background client mvp

faster_whisper_server/config.py

@@ -184,20 +184,30 @@ class Config(BaseSettings):
     """
 
     default_language: Language | None = None
+    """
+    Default language to use for transcription. If not set, the language will be detected automatically.
+    It is recommended to set this as it will improve the performance.
+    """
     default_response_format: ResponseFormat = ResponseFormat.JSON
     whisper: WhisperConfig = WhisperConfig()
     max_models: int = 1
+    """
+    Maximum number of models that can be loaded at a time.
+    """
     max_no_data_seconds: float = 1.0
     """
-    Max duration to for the next audio chunk before transcription is finilized and connection is closed.
+    Max duration to wait for the next audio chunk before transcription is finilized and connection is closed.
     """
     min_duration: float = 1.0
+    """
+    Minimum duration of an audio chunk that will be transcribed.
+    """
     word_timestamp_error_margin: float = 0.2
-    max_inactivity_seconds: float = 5.0
+    max_inactivity_seconds: float = 2.5
     """
     Max allowed audio duration without any speech being detected before transcription is finilized and connection is closed.
     """  # noqa: E501
-    inactivity_window_seconds: float = 10.0
+    inactivity_window_seconds: float = 5.0
     """
     Controls how many latest seconds of audio are being passed through VAD.
     Should be greater than `max_inactivity_seconds`

pyproject.toml

@@ -14,11 +14,21 @@ dependencies = [
     "httpx-sse",
     "httpx",
     "gradio",
-    "openai"
+    "openai",
 ]
 
 [project.optional-dependencies]
-dev = ["ruff==0.5.3", "pytest", "webvtt-py", "srt", "basedpyright==1.13.0", "pytest-xdist"]
+client = [
+    "keyboard",
+]
+dev = [
+    "ruff==0.5.3",
+    "pytest",
+    "webvtt-py",
+    "srt",
+    "basedpyright==1.13.0",
+    "pytest-xdist"
+]
 
 other = ["youtube-dl @ git+https://github.com/ytdl-org/youtube-dl.git@37cea84f775129ad715b9bcd617251c831fcc980", "aider-chat==0.39.0"]
 
@@ -58,7 +68,8 @@ ignore = [
     "T201", # print
     "TRY003",
     "W505",
-    "ISC001" # recommended to disable for formatting
+    "ISC001", # recommended to disable for formatting
+    "INP001",
 ]
 
 [tool.ruff.lint.isort]

requirements-all.txt

@@ -15,7 +15,6 @@ aiosignal==1.3.1
 altair==5.3.0
     # via
     #   aider-chat
-    #   gradio
     #   streamlit
 annotated-types==0.7.0
     # via
@@ -24,17 +23,17 @@ annotated-types==0.7.0
 anyio==4.4.0
     # via
     #   aider-chat
+    #   gradio
     #   httpx
     #   openai
     #   starlette
-    #   watchfiles
 attrs==23.2.0
     # via
     #   aider-chat
     #   aiohttp
     #   jsonschema
     #   referencing
-av==12.2.0
+av==12.3.0
     # via faster-whisper
 backoff==2.2.1
     # via aider-chat
@@ -91,21 +90,15 @@ distro==1.9.0
     # via
     #   aider-chat
     #   openai
-dnspython==2.6.1
-    # via email-validator
-email-validator==2.2.0
-    # via fastapi
 execnet==2.1.1
     # via pytest-xdist
-fastapi==0.111.1
+fastapi==0.112.2
     # via
     #   faster-whisper-server (pyproject.toml)
     #   gradio
-fastapi-cli==0.0.4
-    # via fastapi
 faster-whisper==1.0.3
     # via faster-whisper-server (pyproject.toml)
-ffmpy==0.3.2
+ffmpy==0.4.0
     # via gradio
 filelock==3.15.1
     # via
@@ -168,9 +161,9 @@ googleapis-common-protos==1.63.1
     #   aider-chat
     #   google-api-core
     #   grpcio-status
-gradio==4.38.1
+gradio==4.42.0
     # via faster-whisper-server (pyproject.toml)
-gradio-client==1.1.0
+gradio-client==1.3.0
     # via gradio
 greenlet==3.0.3
     # via
@@ -201,13 +194,10 @@ httplib2==0.22.0
     #   aider-chat
     #   google-api-python-client
     #   google-auth-httplib2
-httptools==0.6.1
-    # via uvicorn
 httpx==0.27.0
     # via
     #   faster-whisper-server (pyproject.toml)
     #   aider-chat
-    #   fastapi
     #   gradio
     #   gradio-client
     #   openai
@@ -227,23 +217,21 @@ idna==3.7
     # via
     #   aider-chat
     #   anyio
-    #   email-validator
     #   httpx
     #   requests
     #   yarl
 importlib-metadata==7.1.0
     # via
     #   aider-chat
     #   litellm
-importlib-resources==6.4.0
+importlib-resources==6.4.4
     # via gradio
 iniconfig==2.0.0
     # via pytest
 jinja2==3.1.4
     # via
     #   aider-chat
     #   altair
-    #   fastapi
     #   gradio
     #   litellm
     #   pydeck
@@ -255,6 +243,8 @@ jsonschema-specifications==2023.12.1
     # via
     #   aider-chat
     #   jsonschema
+keyboard==0.13.5
+    # via faster-whisper-server (pyproject.toml)
 kiwisolver==1.4.5
     # via matplotlib
 litellm==1.40.15
@@ -268,7 +258,7 @@ markupsafe==2.1.5
     #   aider-chat
     #   gradio
     #   jinja2
-matplotlib==3.9.1
+matplotlib==3.9.2
     # via gradio
 mccabe==0.7.0
     # via
@@ -287,7 +277,7 @@ multidict==6.0.5
     #   yarl
 networkx==3.2.1
     # via aider-chat
-nodejs-wheel-binaries==20.15.1
+nodejs-wheel-binaries==20.17.0
     # via basedpyright
 numpy==1.26.4
     # via
@@ -304,14 +294,14 @@ numpy==1.26.4
     #   pydeck
     #   scipy
     #   streamlit
-onnxruntime==1.18.1
+onnxruntime==1.19.0
     # via faster-whisper
 openai==1.34.0
     # via
     #   faster-whisper-server (pyproject.toml)
     #   aider-chat
     #   litellm
-orjson==3.10.6
+orjson==3.10.7
     # via gradio
 packaging==24.1
     # via
@@ -397,7 +387,7 @@ pydantic-core==2.18.4
     # via
     #   aider-chat
     #   pydantic
-pydantic-settings==2.3.4
+pydantic-settings==2.4.0
     # via faster-whisper-server (pyproject.toml)
 pydeck==0.9.1
     # via
@@ -424,7 +414,7 @@ pyparsing==3.1.2
     #   aider-chat
     #   httplib2
     #   matplotlib
-pytest==8.2.2
+pytest==8.3.2
     # via
     #   faster-whisper-server (pyproject.toml)
     #   pytest-xdist
@@ -440,11 +430,9 @@ python-dotenv==1.0.1
     #   aider-chat
     #   litellm
     #   pydantic-settings
-    #   uvicorn
 python-multipart==0.0.9
     # via
     #   faster-whisper-server (pyproject.toml)
-    #   fastapi
     #   gradio
 pytz==2024.1
     # via
@@ -456,7 +444,6 @@ pyyaml==6.0.1
     #   ctranslate2
     #   gradio
     #   huggingface-hub
-    #   uvicorn
 referencing==0.35.1
     # via
     #   aider-chat
@@ -496,7 +483,7 @@ scipy==1.13.1
     # via aider-chat
 semantic-version==2.10.0
     # via gradio
-setuptools==71.0.4
+setuptools==73.0.1
     # via ctranslate2
 shellingham==1.5.4
     # via typer
@@ -526,11 +513,11 @@ soupsieve==2.5
     #   beautifulsoup4
 srt==3.5.3
     # via faster-whisper-server (pyproject.toml)
-starlette==0.37.2
+starlette==0.38.2
     # via fastapi
 streamlit==1.35.0
     # via aider-chat
-sympy==1.13.1
+sympy==1.13.2
     # via onnxruntime
 tenacity==8.3.0
     # via
@@ -573,10 +560,8 @@ tree-sitter-languages==1.10.2
     # via
     #   aider-chat
     #   grep-ast
-typer==0.12.3
-    # via
-    #   fastapi-cli
-    #   gradio
+typer==0.12.5
+    # via gradio
 typing-extensions==4.12.2
     # via
     #   aider-chat
@@ -604,27 +589,20 @@ urllib3==2.2.1
     #   aider-chat
     #   gradio
     #   requests
-uvicorn==0.30.1
+uvicorn==0.30.6
     # via
     #   faster-whisper-server (pyproject.toml)
-    #   fastapi
     #   gradio
-uvloop==0.19.0
-    # via uvicorn
 watchdog==4.0.1
     # via
     #   aider-chat
     #   streamlit
-watchfiles==0.22.0
-    # via uvicorn
 wcwidth==0.2.13
     # via
     #   aider-chat
     #   prompt-toolkit
-websockets==11.0.3
-    # via
-    #   gradio-client
-    #   uvicorn
+websockets==12.0
+    # via gradio-client
 webvtt-py==0.5.1
     # via faster-whisper-server (pyproject.toml)
 yarl==1.9.4