisper_ctranslate2 import Transcribe
 from src.whisper_ctranslate2.writers import format_timestamp
-import opencc
+from faster_whisper.transcribe import Segment, TranscriptionInfo
 from prometheus_fastapi_instrumentator import Instrumentator
 
 parser = argparse.ArgumentParser()
 parser.add_argument("--host", default="0.0.0.0", type=str)
 parser.add_argument("--port", default=5000, type=int)
-parser.add_argument("--model", default="large-v2", type=str)
+parser.add_argument("--model", default="large-v3", type=str)
 parser.add_argument("--device", default="auto", type=str)
 parser.add_argument("--cache_dir", default=None, type=str)
+parser.add_argument("--local_files_only", default=False, type=bool)
+parser.add_argument("--threads", default=4, type=int)
 args = parser.parse_args()
 app = FastAPI()
 # Instrument your app with default metrics and expose the metrics
 Instrumentator().instrument(app).expose(app, endpoint="/konele/metrics")
-ccc = opencc.OpenCC("t2s.json")
 
 print("Loading model...")
 transcriber = Transcribe(
@@ -43,9 +43,9 @@ transcriber = Transcribe(
     device=args.device,
     device_index=0,
     compute_type="default",
-    threads=1,
+    threads=args.threads,
     cache_directory=args.cache_dir,
-    local_files_only=False,
+    local_files_only=args.local_files_only,
 )
 print("Model loaded!")
 
@@ -102,10 +102,11 @@ def vtt_writer(generator: Generator[dict[str, Any], Any, None]):
 
 
 def build_json_result(
-    generator: Generator[dict[str, Any], Any, None]
+    generator: Iterable[Segment],
+    info:  dict,
 ) -> dict[str, Any]:
     segments = [i for i in generator]
-    return {
+    return info | {
         "text": "\n".join(i["text"] for i in segments),
         "segments": segments,
     }
@@ -117,46 +118,39 @@ def stream_builder(
     vad_filter: bool,
     language: str | None,
     initial_prompt: str = "",
-):
+    repetition_penalty: float = 1.0,
+) -> Tuple[Iterable[dict], dict]:
     segments, info = transcriber.model.transcribe(
         audio=audio,
         language=language,
         task=task,
-        beam_size=5,
-        best_of=5,
-        patience=1.0,
-        length_penalty=-1.0,
-        repetition_penalty=1.0,
-        no_repeat_ngram_size=0,
-        temperature=[0.0, 1.0 + 1e-6, 0.2],
-        compression_ratio_threshold=2.4,
-        log_prob_threshold=-1.0,
-        no_speech_threshold=0.6,
-        condition_on_previous_text=True,
-        prompt_reset_on_temperature=False,
         initial_prompt=initial_prompt,
-        suppress_blank=False,
-        suppress_tokens=[],
         word_timestamps=True,
-        prepend_punctuations="\"'“¿([{-",
+        repetition_penalty=repetition_penalty,
-        append_punctuations="\"'.。,，!！?？:：”)]}、",
-        vad_filter=vad_filter,
-        vad_parameters=None,
     )
     print(
         "Detected language '%s' with probability %f"
         % (info.language, info.language_probability)
     )
-    last_pos = 0
+    def wrap():
-    with tqdm.tqdm(total=info.duration, unit="seconds", disable=True) as pbar:
+        last_pos = 0
-        for segment in segments:
+        with tqdm.tqdm(total=info.duration, unit="seconds", disable=True) as pbar:
-            start, end, text = segment.start, segment.end, segment.text
+            for segment in segments:
-            pbar.update(end - last_pos)
+                start, end, text = segment.start, segment.end, segment.text
-            last_pos = end
+                pbar.update(end - last_pos)
-            data = segment._asdict()
+                last_pos = end
-            data["total"] = info.duration
+                data = segment._asdict()
-            data["text"] = ccc.convert(data["text"])
+                if data.get('words') is not None:
-            yield data
+                    data["words"] = [i._asdict() for i in data["words"]]
+                yield data
+
+    info_dict = info._asdict()
+    if info_dict['transcription_options'] is not None:
+        info_dict['transcription_options'] = info_dict['transcription_options']._asdict()
+    if info_dict['vad_options'] is not None:
+        info_dict['vad_options'] = info_dict['vad_options']._asdict()
+    
+    return wrap(), info_dict
 
 
 @app.websocket("/k6nele/status")
@@ -215,14 +209,14 @@ async def konele_ws(
 
     file_obj.seek(0)
 
-    generator = stream_builder(
+    generator, info = stream_builder(
         audio=file_obj,
         task=task,
         vad_filter=vad_filter,
         language=None if lang == "und" else lang,
         initial_prompt=initial_prompt,
     )
-    result = build_json_result(generator)
+    result = build_json_result(generator, info)
 
     text = result.get("text", "")
     print("result", text)
@@ -279,14 +273,14 @@ async def translateapi(
 
     file_obj.seek(0)
 
-    generator = stream_builder(
+    generator, info = stream_builder(
         audio=file_obj,
         task=task,
         vad_filter=vad_filter,
         language=None if lang == "und" else lang,
         initial_prompt=initial_prompt,
     )
-    result = build_json_result(generator)
+    result = build_json_result(generator, info)
 
     text = result.get("text", "")
     print("result", text)
@@ -306,6 +300,7 @@ async def transcription(
     task: str = Form("transcribe"),
     language: str = Form("und"),
     vad_filter: bool = Form(False),
+    repetition_penalty: float = Form(1.0),
 ):
     """Transcription endpoint
 
@@ -313,11 +308,12 @@ async def transcription(
     """
 
     # timestamp as filename, keep original extension
-    generator = stream_builder(
+    generator, info = stream_builder(
         audio=io.BytesIO(file.file.read()),
         task=task,
         vad_filter=vad_filter,
         language=None if language == "und" else language,
+        repetition_penalty=repetition_penalty,
     )
 
     # special function for streaming response (OpenAI API does not have this)
@@ -327,7 +323,7 @@ async def transcription(
             media_type="text/event-stream",
         )
     elif response_format == "json":
-        return build_json_result(generator)
+        return build_json_result(generator, info)
     elif response_format == "text":
         return StreamingResponse(text_writer(generator), media_type="text/plain")
     elif response_format == "tsv":