sarvamai 0.1.18a0__tar.gz → 0.1.19a1__tar.gz

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.18a0
+Version: 0.1.19a1
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/pyproject.toml

@@ -3,7 +3,7 @@ name = "sarvamai"
 
 [tool.poetry]
 name = "sarvamai"
-version = "0.1.18a0"
+version = "0.1.19a1"
 description = ""
 readme = "README.md"
 authors = []

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/__init__.py

@@ -74,6 +74,7 @@ from .types import (
     SpokenFormNumeralsFormat,
     StopConfiguration,
     StorageContainerType,
+    SttFlushSignal,
     TaskDetailV1,
     TaskFileDetails,
     TaskState,
@@ -167,6 +168,7 @@ from .requests import (
     SpeechToTextTranslateStreamingResponseParams,
     SpeechToTextTranslateTranscriptionDataParams,
     StopConfigurationParams,
+    SttFlushSignalParams,
     TaskDetailV1Params,
     TaskFileDetailsParams,
     TextToSpeechResponseParams,
@@ -176,12 +178,14 @@ from .requests import (
     TransliterationResponseParams,
 )
 from .speech_to_text_streaming import (
+    SpeechToTextStreamingFlushSignal,
     SpeechToTextStreamingHighVadSensitivity,
     SpeechToTextStreamingLanguageCode,
     SpeechToTextStreamingModel,
     SpeechToTextStreamingVadSignals,
 )
 from .speech_to_text_translate_streaming import (
+    SpeechToTextTranslateStreamingFlushSignal,
     SpeechToTextTranslateStreamingHighVadSensitivity,
     SpeechToTextTranslateStreamingModel,
     SpeechToTextTranslateStreamingVadSignals,
@@ -297,6 +301,7 @@ __all__ = [
     "SpeechToTextResponseData",
     "SpeechToTextResponseDataParams",
     "SpeechToTextResponseParams",
+    "SpeechToTextStreamingFlushSignal",
     "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingLanguageCode",
     "SpeechToTextStreamingModel",
@@ -313,6 +318,7 @@ __all__ = [
     "SpeechToTextTranslateResponseData",
     "SpeechToTextTranslateResponseDataParams",
     "SpeechToTextTranslateResponseParams",
+    "SpeechToTextTranslateStreamingFlushSignal",
     "SpeechToTextTranslateStreamingHighVadSensitivity",
     "SpeechToTextTranslateStreamingModel",
     "SpeechToTextTranslateStreamingResponse",
@@ -324,6 +330,8 @@ __all__ = [
     "StopConfiguration",
     "StopConfigurationParams",
     "StorageContainerType",
+    "SttFlushSignal",
+    "SttFlushSignalParams",
     "TaskDetailV1",
     "TaskDetailV1Params",
     "TaskFileDetails",

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/core/client_wrapper.py

@@ -23,10 +23,10 @@ class BaseClientWrapper:
 
     def get_headers(self) -> typing.Dict[str, str]:
         headers: typing.Dict[str, str] = {
-            "User-Agent": "sarvamai/0.1.18a0",
+            "User-Agent": "sarvamai/0.1.19a1",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.18a0",
+            "X-Fern-SDK-Version": "0.1.19a1",
             **(self.get_custom_headers() or {}),
         }
         headers["api-subscription-key"] = self.api_subscription_key

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/requests/__init__.py

@@ -54,6 +54,7 @@ from .speech_to_text_translate_response_data import SpeechToTextTranslateRespons
 from .speech_to_text_translate_streaming_response import SpeechToTextTranslateStreamingResponseParams
 from .speech_to_text_translate_transcription_data import SpeechToTextTranslateTranscriptionDataParams
 from .stop_configuration import StopConfigurationParams
+from .stt_flush_signal import SttFlushSignalParams
 from .task_detail_v_1 import TaskDetailV1Params
 from .task_file_details import TaskFileDetailsParams
 from .text_to_speech_response import TextToSpeechResponseParams
@@ -113,6 +114,7 @@ __all__ = [
     "SpeechToTextTranslateStreamingResponseParams",
     "SpeechToTextTranslateTranscriptionDataParams",
     "StopConfigurationParams",
+    "SttFlushSignalParams",
     "TaskDetailV1Params",
     "TaskFileDetailsParams",
     "TextToSpeechResponseParams",

sarvamai-0.1.19a1/src/sarvamai/requests/stt_flush_signal.py

@@ -0,0 +1,21 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+import typing_extensions
+
+
+class SttFlushSignalParams(typing_extensions.TypedDict):
+    """
+    Signal to flush the audio buffer and force finalize partial transcriptions/translations
+    """
+
+    type: typing.Literal["flush"]
+    """
+    Type identifier for flush signal
+    """
+
+    reason: typing_extensions.NotRequired[str]
+    """
+    Optional reason for flushing (e.g., "end_of_segment", "manual_flush")
+    """

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_streaming/__init__.py

@@ -3,6 +3,7 @@
 # isort: skip_file
 
 from .types import (
+    SpeechToTextStreamingFlushSignal,
     SpeechToTextStreamingHighVadSensitivity,
     SpeechToTextStreamingLanguageCode,
     SpeechToTextStreamingModel,
@@ -10,6 +11,7 @@ from .types import (
 )
 
 __all__ = [
+    "SpeechToTextStreamingFlushSignal",
     "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingLanguageCode",
     "SpeechToTextStreamingModel",

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_streaming/client.py

@@ -11,6 +11,7 @@ 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_flush_signal import SpeechToTextStreamingFlushSignal
 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
@@ -45,6 +46,7 @@ class SpeechToTextStreamingClient:
         model: typing.Optional[SpeechToTextStreamingModel] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
+        flush_signal: typing.Optional[SpeechToTextStreamingFlushSignal] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextStreamingSocketClient]:
@@ -65,6 +67,9 @@ class SpeechToTextStreamingClient:
         vad_signals : typing.Optional[SpeechToTextStreamingVadSignals]
             Enable VAD signals in response
 
+        flush_signal : typing.Optional[SpeechToTextStreamingFlushSignal]
+            Signal to flush the audio buffer and finalize transcription
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -85,6 +90,8 @@ class SpeechToTextStreamingClient:
             query_params = query_params.add("high_vad_sensitivity", high_vad_sensitivity)
         if vad_signals is not None:
             query_params = query_params.add("vad_signals", vad_signals)
+        if flush_signal is not None:
+            query_params = query_params.add("flush_signal", flush_signal)
         ws_url = ws_url + f"?{query_params}"
         headers = self._raw_client._client_wrapper.get_headers()
         if api_subscription_key is not None:
@@ -132,6 +139,7 @@ class AsyncSpeechToTextStreamingClient:
         model: typing.Optional[SpeechToTextStreamingModel] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
+        flush_signal: typing.Optional[SpeechToTextStreamingFlushSignal] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextStreamingSocketClient]:
@@ -152,6 +160,9 @@ class AsyncSpeechToTextStreamingClient:
         vad_signals : typing.Optional[SpeechToTextStreamingVadSignals]
             Enable VAD signals in response
 
+        flush_signal : typing.Optional[SpeechToTextStreamingFlushSignal]
+            Signal to flush the audio buffer and finalize transcription
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -172,6 +183,8 @@ class AsyncSpeechToTextStreamingClient:
             query_params = query_params.add("high_vad_sensitivity", high_vad_sensitivity)
         if vad_signals is not None:
             query_params = query_params.add("vad_signals", vad_signals)
+        if flush_signal is not None:
+            query_params = query_params.add("flush_signal", flush_signal)
         ws_url = ws_url + f"?{query_params}"
         headers = self._raw_client._client_wrapper.get_headers()
         if api_subscription_key is not None:

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_streaming/raw_client.py

@@ -10,6 +10,7 @@ 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_flush_signal import SpeechToTextStreamingFlushSignal
 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
@@ -33,6 +34,7 @@ class RawSpeechToTextStreamingClient:
         model: typing.Optional[SpeechToTextStreamingModel] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
+        flush_signal: typing.Optional[SpeechToTextStreamingFlushSignal] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextStreamingSocketClient]:
@@ -53,6 +55,9 @@ class RawSpeechToTextStreamingClient:
         vad_signals : typing.Optional[SpeechToTextStreamingVadSignals]
             Enable VAD signals in response
 
+        flush_signal : typing.Optional[SpeechToTextStreamingFlushSignal]
+            Signal to flush the audio buffer and finalize transcription
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -73,6 +78,8 @@ class RawSpeechToTextStreamingClient:
             query_params = query_params.add("high_vad_sensitivity", high_vad_sensitivity)
         if vad_signals is not None:
             query_params = query_params.add("vad_signals", vad_signals)
+        if flush_signal is not None:
+            query_params = query_params.add("flush_signal", flush_signal)
         ws_url = ws_url + f"?{query_params}"
         headers = self._client_wrapper.get_headers()
         if api_subscription_key is not None:
@@ -109,6 +116,7 @@ class AsyncRawSpeechToTextStreamingClient:
         model: typing.Optional[SpeechToTextStreamingModel] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
+        flush_signal: typing.Optional[SpeechToTextStreamingFlushSignal] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextStreamingSocketClient]:
@@ -129,6 +137,9 @@ class AsyncRawSpeechToTextStreamingClient:
         vad_signals : typing.Optional[SpeechToTextStreamingVadSignals]
             Enable VAD signals in response
 
+        flush_signal : typing.Optional[SpeechToTextStreamingFlushSignal]
+            Signal to flush the audio buffer and finalize transcription
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -149,6 +160,8 @@ class AsyncRawSpeechToTextStreamingClient:
             query_params = query_params.add("high_vad_sensitivity", high_vad_sensitivity)
         if vad_signals is not None:
             query_params = query_params.add("vad_signals", vad_signals)
+        if flush_signal is not None:
+            query_params = query_params.add("flush_signal", flush_signal)
         ws_url = ws_url + f"?{query_params}"
         headers = self._client_wrapper.get_headers()
         if api_subscription_key is not None:

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_streaming/socket_client.py

@@ -9,7 +9,10 @@ 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
+from ..types.speech_to_text_streaming_response import (
+    SpeechToTextStreamingResponse,
+)
+from ..types.stt_flush_signal import SttFlushSignal
 
 SpeechToTextStreamingSocketClientResponse = typing.Union[SpeechToTextStreamingResponse]
 
@@ -22,7 +25,9 @@ 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
+            yield parse_obj_as(
+                SpeechToTextStreamingSocketClientResponse, message
+            )  # type: ignore
 
     async def start_listening(self):
         """
@@ -37,8 +42,14 @@ 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
+                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:
             self._emit(EventType.ERROR, exc)
@@ -54,10 +65,26 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
         """
 
         return await self._send_speech_to_text_streaming_audio_message(
-            message=AudioMessage(audio=AudioData(data=audio, sample_rate=sample_rate, encoding=encoding))
+            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:
+    async def flush(self, reason: typing.Optional[str] = None) -> None:
+        """
+        Signal to flush the audio buffer and force finalize partial transcriptions.
+        Use this to force processing of any remaining audio that hasn't been
+        transcribed yet.
+
+        :param reason: Optional reason for flushing (e.g., "end_of_segment",
+                      "manual_flush")
+        """
+        message = SttFlushSignal(reason=reason)
+        await self._send_model(message)
+
+    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.
@@ -70,7 +97,9 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
         """
         data = await self._websocket.recv()
         data = json.loads(data) if isinstance(data, str) else data
-        return parse_obj_as(SpeechToTextStreamingSocketClientResponse, data)  # type: ignore
+        return parse_obj_as(
+            SpeechToTextStreamingSocketClientResponse, data
+        )  # type: ignore
 
     async def _send(self, data: typing.Any) -> None:
         """
@@ -95,7 +124,9 @@ 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
+            yield parse_obj_as(
+                SpeechToTextStreamingSocketClientResponse, message
+            )  # type: ignore
 
     def start_listening(self):
         """
@@ -110,8 +141,14 @@ 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
+                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:
             self._emit(EventType.ERROR, exc)
@@ -126,18 +163,36 @@ class SpeechToTextStreamingSocketClient(EventEmitterMixin):
         :param sample_rate (Optional): Audio sample rate in Hz (default is 16000)
         """
         return self._send_speech_to_text_streaming_audio_message(
-            message=AudioMessage(audio=AudioData(data=audio, sample_rate=sample_rate, encoding=encoding))
+            message=AudioMessage(
+                audio=AudioData(data=audio, sample_rate=sample_rate, encoding=encoding)
+            )
         )
 
+    def flush(self, reason: typing.Optional[str] = None) -> None:
+        """
+        Signal to flush the audio buffer and force finalize partial transcriptions.
+        Use this to force processing of any remaining audio that hasn't been
+        transcribed yet.
+
+        :param reason: Optional reason for flushing (e.g., "end_of_segment",
+                      "manual_flush")
+        """
+        message = SttFlushSignal(reason=reason)
+        self._send_model(message)
+
     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
+        return parse_obj_as(
+            SpeechToTextStreamingSocketClientResponse, data
+        )  # type: ignore
 
-    def _send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
+    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.

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_streaming/types/__init__.py

@@ -2,12 +2,14 @@
 
 # isort: skip_file
 
+from .speech_to_text_streaming_flush_signal import SpeechToTextStreamingFlushSignal
 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__ = [
+    "SpeechToTextStreamingFlushSignal",
     "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingLanguageCode",
     "SpeechToTextStreamingModel",

sarvamai-0.1.19a1/src/sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_flush_signal.py

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

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_translate_streaming/__init__.py

@@ -3,12 +3,14 @@
 # isort: skip_file
 
 from .types import (
+    SpeechToTextTranslateStreamingFlushSignal,
     SpeechToTextTranslateStreamingHighVadSensitivity,
     SpeechToTextTranslateStreamingModel,
     SpeechToTextTranslateStreamingVadSignals,
 )
 
 __all__ = [
+    "SpeechToTextTranslateStreamingFlushSignal",
     "SpeechToTextTranslateStreamingHighVadSensitivity",
     "SpeechToTextTranslateStreamingModel",
     "SpeechToTextTranslateStreamingVadSignals",

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_translate_streaming/client.py

@@ -11,6 +11,7 @@ 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_flush_signal import SpeechToTextTranslateStreamingFlushSignal
 from .types.speech_to_text_translate_streaming_high_vad_sensitivity import (
     SpeechToTextTranslateStreamingHighVadSensitivity,
 )
@@ -45,6 +46,7 @@ class SpeechToTextTranslateStreamingClient:
         model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
+        flush_signal: typing.Optional[SpeechToTextTranslateStreamingFlushSignal] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextTranslateStreamingSocketClient]:
@@ -62,6 +64,9 @@ class SpeechToTextTranslateStreamingClient:
         vad_signals : typing.Optional[SpeechToTextTranslateStreamingVadSignals]
             Enable VAD signals in response
 
+        flush_signal : typing.Optional[SpeechToTextTranslateStreamingFlushSignal]
+            Signal to flush the audio buffer and finalize transcription and translation
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -80,6 +85,8 @@ class SpeechToTextTranslateStreamingClient:
             query_params = query_params.add("high_vad_sensitivity", high_vad_sensitivity)
         if vad_signals is not None:
             query_params = query_params.add("vad_signals", vad_signals)
+        if flush_signal is not None:
+            query_params = query_params.add("flush_signal", flush_signal)
         ws_url = ws_url + f"?{query_params}"
         headers = self._raw_client._client_wrapper.get_headers()
         if api_subscription_key is not None:
@@ -126,6 +133,7 @@ class AsyncSpeechToTextTranslateStreamingClient:
         model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
+        flush_signal: typing.Optional[SpeechToTextTranslateStreamingFlushSignal] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextTranslateStreamingSocketClient]:
@@ -143,6 +151,9 @@ class AsyncSpeechToTextTranslateStreamingClient:
         vad_signals : typing.Optional[SpeechToTextTranslateStreamingVadSignals]
             Enable VAD signals in response
 
+        flush_signal : typing.Optional[SpeechToTextTranslateStreamingFlushSignal]
+            Signal to flush the audio buffer and finalize transcription and translation
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -161,6 +172,8 @@ class AsyncSpeechToTextTranslateStreamingClient:
             query_params = query_params.add("high_vad_sensitivity", high_vad_sensitivity)
         if vad_signals is not None:
             query_params = query_params.add("vad_signals", vad_signals)
+        if flush_signal is not None:
+            query_params = query_params.add("flush_signal", flush_signal)
         ws_url = ws_url + f"?{query_params}"
         headers = self._raw_client._client_wrapper.get_headers()
         if api_subscription_key is not None:

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_translate_streaming/raw_client.py

@@ -10,6 +10,7 @@ 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_flush_signal import SpeechToTextTranslateStreamingFlushSignal
 from .types.speech_to_text_translate_streaming_high_vad_sensitivity import (
     SpeechToTextTranslateStreamingHighVadSensitivity,
 )
@@ -33,6 +34,7 @@ class RawSpeechToTextTranslateStreamingClient:
         model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
+        flush_signal: typing.Optional[SpeechToTextTranslateStreamingFlushSignal] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextTranslateStreamingSocketClient]:
@@ -50,6 +52,9 @@ class RawSpeechToTextTranslateStreamingClient:
         vad_signals : typing.Optional[SpeechToTextTranslateStreamingVadSignals]
             Enable VAD signals in response
 
+        flush_signal : typing.Optional[SpeechToTextTranslateStreamingFlushSignal]
+            Signal to flush the audio buffer and finalize transcription and translation
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -68,6 +73,8 @@ class RawSpeechToTextTranslateStreamingClient:
             query_params = query_params.add("high_vad_sensitivity", high_vad_sensitivity)
         if vad_signals is not None:
             query_params = query_params.add("vad_signals", vad_signals)
+        if flush_signal is not None:
+            query_params = query_params.add("flush_signal", flush_signal)
         ws_url = ws_url + f"?{query_params}"
         headers = self._client_wrapper.get_headers()
         if api_subscription_key is not None:
@@ -103,6 +110,7 @@ class AsyncRawSpeechToTextTranslateStreamingClient:
         model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
+        flush_signal: typing.Optional[SpeechToTextTranslateStreamingFlushSignal] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextTranslateStreamingSocketClient]:
@@ -120,6 +128,9 @@ class AsyncRawSpeechToTextTranslateStreamingClient:
         vad_signals : typing.Optional[SpeechToTextTranslateStreamingVadSignals]
             Enable VAD signals in response
 
+        flush_signal : typing.Optional[SpeechToTextTranslateStreamingFlushSignal]
+            Signal to flush the audio buffer and finalize transcription and translation
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -138,6 +149,8 @@ class AsyncRawSpeechToTextTranslateStreamingClient:
             query_params = query_params.add("high_vad_sensitivity", high_vad_sensitivity)
         if vad_signals is not None:
             query_params = query_params.add("vad_signals", vad_signals)
+        if flush_signal is not None:
+            query_params = query_params.add("flush_signal", flush_signal)
         ws_url = ws_url + f"?{query_params}"
         headers = self._client_wrapper.get_headers()
         if api_subscription_key is not None:

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/speech_to_text_translate_streaming/types/__init__.py

@@ -2,11 +2,13 @@
 
 # isort: skip_file
 
+from .speech_to_text_translate_streaming_flush_signal import SpeechToTextTranslateStreamingFlushSignal
 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__ = [
+    "SpeechToTextTranslateStreamingFlushSignal",
     "SpeechToTextTranslateStreamingHighVadSensitivity",
     "SpeechToTextTranslateStreamingModel",
     "SpeechToTextTranslateStreamingVadSignals",

sarvamai-0.1.19a1/src/sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_flush_signal.py

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

{sarvamai-0.1.18a0 → sarvamai-0.1.19a1}/src/sarvamai/types/__init__.py

@@ -75,6 +75,7 @@ from .speech_to_text_translate_transcription_data import SpeechToTextTranslateTr
 from .spoken_form_numerals_format import SpokenFormNumeralsFormat
 from .stop_configuration import StopConfiguration
 from .storage_container_type import StorageContainerType
+from .stt_flush_signal import SttFlushSignal
 from .task_detail_v_1 import TaskDetailV1
 from .task_file_details import TaskFileDetails
 from .task_state import TaskState
@@ -168,6 +169,7 @@ __all__ = [
     "SpokenFormNumeralsFormat",
     "StopConfiguration",
     "StorageContainerType",
+    "SttFlushSignal",
     "TaskDetailV1",
     "TaskFileDetails",
     "TaskState",

sarvamai-0.1.19a1/src/sarvamai/types/stt_flush_signal.py

@@ -0,0 +1,31 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+import pydantic
+from ..core.pydantic_utilities import IS_PYDANTIC_V2, UniversalBaseModel
+
+
+class SttFlushSignal(UniversalBaseModel):
+    """
+    Signal to flush the audio buffer and force finalize partial transcriptions/translations
+    """
+
+    type: typing.Literal["flush"] = pydantic.Field(default="flush")
+    """
+    Type identifier for flush signal
+    """
+
+    reason: typing.Optional[str] = pydantic.Field(default=None)
+    """
+    Optional reason for flushing (e.g., "end_of_segment", "manual_flush")
+    """
+
+    if IS_PYDANTIC_V2:
+        model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2
+    else:
+
+        class Config:
+            frozen = True
+            smart_union = True
+            extra = pydantic.Extra.allow