sarvamai 0.1.5a12__py3-none-any.whl → 0.1.5a14__py3-none-any.whl

sarvamai/__init__.py

@@ -4,7 +4,6 @@
 
 from .types import (
     AudioData,
-    AudioDataEncoding,
     AudioMessage,
     ChatCompletionRequestAssistantMessage,
     ChatCompletionRequestMessage,
@@ -113,14 +112,22 @@ from .requests import (
     TranslationResponseParams,
     TransliterationResponseParams,
 )
-from .speech_to_text_streaming import SpeechToTextStreamingLanguageCode, SpeechToTextStreamingModel
-from .speech_to_text_translate_streaming import SpeechToTextTranslateStreamingModel
+from .speech_to_text_streaming import (
+    SpeechToTextStreamingHighVadSensitivity,
+    SpeechToTextStreamingLanguageCode,
+    SpeechToTextStreamingModel,
+    SpeechToTextStreamingVadSignals,
+)
+from .speech_to_text_translate_streaming import (
+    SpeechToTextTranslateStreamingHighVadSensitivity,
+    SpeechToTextTranslateStreamingModel,
+    SpeechToTextTranslateStreamingVadSignals,
+)
 from .version import __version__
 
 __all__ = [
     "AsyncSarvamAI",
     "AudioData",
-    "AudioDataEncoding",
     "AudioDataParams",
     "AudioMessage",
     "AudioMessageParams",
@@ -183,10 +190,12 @@ __all__ = [
     "SpeechToTextResponseData",
     "SpeechToTextResponseDataParams",
     "SpeechToTextResponseParams",
+    "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingLanguageCode",
     "SpeechToTextStreamingModel",
     "SpeechToTextStreamingResponse",
     "SpeechToTextStreamingResponseParams",
+    "SpeechToTextStreamingVadSignals",
     "SpeechToTextTranscriptionData",
     "SpeechToTextTranscriptionDataParams",
     "SpeechToTextTranslateLanguage",
@@ -195,9 +204,11 @@ __all__ = [
     "SpeechToTextTranslateResponseData",
     "SpeechToTextTranslateResponseDataParams",
     "SpeechToTextTranslateResponseParams",
+    "SpeechToTextTranslateStreamingHighVadSensitivity",
     "SpeechToTextTranslateStreamingModel",
     "SpeechToTextTranslateStreamingResponse",
     "SpeechToTextTranslateStreamingResponseParams",
+    "SpeechToTextTranslateStreamingVadSignals",
     "SpeechToTextTranslateTranscriptionData",
     "SpeechToTextTranslateTranscriptionDataParams",
     "SpokenFormNumeralsFormat",

sarvamai/core/client_wrapper.py

@@ -17,10 +17,10 @@ class BaseClientWrapper:
 
     def get_headers(self) -> typing.Dict[str, str]:
         headers: typing.Dict[str, str] = {
-            "User-Agent": "sarvamai/0.1.5a12",
+            "User-Agent": "sarvamai/0.1.5a14",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.5a12",
+            "X-Fern-SDK-Version": "0.1.5a14",
         }
         headers["api-subscription-key"] = self.api_subscription_key
         return headers

sarvamai/requests/audio_data.py

@@ -1,7 +1,8 @@
 # This file was auto-generated by Fern from our API Definition.
 
+import typing
+
 import typing_extensions
-from ..types.audio_data_encoding import AudioDataEncoding
 
 
 class AudioDataParams(typing_extensions.TypedDict):
@@ -15,7 +16,7 @@ class AudioDataParams(typing_extensions.TypedDict):
     Audio sample rate in Hz (16kHz preferred, 8kHz least preferred)
     """
 
-    encoding: AudioDataEncoding
+    encoding: typing.Literal["audio/wav"]
     """
     Audio encoding format
     """

sarvamai/speech_to_text_streaming/__init__.py

@@ -2,6 +2,16 @@
 
 # isort: skip_file
 
-from .types import SpeechToTextStreamingLanguageCode, SpeechToTextStreamingModel
+from .types import (
+    SpeechToTextStreamingHighVadSensitivity,
+    SpeechToTextStreamingLanguageCode,
+    SpeechToTextStreamingModel,
+    SpeechToTextStreamingVadSignals,
+)
 
-__all__ = ["SpeechToTextStreamingLanguageCode", "SpeechToTextStreamingModel"]
+__all__ = [
+    "SpeechToTextStreamingHighVadSensitivity",
+    "SpeechToTextStreamingLanguageCode",
+    "SpeechToTextStreamingModel",
+    "SpeechToTextStreamingVadSignals",
+]

sarvamai/speech_to_text_streaming/client.py

@@ -11,8 +11,10 @@ from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from .raw_client import AsyncRawSpeechToTextStreamingClient, RawSpeechToTextStreamingClient
 from .socket_client import AsyncSpeechToTextStreamingSocketClient, SpeechToTextStreamingSocketClient
+from .types.speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .types.speech_to_text_streaming_language_code import SpeechToTextStreamingLanguageCode
 from .types.speech_to_text_streaming_model import SpeechToTextStreamingModel
+from .types.speech_to_text_streaming_vad_signals import SpeechToTextStreamingVadSignals
 
 
 class SpeechToTextStreamingClient:
@@ -36,8 +38,8 @@ class SpeechToTextStreamingClient:
         *,
         language_code: SpeechToTextStreamingLanguageCode,
         model: typing.Optional[SpeechToTextStreamingModel] = None,
-        high_vad_sensitivity: typing.Optional[str] = None,
-        vad_signals: typing.Optional[str] = None,
+        high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
+        vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextStreamingSocketClient]:
@@ -52,10 +54,10 @@ class SpeechToTextStreamingClient:
         model : typing.Optional[SpeechToTextStreamingModel]
             Speech to text model to use
 
-        high_vad_sensitivity : typing.Optional[str]
+        high_vad_sensitivity : typing.Optional[SpeechToTextStreamingHighVadSensitivity]
             Enable high VAD (Voice Activity Detection) sensitivity
 
-        vad_signals : typing.Optional[str]
+        vad_signals : typing.Optional[SpeechToTextStreamingVadSignals]
             Enable VAD signals in response
 
         api_subscription_key : typing.Optional[str]
@@ -123,8 +125,8 @@ class AsyncSpeechToTextStreamingClient:
         *,
         language_code: SpeechToTextStreamingLanguageCode,
         model: typing.Optional[SpeechToTextStreamingModel] = None,
-        high_vad_sensitivity: typing.Optional[str] = None,
-        vad_signals: typing.Optional[str] = None,
+        high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
+        vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextStreamingSocketClient]:
@@ -139,10 +141,10 @@ class AsyncSpeechToTextStreamingClient:
         model : typing.Optional[SpeechToTextStreamingModel]
             Speech to text model to use
 
-        high_vad_sensitivity : typing.Optional[str]
+        high_vad_sensitivity : typing.Optional[SpeechToTextStreamingHighVadSensitivity]
             Enable high VAD (Voice Activity Detection) sensitivity
 
-        vad_signals : typing.Optional[str]
+        vad_signals : typing.Optional[SpeechToTextStreamingVadSignals]
             Enable VAD signals in response
 
         api_subscription_key : typing.Optional[str]

sarvamai/speech_to_text_streaming/raw_client.py

@@ -10,8 +10,10 @@ from ..core.api_error import ApiError
 from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from .socket_client import AsyncSpeechToTextStreamingSocketClient, SpeechToTextStreamingSocketClient
+from .types.speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .types.speech_to_text_streaming_language_code import SpeechToTextStreamingLanguageCode
 from .types.speech_to_text_streaming_model import SpeechToTextStreamingModel
+from .types.speech_to_text_streaming_vad_signals import SpeechToTextStreamingVadSignals
 
 
 class RawSpeechToTextStreamingClient:
@@ -24,8 +26,8 @@ class RawSpeechToTextStreamingClient:
         *,
         language_code: SpeechToTextStreamingLanguageCode,
         model: typing.Optional[SpeechToTextStreamingModel] = None,
-        high_vad_sensitivity: typing.Optional[str] = None,
-        vad_signals: typing.Optional[str] = None,
+        high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
+        vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextStreamingSocketClient]:
@@ -40,10 +42,10 @@ class RawSpeechToTextStreamingClient:
         model : typing.Optional[SpeechToTextStreamingModel]
             Speech to text model to use
 
-        high_vad_sensitivity : typing.Optional[str]
+        high_vad_sensitivity : typing.Optional[SpeechToTextStreamingHighVadSensitivity]
             Enable high VAD (Voice Activity Detection) sensitivity
 
-        vad_signals : typing.Optional[str]
+        vad_signals : typing.Optional[SpeechToTextStreamingVadSignals]
             Enable VAD signals in response
 
         api_subscription_key : typing.Optional[str]
@@ -100,8 +102,8 @@ class AsyncRawSpeechToTextStreamingClient:
         *,
         language_code: SpeechToTextStreamingLanguageCode,
         model: typing.Optional[SpeechToTextStreamingModel] = None,
-        high_vad_sensitivity: typing.Optional[str] = None,
-        vad_signals: typing.Optional[str] = None,
+        high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
+        vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextStreamingSocketClient]:
@@ -116,10 +118,10 @@ class AsyncRawSpeechToTextStreamingClient:
         model : typing.Optional[SpeechToTextStreamingModel]
             Speech to text model to use
 
-        high_vad_sensitivity : typing.Optional[str]
+        high_vad_sensitivity : typing.Optional[SpeechToTextStreamingHighVadSensitivity]
             Enable high VAD (Voice Activity Detection) sensitivity
 
-        vad_signals : typing.Optional[str]
+        vad_signals : typing.Optional[SpeechToTextStreamingVadSignals]
             Enable VAD signals in response
 
         api_subscription_key : typing.Optional[str]

sarvamai/speech_to_text_streaming/socket_client.py

@@ -7,6 +7,7 @@ import websockets
 import websockets.sync.connection as websockets_sync_connection
 from ..core.events import EventEmitterMixin, EventType
 from ..core.pydantic_utilities import parse_obj_as
+from ..types.audio_data import AudioData
 from ..types.audio_message import AudioMessage
 from ..types.speech_to_text_streaming_response import SpeechToTextStreamingResponse
 
@@ -20,6 +21,7 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
 
     async def __aiter__(self):
         async for message in self._websocket:
+            message = json.loads(message) if isinstance(message, str) else message
             yield parse_obj_as(SpeechToTextStreamingSocketClientResponse, message)  # type: ignore
 
     async def start_listening(self):
@@ -35,6 +37,7 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
         self._emit(EventType.OPEN, None)
         try:
             async for raw_message in self._websocket:
+                raw_message = json.loads(raw_message) if isinstance(raw_message, str) else raw_message
                 parsed = parse_obj_as(SpeechToTextStreamingSocketClientResponse, raw_message)  # type: ignore
                 self._emit(EventType.MESSAGE, parsed)
         except websockets.WebSocketException as exc:
@@ -42,7 +45,19 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    async def send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
+    async def transcribe(self, audio: str, encoding="audio/wav", sample_rate=16000):
+        """
+        Sends transcription request to the server.
+        :param audio: Base64 encoded audio data
+        :param encoding: Audio encoding format (default is "audio/wav")
+        :param sample_rate: Audio sample rate in Hz (default is 16000)
+        """
+
+        return await self._send_speech_to_text_streaming_audio_message(
+            message=AudioMessage(audio=AudioData(data=audio, sample_rate=sample_rate, encoding=encoding))
+        )
+
+    async def _send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
         """
         Send a message to the websocket connection.
         The message will be sent as a AudioMessage.
@@ -54,6 +69,7 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
         Receive a message from the websocket connection.
         """
         data = await self._websocket.recv()
+        data = json.loads(data) if isinstance(data, str) else data
         return parse_obj_as(SpeechToTextStreamingSocketClientResponse, data)  # type: ignore
 
     async def _send(self, data: typing.Any) -> None:
@@ -78,6 +94,7 @@ class SpeechToTextStreamingSocketClient(EventEmitterMixin):
 
     def __iter__(self):
         for message in self._websocket:
+            message = json.loads(message) if isinstance(message, str) else message
             yield parse_obj_as(SpeechToTextStreamingSocketClientResponse, message)  # type: ignore
 
     def start_listening(self):
@@ -93,6 +110,7 @@ class SpeechToTextStreamingSocketClient(EventEmitterMixin):
         self._emit(EventType.OPEN, None)
         try:
             for raw_message in self._websocket:
+                raw_message = json.loads(raw_message) if isinstance(raw_message, str) else raw_message
                 parsed = parse_obj_as(SpeechToTextStreamingSocketClientResponse, raw_message)  # type: ignore
                 self._emit(EventType.MESSAGE, parsed)
         except websockets.WebSocketException as exc:
@@ -100,20 +118,32 @@ class SpeechToTextStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    def send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
+    def transcribe(self, audio: str, encoding="audio/wav", sample_rate=16000) -> None:
         """
-        Send a message to the websocket connection.
-        The message will be sent as a AudioMessage.
+        Sends transcription request to the server.
+        :param audio: Base64 encoded audio data
+        :param encoding (Optional): Audio encoding format (default is "audio/wav")
+        :param sample_rate (Optional): Audio sample rate in Hz (default is 16000)
         """
-        self._send_model(message)
+        return self._send_speech_to_text_streaming_audio_message(
+            message=AudioMessage(audio=AudioData(data=audio, sample_rate=sample_rate, encoding=encoding))
+        )
 
     def recv(self) -> SpeechToTextStreamingSocketClientResponse:
         """
         Receive a message from the websocket connection.
         """
         data = self._websocket.recv()
+        data = json.loads(data) if isinstance(data, str) else data
         return parse_obj_as(SpeechToTextStreamingSocketClientResponse, data)  # type: ignore
 
+    def _send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a AudioMessage.
+        """
+        self._send_model(message)
+
     def _send(self, data: typing.Any) -> None:
         """
         Send a message to the websocket connection.

sarvamai/speech_to_text_streaming/types/__init__.py

@@ -2,7 +2,14 @@
 
 # isort: skip_file
 
+from .speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .speech_to_text_streaming_language_code import SpeechToTextStreamingLanguageCode
 from .speech_to_text_streaming_model import SpeechToTextStreamingModel
+from .speech_to_text_streaming_vad_signals import SpeechToTextStreamingVadSignals
 
-__all__ = ["SpeechToTextStreamingLanguageCode", "SpeechToTextStreamingModel"]
+__all__ = [
+    "SpeechToTextStreamingHighVadSensitivity",
+    "SpeechToTextStreamingLanguageCode",
+    "SpeechToTextStreamingModel",
+    "SpeechToTextStreamingVadSignals",
+]

sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_high_vad_sensitivity.py

@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+SpeechToTextStreamingHighVadSensitivity = typing.Union[typing.Literal["true", "false"], typing.Any]

sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_vad_signals.py

@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+SpeechToTextStreamingVadSignals = typing.Union[typing.Literal["true", "false"], typing.Any]

sarvamai/speech_to_text_translate_streaming/__init__.py

@@ -2,6 +2,14 @@
 
 # isort: skip_file
 
-from .types import SpeechToTextTranslateStreamingModel
+from .types import (
+    SpeechToTextTranslateStreamingHighVadSensitivity,
+    SpeechToTextTranslateStreamingModel,
+    SpeechToTextTranslateStreamingVadSignals,
+)
 
-__all__ = ["SpeechToTextTranslateStreamingModel"]
+__all__ = [
+    "SpeechToTextTranslateStreamingHighVadSensitivity",
+    "SpeechToTextTranslateStreamingModel",
+    "SpeechToTextTranslateStreamingVadSignals",
+]

sarvamai/speech_to_text_translate_streaming/client.py

@@ -11,7 +11,11 @@ from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from .raw_client import AsyncRawSpeechToTextTranslateStreamingClient, RawSpeechToTextTranslateStreamingClient
 from .socket_client import AsyncSpeechToTextTranslateStreamingSocketClient, SpeechToTextTranslateStreamingSocketClient
+from .types.speech_to_text_translate_streaming_high_vad_sensitivity import (
+    SpeechToTextTranslateStreamingHighVadSensitivity,
+)
 from .types.speech_to_text_translate_streaming_model import SpeechToTextTranslateStreamingModel
+from .types.speech_to_text_translate_streaming_vad_signals import SpeechToTextTranslateStreamingVadSignals
 
 
 class SpeechToTextTranslateStreamingClient:
@@ -34,8 +38,8 @@ class SpeechToTextTranslateStreamingClient:
         self,
         *,
         model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
-        high_vad_sensitivity: typing.Optional[str] = None,
-        vad_signals: typing.Optional[str] = None,
+        high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
+        vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextTranslateStreamingSocketClient]:
@@ -47,10 +51,10 @@ class SpeechToTextTranslateStreamingClient:
         model : typing.Optional[SpeechToTextTranslateStreamingModel]
             Speech to text model to use (defaults to "saaras:v2.5" if not specified)
 
-        high_vad_sensitivity : typing.Optional[str]
+        high_vad_sensitivity : typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity]
             Enable high VAD (Voice Activity Detection) sensitivity
 
-        vad_signals : typing.Optional[str]
+        vad_signals : typing.Optional[SpeechToTextTranslateStreamingVadSignals]
             Enable VAD signals in response
 
         api_subscription_key : typing.Optional[str]
@@ -115,8 +119,8 @@ class AsyncSpeechToTextTranslateStreamingClient:
         self,
         *,
         model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
-        high_vad_sensitivity: typing.Optional[str] = None,
-        vad_signals: typing.Optional[str] = None,
+        high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
+        vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextTranslateStreamingSocketClient]:
@@ -128,10 +132,10 @@ class AsyncSpeechToTextTranslateStreamingClient:
         model : typing.Optional[SpeechToTextTranslateStreamingModel]
             Speech to text model to use (defaults to "saaras:v2.5" if not specified)
 
-        high_vad_sensitivity : typing.Optional[str]
+        high_vad_sensitivity : typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity]
             Enable high VAD (Voice Activity Detection) sensitivity
 
-        vad_signals : typing.Optional[str]
+        vad_signals : typing.Optional[SpeechToTextTranslateStreamingVadSignals]
             Enable VAD signals in response
 
         api_subscription_key : typing.Optional[str]

sarvamai/speech_to_text_translate_streaming/raw_client.py

@@ -10,7 +10,11 @@ from ..core.api_error import ApiError
 from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from .socket_client import AsyncSpeechToTextTranslateStreamingSocketClient, SpeechToTextTranslateStreamingSocketClient
+from .types.speech_to_text_translate_streaming_high_vad_sensitivity import (
+    SpeechToTextTranslateStreamingHighVadSensitivity,
+)
 from .types.speech_to_text_translate_streaming_model import SpeechToTextTranslateStreamingModel
+from .types.speech_to_text_translate_streaming_vad_signals import SpeechToTextTranslateStreamingVadSignals
 
 
 class RawSpeechToTextTranslateStreamingClient:
@@ -22,8 +26,8 @@ class RawSpeechToTextTranslateStreamingClient:
         self,
         *,
         model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
-        high_vad_sensitivity: typing.Optional[str] = None,
-        vad_signals: typing.Optional[str] = None,
+        high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
+        vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextTranslateStreamingSocketClient]:
@@ -35,10 +39,10 @@ class RawSpeechToTextTranslateStreamingClient:
         model : typing.Optional[SpeechToTextTranslateStreamingModel]
             Speech to text model to use (defaults to "saaras:v2.5" if not specified)
 
-        high_vad_sensitivity : typing.Optional[str]
+        high_vad_sensitivity : typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity]
             Enable high VAD (Voice Activity Detection) sensitivity
 
-        vad_signals : typing.Optional[str]
+        vad_signals : typing.Optional[SpeechToTextTranslateStreamingVadSignals]
             Enable VAD signals in response
 
         api_subscription_key : typing.Optional[str]
@@ -92,8 +96,8 @@ class AsyncRawSpeechToTextTranslateStreamingClient:
         self,
         *,
         model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
-        high_vad_sensitivity: typing.Optional[str] = None,
-        vad_signals: typing.Optional[str] = None,
+        high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
+        vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextTranslateStreamingSocketClient]:
@@ -105,10 +109,10 @@ class AsyncRawSpeechToTextTranslateStreamingClient:
         model : typing.Optional[SpeechToTextTranslateStreamingModel]
             Speech to text model to use (defaults to "saaras:v2.5" if not specified)
 
-        high_vad_sensitivity : typing.Optional[str]
+        high_vad_sensitivity : typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity]
             Enable high VAD (Voice Activity Detection) sensitivity
 
-        vad_signals : typing.Optional[str]
+        vad_signals : typing.Optional[SpeechToTextTranslateStreamingVadSignals]
             Enable VAD signals in response
 
         api_subscription_key : typing.Optional[str]

sarvamai/speech_to_text_translate_streaming/socket_client.py

@@ -7,6 +7,7 @@ import websockets
 import websockets.sync.connection as websockets_sync_connection
 from ..core.events import EventEmitterMixin, EventType
 from ..core.pydantic_utilities import parse_obj_as
+from ..types.audio_data import AudioData
 from ..types.audio_message import AudioMessage
 from ..types.config_message import ConfigMessage
 from ..types.speech_to_text_translate_streaming_response import SpeechToTextTranslateStreamingResponse
@@ -21,6 +22,7 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
 
     async def __aiter__(self):
         async for message in self._websocket:
+            message = json.loads(message) if isinstance(message, str) else message
             yield parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, message)  # type: ignore
 
     async def start_listening(self):
@@ -36,6 +38,7 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         self._emit(EventType.OPEN, None)
         try:
             async for raw_message in self._websocket:
+                raw_message = json.loads(raw_message) if isinstance(raw_message, str) else raw_message
                 parsed = parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, raw_message)  # type: ignore
                 self._emit(EventType.MESSAGE, parsed)
         except websockets.WebSocketException as exc:
@@ -43,27 +46,48 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    async def send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+    async def translate(self, audio: str, encoding: str = "audio/wav", sample_rate: int = 16000):
         """
-        Send a message to the websocket connection.
-        The message will be sent as a AudioMessage.
+        Sends audio translation request to the server.
+        :param audio: Base64 encoded audio data
+        :param encoding (Optional): Audio encoding format (default is "audio/wav")
+        :param sample_rate (Optional): Audio sample rate in Hz (default is 16000)
         """
-        await self._send_model(message)
+        return await self._send_speech_to_text_translate_streaming_audio_message(
+            message=AudioMessage(audio=AudioData(data=audio, encoding=encoding, sample_rate=sample_rate))
+        )
 
-    async def send_config_message(self, message: ConfigMessage) -> None:
+    async def set_prompt(self, prompt: str) -> None:
         """
-        Send a message to the websocket connection.
-        The message will be sent as a ConfigMessage.
+        Set the prompt for the translation.
+        The prompt is useful for detecting hotwords in the audio stream.
+        This will send a ConfigMessage to the websocket connection.
         """
-        await self._send_model(message)
+        message = ConfigMessage(prompt=prompt)
+        await self._send_config_message(message)
 
     async def recv(self) -> SpeechToTextTranslateStreamingSocketClientResponse:
         """
         Receive a message from the websocket connection.
         """
         data = await self._websocket.recv()
+        data = json.loads(data) if isinstance(data, str) else data
         return parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, data)  # type: ignore
 
+    async def _send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a AudioMessage.
+        """
+        await self._send_model(message)
+
+    async def _send_config_message(self, message: ConfigMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a ConfigMessage.
+        """
+        await self._send_model(message)
+
     async def _send(self, data: typing.Any) -> None:
         """
         Send a message to the websocket connection.
@@ -86,6 +110,7 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
 
     def __iter__(self):
         for message in self._websocket:
+            message = json.loads(message) if isinstance(message, str) else message
             yield parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, message)  # type: ignore
 
     def start_listening(self):
@@ -101,6 +126,7 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         self._emit(EventType.OPEN, None)
         try:
             for raw_message in self._websocket:
+                raw_message = json.loads(raw_message) if isinstance(raw_message, str) else raw_message
                 parsed = parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, raw_message)  # type: ignore
                 self._emit(EventType.MESSAGE, parsed)
         except websockets.WebSocketException as exc:
@@ -108,27 +134,48 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    def send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+    def translate(self, audio: str, encoding: str = "audio/wav", sample_rate: int = 16000):
         """
-        Send a message to the websocket connection.
-        The message will be sent as a AudioMessage.
+        Sends audio translation request to the server.
+        :param audio: Base64 encoded audio data
+        :param encoding: Audio encoding format (default is "audio/wav")
+        :param sample_rate: Audio sample rate in Hz (default is 16000)
         """
-        self._send_model(message)
+        return self._send_speech_to_text_translate_streaming_audio_message(
+            message=AudioMessage(audio=AudioData(data=audio, encoding=encoding, sample_rate=sample_rate))
+        )
 
-    def send_config_message(self, message: ConfigMessage) -> None:
+    def set_prompt(self, prompt: str) -> None:
         """
-        Send a message to the websocket connection.
-        The message will be sent as a ConfigMessage.
+        Set the prompt for the translation.
+        The prompt is useful for detecting hotwords in the audio stream.
+        This will send a ConfigMessage to the websocket connection.
         """
-        self._send_model(message)
+        message = ConfigMessage(prompt=prompt)
+        self._send_config_message(message)
 
     def recv(self) -> SpeechToTextTranslateStreamingSocketClientResponse:
         """
         Receive a message from the websocket connection.
         """
         data = self._websocket.recv()
+        data = json.loads(data) if isinstance(data, str) else data
         return parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, data)  # type: ignore
 
+    def _send_config_message(self, message: ConfigMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a ConfigMessage.
+        """
+        self._send_model(message)
+
+    def _send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a AudioMessage.
+        """
+        self._send_model(message)
+
     def _send(self, data: typing.Any) -> None:
         """
         Send a message to the websocket connection.

sarvamai/speech_to_text_translate_streaming/types/__init__.py

@@ -2,6 +2,12 @@
 
 # isort: skip_file
 
+from .speech_to_text_translate_streaming_high_vad_sensitivity import SpeechToTextTranslateStreamingHighVadSensitivity
 from .speech_to_text_translate_streaming_model import SpeechToTextTranslateStreamingModel
+from .speech_to_text_translate_streaming_vad_signals import SpeechToTextTranslateStreamingVadSignals
 
-__all__ = ["SpeechToTextTranslateStreamingModel"]
+__all__ = [
+    "SpeechToTextTranslateStreamingHighVadSensitivity",
+    "SpeechToTextTranslateStreamingModel",
+    "SpeechToTextTranslateStreamingVadSignals",
+]

sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_high_vad_sensitivity.py

@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+SpeechToTextTranslateStreamingHighVadSensitivity = typing.Union[typing.Literal["true", "false"], typing.Any]

sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_vad_signals.py

@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+SpeechToTextTranslateStreamingVadSignals = typing.Union[typing.Literal["true", "false"], typing.Any]

sarvamai/types/__init__.py

@@ -3,7 +3,6 @@
 # isort: skip_file
 
 from .audio_data import AudioData
-from .audio_data_encoding import AudioDataEncoding
 from .audio_message import AudioMessage
 from .chat_completion_request_assistant_message import ChatCompletionRequestAssistantMessage
 from .chat_completion_request_message import (
@@ -68,7 +67,6 @@ from .transliteration_response import TransliterationResponse
 
 __all__ = [
     "AudioData",
-    "AudioDataEncoding",
     "AudioMessage",
     "ChatCompletionRequestAssistantMessage",
     "ChatCompletionRequestMessage",

sarvamai/types/audio_data.py

@@ -4,7 +4,6 @@ import typing
 
 import pydantic
 from ..core.pydantic_utilities import IS_PYDANTIC_V2, UniversalBaseModel
-from .audio_data_encoding import AudioDataEncoding
 
 
 class AudioData(UniversalBaseModel):
@@ -18,7 +17,7 @@ class AudioData(UniversalBaseModel):
     Audio sample rate in Hz (16kHz preferred, 8kHz least preferred)
     """
 
-    encoding: AudioDataEncoding = pydantic.Field()
+    encoding: typing.Literal["audio/wav"] = pydantic.Field(default="audio/wav")
     """
     Audio encoding format
     """

{sarvamai-0.1.5a12.dist-info → sarvamai-0.1.5a14.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.5a12
+Version: 0.1.5a14
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{sarvamai-0.1.5a12.dist-info → sarvamai-0.1.5a14.dist-info}/RECORD

@@ -1,11 +1,11 @@
-sarvamai/__init__.py,sha256=rN99IDvheQJQijubWD47TaGNlx8Q4IEUZ5CHh8QrIuY,7165
+sarvamai/__init__.py,sha256=-XV5i09uHn5nMVvhydUPG84_lryvCgcEmOwzOgCEsMk,7511
 sarvamai/chat/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
 sarvamai/chat/client.py,sha256=xOSj83Gr6Q7eY2qUeATiuXYQqBqWqSCQlIEopK5fKus,11022
 sarvamai/chat/raw_client.py,sha256=A2kRuZcVWlJhyYCD7YKgqNkZEp3cYa1731KhRkhirU0,17885
 sarvamai/client.py,sha256=5-fW9679vlfqw6hQCtFIG0gnqd6tdz2W8BWx2ypFUHE,6855
 sarvamai/core/__init__.py,sha256=YE2CtXeASe1RAbaI39twKWYKCuT4tW5is9HWHhJjR_g,1653
 sarvamai/core/api_error.py,sha256=44vPoTyWN59gonCIZMdzw7M1uspygiLnr3GNFOoVL2Q,614
-sarvamai/core/client_wrapper.py,sha256=7S93aF0oS2Ui73y8XddO9WmXgzn3TGlnVjzJ9iRKLiA,2080
+sarvamai/core/client_wrapper.py,sha256=a9EClnhAIjINM2aTSKgrG-q3k7M8Z85hiQhB3X1KytE,2080
 sarvamai/core/datetime_utils.py,sha256=nBys2IsYrhPdszxGKCNRPSOCwa-5DWOHG95FB8G9PKo,1047
 sarvamai/core/events.py,sha256=j7VWXgMpOsjCXdzY22wIhI7Q-v5InZ4WchRzA88x_Sk,856
 sarvamai/core/file.py,sha256=d4NNbX8XvXP32z8KpK2Xovv33nFfruIrpz0QWxlgpZk,2663
@@ -29,7 +29,7 @@ sarvamai/errors/unprocessable_entity_error.py,sha256=JqxtzIhvjkpQDqbT9Q-go1n-gyv
 sarvamai/play.py,sha256=4fh86zy8g8IPU2O8yPBY7QxXQOivv_nWQvPQsOa1arw,2183
 sarvamai/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 sarvamai/requests/__init__.py,sha256=orINJGeCXuK0ep5LLDhoiGp0kCYICrfJiLC6S-JmYNw,3697
-sarvamai/requests/audio_data.py,sha256=zbGm4jK04s1l9d9gCM-5F9c7H1qzhxbi_0YjoOnKFg8,443
+sarvamai/requests/audio_data.py,sha256=QI3SK5aiAg2yJ-m3l9CxOkONnH3CCKMFCl9kAdMs19o,410
 sarvamai/requests/audio_message.py,sha256=ZBeogjGE6YFXXM-0g8zq9SoizDk21reR0YXSB-0fMjg,214
 sarvamai/requests/chat_completion_request_assistant_message.py,sha256=xI6nqqY2t4j56DGEAt2aasDnI7no_mxxCBk_ChxNQjg,247
 sarvamai/requests/chat_completion_request_message.py,sha256=B5tOPGNdSaMOJRl0k26uuXaqvpTrftiu-99CDDBTnSI,736
@@ -64,28 +64,31 @@ sarvamai/requests/transliteration_response.py,sha256=KqRkqnegLmt7LjdVxjRePX6Roqa
 sarvamai/speech_to_text/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
 sarvamai/speech_to_text/client.py,sha256=E76V1BZ236AN37cvm7c-NTRTnDleiYQ4_hKHE3wfevY,11435
 sarvamai/speech_to_text/raw_client.py,sha256=rencVGZ5cVX1eHhYoVMhcrlkqVmpI4aAGTrSC7kcYUI,25690
-sarvamai/speech_to_text_streaming/__init__.py,sha256=S4UmJOq-iqUhm1LoHBry0tAc-ojsyXhYyiPs0BxlYm8,245
-sarvamai/speech_to_text_streaming/client.py,sha256=OXBYbfRY7DwF70xxprp1xDyNrviMhdoLJePsjm-BbGM,7775
-sarvamai/speech_to_text_streaming/raw_client.py,sha256=oyCysLboisbmwKd8p_IIspg3GbxvGC5C3iXVSqgKYvg,6956
-sarvamai/speech_to_text_streaming/socket_client.py,sha256=w7rVywKQYkvQJZe-ydz5o5a0NssRRpQJT0bDBKt3GNQ,4771
-sarvamai/speech_to_text_streaming/types/__init__.py,sha256=EAmovJvXu5XZTWa079d3MoZMTO4au1SxjdjxrJfGJPM,321
+sarvamai/speech_to_text_streaming/__init__.py,sha256=q7QygMmZCHJ-4FMhhL_6XNV_dsqlIFRCO1iSxoyxaaY,437
+sarvamai/speech_to_text_streaming/client.py,sha256=WdkzZxKMdnQ2hHv9hzJlfSNggRJLKFljRiC7695Jcog,8224
+sarvamai/speech_to_text_streaming/raw_client.py,sha256=7zcgVw7CXA2TySkJKQkS-NdnJOMRudmV_m4NaigICN0,7405
+sarvamai/speech_to_text_streaming/socket_client.py,sha256=QI0vEjDvNTG9-EH99NolIuARYwD-P8Fms8rqKOD9HJQ,6419
+sarvamai/speech_to_text_streaming/types/__init__.py,sha256=hzEDbcyy6K0Q9-Zx5k5lxAHxIgeHnAoPpvTLrL13YT0,599
+sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_high_vad_sensitivity.py,sha256=OwPwffa8TkLPGMnOTn5S7d-HmV8QmN3B7fHz8I1-VT8,180
 sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_language_code.py,sha256=LxgEifmgWTCFZn9U-f-TWKxRPng3a2J26Zt526QrA0Y,267
 sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_model.py,sha256=b6F4ymgz4got6KVDqrweYvkET8itze63wUwWyjqDlO4,180
-sarvamai/speech_to_text_translate_streaming/__init__.py,sha256=TQyn8tpGwhx5iAVwgJzLP4Hxp0M2VXej2JCIESMiJxQ,191
-sarvamai/speech_to_text_translate_streaming/client.py,sha256=nd0_fBWydKbptgfITSQhIdq47Fm-aGyLRUL1qcIeb8o,7500
-sarvamai/speech_to_text_translate_streaming/raw_client.py,sha256=ShPaomtppKUhfgvSm4sYLXJu1h3Mg-o1ZMyFcx7KJGQ,6609
-sarvamai/speech_to_text_translate_streaming/socket_client.py,sha256=zfb6Y6XVqVjEt5TJVIZ3osUkMMf3tVZM4DPtAqy26ao,5441
-sarvamai/speech_to_text_translate_streaming/types/__init__.py,sha256=gYC5MLZAotSa5ChIJnyFV5dNP18wC8vcHVeRm9DAjvQ,226
+sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_vad_signals.py,sha256=8wiFOB7WDMbYCcMTYgNFJaIjEytYeXpJLwr_O_mH0TI,172
+sarvamai/speech_to_text_translate_streaming/__init__.py,sha256=_hmlce1Zs1grylysZhBUdtKfkaUROwVydtwz6l-1qqg,411
+sarvamai/speech_to_text_translate_streaming/client.py,sha256=TnHCcspbbYFaimcEk8km3QNrNkm8JlX7e2ydpeHL9EE,8068
+sarvamai/speech_to_text_translate_streaming/raw_client.py,sha256=_TlUro1NwWc1dAvC0SHcOoTnsnAqRRXxzxKHxw8BGTs,7177
+sarvamai/speech_to_text_translate_streaming/socket_client.py,sha256=Xy86G9Mv1kGLf8xwFo3OBf98vqGQ50iigUAoOaApLdI,7842
+sarvamai/speech_to_text_translate_streaming/types/__init__.py,sha256=zyKoGAbKW4d0-Zi56F0RfPUqtk_xUjWjF_RjTxbXXW4,556
+sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_high_vad_sensitivity.py,sha256=r6MvTlkM0VEpb4dpnMHtINOZ-gYc22o0Fx_Xce2rjvo,189
 sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_model.py,sha256=6B8VxkpJG_pNprCSctseDtJb_ULVdKrPaeENkQ6Jvjg,187
+sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_vad_signals.py,sha256=EV3xd9qyKMnMvA9rO-qFDDIac4b84roBu7n-maaPxG8,181
 sarvamai/text/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
 sarvamai/text/client.py,sha256=CMwDVtMsPor08e8F9a7yhHtgrdLzd1__samwm4S9CGM,30525
 sarvamai/text/raw_client.py,sha256=lQ7bV9aVqxjwEUHMPEZ4x0_Xs036_yFArMK9rnYT4ZI,48625
 sarvamai/text_to_speech/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
 sarvamai/text_to_speech/client.py,sha256=dyJlmHhVHUiXsi_5uh8Oged-ynN_NcDzD7JC_MByzV4,9164
 sarvamai/text_to_speech/raw_client.py,sha256=dwVPzx1kjXjuYGuNUjB1ibxHrYHXQJOJ6TnMGpDZOdo,15102
-sarvamai/types/__init__.py,sha256=BQZ6YuNB-iuXpHZZWoHzTKCTMA1cyK8PyUYEUkiVuU8,5302
-sarvamai/types/audio_data.py,sha256=yKkMcUbg9QGWhpfJSpWXEmwgun0K4NV7OZQiN_YkoVw,851
-sarvamai/types/audio_data_encoding.py,sha256=bfXb83yGdbLChm-XoN73PW-ak-iFPR24lVti07Ej00A,194
+sarvamai/types/__init__.py,sha256=OWVmsa_5dWjmCtLRsCpSlXrlTPd1ZLNAxqfAYvxAF80,5226
+sarvamai/types/audio_data.py,sha256=rgOukLkLNJ_HBBVE2g5dfEL2CWjRoGiMvCtpq0qTB1Y,829
 sarvamai/types/audio_message.py,sha256=sB4EgkWkWJzipYXobkmM9AYZTTZtCpg_ySKssUeznUE,560
 sarvamai/types/chat_completion_request_assistant_message.py,sha256=pFSONJ6CBsv3frcteid66SOKMkFwQ1UJs_e0XwwbKis,624
 sarvamai/types/chat_completion_request_message.py,sha256=4ZfaNvaphnPdJqYSeSXMZwBhkrAqBz3aob6j-4Cklho,1638
@@ -143,6 +146,6 @@ sarvamai/types/transliterate_mode.py,sha256=1jSEMlGcoLkWuk12TgoOpSgwifa4rThGKZ1h
 sarvamai/types/transliterate_source_language.py,sha256=bSY9wJszF0sg-Cgg6F-YcWC8ly1mIlj9rqa15-jBtx8,283
 sarvamai/types/transliteration_response.py,sha256=yt-lzTbDeJ_ZL4I8kQa6oESxA9ebeJJY7LfFHpdEsmM,815
 sarvamai/version.py,sha256=Qkp3Ee9YH-O9RTix90e0i7iNrFAGN-QDt2AFwGA4n8k,75
-sarvamai-0.1.5a12.dist-info/METADATA,sha256=L0swoNVGJyVhMfQaYDz0bSbxEHmx8vrzGGQS5hfmqJs,1038
-sarvamai-0.1.5a12.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
-sarvamai-0.1.5a12.dist-info/RECORD,,
+sarvamai-0.1.5a14.dist-info/METADATA,sha256=lNBU8MkacRV__lZfNW49jdBk1-rMfRkeGQA7o4fKb0o,1038
+sarvamai-0.1.5a14.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
+sarvamai-0.1.5a14.dist-info/RECORD,,

sarvamai/types/audio_data_encoding.py

@@ -1,5 +0,0 @@
-# This file was auto-generated by Fern from our API Definition.
-
-import typing
-
-AudioDataEncoding = typing.Union[typing.Literal["audio/wav", "audio/pcm", "audio/flac", "audio/mp3"], typing.Any]