sarvamai 0.1.19a1__py3-none-any.whl → 0.1.19a3__py3-none-any.whl

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/0.1.19a1",
+            "User-Agent": "sarvamai/0.1.19a3",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.19a1",
+            "X-Fern-SDK-Version": "0.1.19a3",
             **(self.get_custom_headers() or {}),
         }
         headers["api-subscription-key"] = self.api_subscription_key

sarvamai/requests/config_message.py

@@ -13,5 +13,5 @@ class ConfigMessageParams(typing_extensions.TypedDict):
 
     prompt: typing_extensions.NotRequired[str]
     """
-    Prompt for ASR model to improve transcription accuracy
+    Prompt for ASR model to improve transcription accuracy.
     """

sarvamai/speech_to_text_translate_streaming/socket_client.py

@@ -10,9 +10,14 @@ 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
+from ..types.speech_to_text_translate_streaming_response import (
+    SpeechToTextTranslateStreamingResponse,
+)
+from ..types.stt_flush_signal import SttFlushSignal
 
-SpeechToTextTranslateStreamingSocketClientResponse = typing.Union[SpeechToTextTranslateStreamingResponse]
+SpeechToTextTranslateStreamingSocketClientResponse = typing.Union[
+    SpeechToTextTranslateStreamingResponse
+]
 
 
 class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
@@ -38,7 +43,11 @@ 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
+                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:
@@ -46,7 +55,9 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    async def translate(self, audio: str, encoding: str = "audio/wav", sample_rate: int = 16000):
+    async def translate(
+        self, audio: str, encoding: str = "audio/wav", sample_rate: int = 16000
+    ):
         """
         Sends audio translation request to the server.
         :param audio: Base64 encoded audio data
@@ -54,7 +65,9 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         :param sample_rate (Optional): Audio sample rate in Hz (default is 16000)
         """
         return await self._send_speech_to_text_translate_streaming_audio_message(
-            message=AudioMessage(audio=AudioData(data=audio, encoding=encoding, sample_rate=sample_rate))
+            message=AudioMessage(
+                audio=AudioData(data=audio, encoding=encoding, sample_rate=sample_rate)
+            )
         )
 
     async def set_prompt(self, prompt: str) -> None:
@@ -66,15 +79,31 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         message = ConfigMessage(prompt=prompt)
         await self._send_config_message(message)
 
+    async def flush(self, reason: typing.Optional[str] = None) -> None:
+        """
+        Signal to flush the audio buffer and force finalize partial
+        transcriptions and translations. Use this to force processing of any
+        remaining audio that hasn't been transcribed and translated 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 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
+        return parse_obj_as(
+            SpeechToTextTranslateStreamingSocketClientResponse, data
+        )  # type: ignore
 
-    async def _send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+    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.
@@ -126,7 +155,11 @@ 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
+                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:
@@ -134,7 +167,9 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    def translate(self, audio: str, encoding: str = "audio/wav", sample_rate: int = 16000):
+    def translate(
+        self, audio: str, encoding: str = "audio/wav", sample_rate: int = 16000
+    ):
         """
         Sends audio translation request to the server.
         :param audio: Base64 encoded audio data
@@ -142,7 +177,9 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         :param sample_rate: Audio sample rate in Hz (default is 16000)
         """
         return self._send_speech_to_text_translate_streaming_audio_message(
-            message=AudioMessage(audio=AudioData(data=audio, encoding=encoding, sample_rate=sample_rate))
+            message=AudioMessage(
+                audio=AudioData(data=audio, encoding=encoding, sample_rate=sample_rate)
+            )
         )
 
     def set_prompt(self, prompt: str) -> None:
@@ -154,6 +191,18 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         message = ConfigMessage(prompt=prompt)
         self._send_config_message(message)
 
+    def flush(self, reason: typing.Optional[str] = None) -> None:
+        """
+        Signal to flush the audio buffer and force finalize partial transcriptions and translations.
+        Use this to force processing of any remaining audio that hasn't been
+        transcribed and translated 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) -> SpeechToTextTranslateStreamingSocketClientResponse:
         """
         Receive a message from the websocket connection.
@@ -169,7 +218,9 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         """
         self._send_model(message)
 
-    def _send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+    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.

sarvamai/types/config_message.py

@@ -14,7 +14,7 @@ class ConfigMessage(UniversalBaseModel):
 
     prompt: typing.Optional[str] = pydantic.Field(default=None)
     """
-    Prompt for ASR model to improve transcription accuracy
+    Prompt for ASR model to improve transcription accuracy.
     """
 
     if IS_PYDANTIC_V2:

{sarvamai-0.1.19a1.dist-info → sarvamai-0.1.19a3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.19a1
+Version: 0.1.19a3
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{sarvamai-0.1.19a1.dist-info → sarvamai-0.1.19a3.dist-info}/RECORD

@@ -5,7 +5,7 @@ sarvamai/chat/raw_client.py,sha256=A2kRuZcVWlJhyYCD7YKgqNkZEp3cYa1731KhRkhirU0,1
 sarvamai/client.py,sha256=J30X_os1lPf8Wml0KDFEf6p8VGHhgF_lf3nw1T2D3qo,8207
 sarvamai/core/__init__.py,sha256=YE2CtXeASe1RAbaI39twKWYKCuT4tW5is9HWHhJjR_g,1653
 sarvamai/core/api_error.py,sha256=44vPoTyWN59gonCIZMdzw7M1uspygiLnr3GNFOoVL2Q,614
-sarvamai/core/client_wrapper.py,sha256=zeaP6DuMbs7G4AqP-w_QTivx8CVyHsw-Dycgh5yhdiI,2570
+sarvamai/core/client_wrapper.py,sha256=2raeN4Q7mBnfXfOn3pAn2w8SM1aeo2yVzgS1q0nlEl8,2570
 sarvamai/core/datetime_utils.py,sha256=nBys2IsYrhPdszxGKCNRPSOCwa-5DWOHG95FB8G9PKo,1047
 sarvamai/core/events.py,sha256=HvKBdSoYcFetk7cgNXb7FxuY-FtY8NtUhZIN7mGVx8U,1159
 sarvamai/core/file.py,sha256=d4NNbX8XvXP32z8KpK2Xovv33nFfruIrpz0QWxlgpZk,2663
@@ -43,7 +43,7 @@ sarvamai/requests/chat_completion_request_user_message.py,sha256=IqYy7K-qF9oQ8AU
 sarvamai/requests/chat_completion_response_message.py,sha256=JFazj4zK-nj_wjdvNLDkcfIFxIlqw49Xf_P8o7d70aY,336
 sarvamai/requests/choice.py,sha256=uulX4MZUoThEMcD3a80o_3V5YpnpqN8DfPaNZWVz-1o,867
 sarvamai/requests/completion_usage.py,sha256=LbZV-RxcxKdCAYqhCiaRtSFF3VwMJq71A989Z1rm-I8,428
-sarvamai/requests/config_message.py,sha256=EpYioGvDhCXDMvGH7Q1F7448zJzoHmlkQ1owoNGbWAw,383
+sarvamai/requests/config_message.py,sha256=SeWZB5OEVzgL3aAtLehpGZzLdSEkxr-6RI3372j_Blg,384
 sarvamai/requests/configure_connection.py,sha256=a-foQtLxArL4CulvKEdeebbRqmS1GRmko3MZdnHVPEk,716
 sarvamai/requests/configure_connection_data.py,sha256=lRk_4rYPQLLlwS2HXjQ9Abxdf98_DuOOja-VkrIR44Q,3016
 sarvamai/requests/create_chat_completion_response.py,sha256=TqS9u5_WVWMok_NreT4TeOsLJQeybPkbJm45Q0Zxw30,857
@@ -108,7 +108,7 @@ sarvamai/speech_to_text_translate_job/raw_client.py,sha256=dAitbu2B9afPK6iT9zNjU
 sarvamai/speech_to_text_translate_streaming/__init__.py,sha256=sDeWLagKUTg34tL7vpNNwr2gX4AznEn6NAy9jY-Znf4,507
 sarvamai/speech_to_text_translate_streaming/client.py,sha256=41FAJekotqq08tDdxWqhT966B5Ofx9MuZgHAJ9xmwRQ,9137
 sarvamai/speech_to_text_translate_streaming/raw_client.py,sha256=P-p7pNJNLrgra4HoUx1vG34RcLADYfQEpoVMYe-aMa0,8246
-sarvamai/speech_to_text_translate_streaming/socket_client.py,sha256=Xy86G9Mv1kGLf8xwFo3OBf98vqGQ50iigUAoOaApLdI,7842
+sarvamai/speech_to_text_translate_streaming/socket_client.py,sha256=N0-mlrD1c6J34hb3YOjUrpvhGyCE62j_CeFnrdQGjUU,9233
 sarvamai/speech_to_text_translate_streaming/types/__init__.py,sha256=sK4Zv64ZcV33opFFaVd1kgnQnzjfXrSDhUwiEM85sZc,708
 sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_flush_signal.py,sha256=jkjvCGJ1pFKi3AOTkwMW-lo18WGgrgAhMpoe5P0AMzA,182
 sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_high_vad_sensitivity.py,sha256=r6MvTlkM0VEpb4dpnMHtINOZ-gYc22o0Fx_Xce2rjvo,189
@@ -139,7 +139,7 @@ sarvamai/types/chat_completion_request_user_message.py,sha256=J3WhlrfOfCCe7ugmJI
 sarvamai/types/chat_completion_response_message.py,sha256=wz935eBnCkSIl0I0qMxBuH4vAUCso1aHDGReMW1VHGE,744
 sarvamai/types/choice.py,sha256=uXBCsjWP9VK3XWQWZUeI4EnU10w0G9nAfKn2tJZvxko,1244
 sarvamai/types/completion_usage.py,sha256=xYQGlQUbKqsksuV73H-1ajjfT5M7w47eLfdWXSlrI5M,843
-sarvamai/types/config_message.py,sha256=sGrT-qYTRqLVfIo5nRUuRlqPtPVmiAkUAnaMtlmQYCU,778
+sarvamai/types/config_message.py,sha256=xLD2wZcXejYrmREMd-cn38da4hKfsNPKRtyAGCW0Zcg,779
 sarvamai/types/configure_connection.py,sha256=SnSNk02gQqP8e4VB4y88jjeFQ4ClpImjGLn2ANI8cZ4,1058
 sarvamai/types/configure_connection_data.py,sha256=uXC7fhNJWCpaKc2Vrz2DNpUxx1gN3PwAoDL-H8L401A,3537
 sarvamai/types/configure_connection_data_output_audio_bitrate.py,sha256=h00YvKLxsZC8L3__rH4XH53nN_GY40UElW1EjysCwUs,208
@@ -215,6 +215,6 @@ 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-0.1.19a1.dist-info/METADATA,sha256=x59PbuWdgTKrL5EAZfr0SlbC0jqQKP5BfP0fGNjKQPE,26753
-sarvamai-0.1.19a1.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
-sarvamai-0.1.19a1.dist-info/RECORD,,
+sarvamai-0.1.19a3.dist-info/METADATA,sha256=lwfEWnVj0lP9LBb6B4ixr2995S8ph7-LNOsTbtWgdPU,26753
+sarvamai-0.1.19a3.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
+sarvamai-0.1.19a3.dist-info/RECORD,,