sarvamai 0.1.20a2__tar.gz → 0.1.21__tar.gz

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.20a2
+Version: 0.1.21
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/pyproject.toml

@@ -3,7 +3,7 @@ name = "sarvamai"
 
 [tool.poetry]
 name = "sarvamai"
-version = "0.1.20a2"
+version = "0.1.21"
 description = ""
 readme = "README.md"
 authors = []

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/__init__.py

@@ -189,14 +189,12 @@ from .speech_to_text_streaming import (
     SpeechToTextStreamingHighVadSensitivity,
     SpeechToTextStreamingInputAudioCodec,
     SpeechToTextStreamingLanguageCode,
-    SpeechToTextStreamingModel,
     SpeechToTextStreamingVadSignals,
 )
 from .speech_to_text_translate_streaming import (
     SpeechToTextTranslateStreamingFlushSignal,
     SpeechToTextTranslateStreamingHighVadSensitivity,
     SpeechToTextTranslateStreamingInputAudioCodec,
-    SpeechToTextTranslateStreamingModel,
     SpeechToTextTranslateStreamingVadSignals,
 )
 from .text_to_speech_streaming import TextToSpeechStreamingSendCompletionEvent
@@ -322,7 +320,6 @@ __all__ = [
     "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingInputAudioCodec",
     "SpeechToTextStreamingLanguageCode",
-    "SpeechToTextStreamingModel",
     "SpeechToTextStreamingResponse",
     "SpeechToTextStreamingResponseParams",
     "SpeechToTextStreamingVadSignals",
@@ -339,7 +336,6 @@ __all__ = [
     "SpeechToTextTranslateStreamingFlushSignal",
     "SpeechToTextTranslateStreamingHighVadSensitivity",
     "SpeechToTextTranslateStreamingInputAudioCodec",
-    "SpeechToTextTranslateStreamingModel",
     "SpeechToTextTranslateStreamingResponse",
     "SpeechToTextTranslateStreamingResponseParams",
     "SpeechToTextTranslateStreamingVadSignals",

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/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/0.1.20a2",
+            "User-Agent": "sarvamai/0.1.21",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.20a2",
+            "X-Fern-SDK-Version": "0.1.21",
             **(self.get_custom_headers() or {}),
         }
         headers["api-subscription-key"] = self.api_subscription_key

sarvamai-0.1.21/src/sarvamai/play.py

@@ -0,0 +1,140 @@
+import base64
+import shutil
+import subprocess
+import wave
+from .types.text_to_speech_response import TextToSpeechResponse
+
+
+def is_installed(lib_name: str) -> bool:
+    lib = shutil.which(lib_name)
+    if lib is None:
+        return False
+    return True
+
+
+def play(
+    audio: TextToSpeechResponse, notebook: bool = False, use_ffmpeg: bool = True
+) -> None:
+    # Iterate through all audio chunks and concatenate them properly
+    combined_audio_data = b""
+
+    for i, audio_chunk in enumerate(audio.audios):
+        # Decode each base64 chunk to get raw WAV data
+        chunk_data = base64.b64decode(audio_chunk)
+
+        if i == 0:
+            # For the first chunk, keep the entire WAV file
+            combined_audio_data = chunk_data
+        else:
+            # For subsequent chunks, find the data chunk and append only audio data
+            data_pos = chunk_data.find(b"data")
+            if data_pos != -1:
+                # Skip the 'data' header (8 bytes: 'data' + size)
+                audio_data_start = data_pos + 8
+                combined_audio_data += chunk_data[audio_data_start:]
+
+    # Update the WAV header with the correct total file size
+    if len(audio.audios) > 1:
+        # Update the RIFF chunk size (bytes 4-7)
+        total_size = len(combined_audio_data) - 8
+        combined_audio_data = (
+            combined_audio_data[:4]
+            + total_size.to_bytes(4, "little")
+            + combined_audio_data[8:]
+        )
+
+        # Update the data chunk size (find data chunk and update its size)
+        data_pos = combined_audio_data.find(b"data")
+        if data_pos != -1:
+            data_size = len(combined_audio_data) - data_pos - 8
+            combined_audio_data = (
+                combined_audio_data[: data_pos + 4]
+                + data_size.to_bytes(4, "little")
+                + combined_audio_data[data_pos + 8 :]
+            )
+
+    af_bytes = combined_audio_data
+    if notebook:
+        try:
+            from IPython.display import Audio, display  # type: ignore
+        except ModuleNotFoundError:
+            message = "`pip install ipython` required when `notebook=False` "
+            raise ValueError(message)
+
+        display(Audio(af_bytes, rate=22050, autoplay=True))
+    elif use_ffmpeg:
+        if not is_installed("ffplay"):
+            message = (
+                "ffplay from ffmpeg not found, necessary to play audio. "
+                "On mac you can install it with 'brew install ffmpeg'. "
+                "On linux and windows you can install it from "
+                "https://ffmpeg.org/"
+            )
+            raise ValueError(message)
+        args = ["ffplay", "-autoexit", "-", "-nodisp"]
+        proc = subprocess.Popen(
+            args=args,
+            stdout=subprocess.PIPE,
+            stdin=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+        )
+        out, err = proc.communicate(input=af_bytes)
+        proc.poll()
+    else:
+        try:
+            import io
+
+            import sounddevice as sd  # type: ignore
+            import soundfile as sf  # type: ignore
+        except ModuleNotFoundError:
+            message = (
+                "`pip install sounddevice soundfile` required when "
+                "`use_ffmpeg=False` "
+            )
+            raise ValueError(message)
+        sd.play(*sf.read(io.BytesIO(af_bytes)))
+        sd.wait()
+
+
+def save(audio: TextToSpeechResponse, filename: str) -> None:
+    if isinstance(audio.audios, list):
+        # Iterate through all audio chunks and concatenate them properly
+        combined_audio_data = b""
+
+        for i, audio_chunk in enumerate(audio.audios):
+            # Decode each base64 chunk to get raw WAV data
+            chunk_data = base64.b64decode(audio_chunk)
+
+            if i == 0:
+                # For the first chunk, keep the entire WAV file
+                combined_audio_data = chunk_data
+            else:
+                # For subsequent chunks, find the data chunk and append only audio data
+                data_pos = chunk_data.find(b"data")
+                if data_pos != -1:
+                    # Skip the 'data' header (8 bytes: 'data' + size)
+                    audio_data_start = data_pos + 8
+                    combined_audio_data += chunk_data[audio_data_start:]
+
+        # Update the WAV header with the correct total file size
+        if len(audio.audios) > 1:
+            # Update the RIFF chunk size (bytes 4-7)
+            total_size = len(combined_audio_data) - 8
+            combined_audio_data = (
+                combined_audio_data[:4]
+                + total_size.to_bytes(4, "little")
+                + combined_audio_data[8:]
+            )
+
+            # Update the data chunk size (find data chunk and update its size)
+            data_pos = combined_audio_data.find(b"data")
+            if data_pos != -1:
+                data_size = len(combined_audio_data) - data_pos - 8
+                combined_audio_data = (
+                    combined_audio_data[: data_pos + 4]
+                    + data_size.to_bytes(4, "little")
+                    + combined_audio_data[data_pos + 8 :]
+                )
+
+        with open(filename, "wb") as f:
+            f.write(combined_audio_data)

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/requests/audio_output_data.py

@@ -13,3 +13,8 @@ class AudioOutputDataParams(typing_extensions.TypedDict):
     """
     Base64-encoded audio data ready for playback or download
     """
+
+    request_id: typing_extensions.NotRequired[str]
+    """
+    Unique identifier for the request
+    """

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/requests/error_response_data.py

@@ -16,3 +16,8 @@ class ErrorResponseDataParams(typing_extensions.TypedDict):
     """
     Additional error details and context information
     """
+
+    request_id: typing_extensions.NotRequired[str]
+    """
+    Unique identifier for the request
+    """

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/speech_to_text_streaming/__init__.py

@@ -7,7 +7,6 @@ from .types import (
     SpeechToTextStreamingHighVadSensitivity,
     SpeechToTextStreamingInputAudioCodec,
     SpeechToTextStreamingLanguageCode,
-    SpeechToTextStreamingModel,
     SpeechToTextStreamingVadSignals,
 )
 
@@ -16,6 +15,5 @@ __all__ = [
     "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingInputAudioCodec",
     "SpeechToTextStreamingLanguageCode",
-    "SpeechToTextStreamingModel",
     "SpeechToTextStreamingVadSignals",
 ]

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/speech_to_text_streaming/client.py

@@ -15,7 +15,6 @@ from .types.speech_to_text_streaming_flush_signal import SpeechToTextStreamingFl
 from .types.speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .types.speech_to_text_streaming_input_audio_codec import SpeechToTextStreamingInputAudioCodec
 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
 
 try:
@@ -44,7 +43,7 @@ class SpeechToTextStreamingClient:
         self,
         *,
         language_code: SpeechToTextStreamingLanguageCode,
-        model: typing.Optional[SpeechToTextStreamingModel] = None,
+        model: typing.Optional[typing.Literal["saarika:v2.5"]] = None,
         input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
@@ -61,7 +60,7 @@ class SpeechToTextStreamingClient:
         language_code : SpeechToTextStreamingLanguageCode
             Language code for speech recognition
 
-        model : typing.Optional[SpeechToTextStreamingModel]
+        model : typing.Optional[typing.Literal["saarika:v2.5"]]
             Speech to text model to use
 
         input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
@@ -149,7 +148,7 @@ class AsyncSpeechToTextStreamingClient:
         self,
         *,
         language_code: SpeechToTextStreamingLanguageCode,
-        model: typing.Optional[SpeechToTextStreamingModel] = None,
+        model: typing.Optional[typing.Literal["saarika:v2.5"]] = None,
         input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
@@ -166,7 +165,7 @@ class AsyncSpeechToTextStreamingClient:
         language_code : SpeechToTextStreamingLanguageCode
             Language code for speech recognition
 
-        model : typing.Optional[SpeechToTextStreamingModel]
+        model : typing.Optional[typing.Literal["saarika:v2.5"]]
             Speech to text model to use
 
         input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/speech_to_text_streaming/raw_client.py

@@ -14,7 +14,6 @@ from .types.speech_to_text_streaming_flush_signal import SpeechToTextStreamingFl
 from .types.speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .types.speech_to_text_streaming_input_audio_codec import SpeechToTextStreamingInputAudioCodec
 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
 
 try:
@@ -32,7 +31,7 @@ class RawSpeechToTextStreamingClient:
         self,
         *,
         language_code: SpeechToTextStreamingLanguageCode,
-        model: typing.Optional[SpeechToTextStreamingModel] = None,
+        model: typing.Optional[typing.Literal["saarika:v2.5"]] = None,
         input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
@@ -49,7 +48,7 @@ class RawSpeechToTextStreamingClient:
         language_code : SpeechToTextStreamingLanguageCode
             Language code for speech recognition
 
-        model : typing.Optional[SpeechToTextStreamingModel]
+        model : typing.Optional[typing.Literal["saarika:v2.5"]]
             Speech to text model to use
 
         input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
@@ -126,7 +125,7 @@ class AsyncRawSpeechToTextStreamingClient:
         self,
         *,
         language_code: SpeechToTextStreamingLanguageCode,
-        model: typing.Optional[SpeechToTextStreamingModel] = None,
+        model: typing.Optional[typing.Literal["saarika:v2.5"]] = None,
         input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
@@ -143,7 +142,7 @@ class AsyncRawSpeechToTextStreamingClient:
         language_code : SpeechToTextStreamingLanguageCode
             Language code for speech recognition
 
-        model : typing.Optional[SpeechToTextStreamingModel]
+        model : typing.Optional[typing.Literal["saarika:v2.5"]]
             Speech to text model to use
 
         input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/speech_to_text_streaming/types/__init__.py

@@ -6,7 +6,6 @@ from .speech_to_text_streaming_flush_signal import SpeechToTextStreamingFlushSig
 from .speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .speech_to_text_streaming_input_audio_codec import SpeechToTextStreamingInputAudioCodec
 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__ = [
@@ -14,6 +13,5 @@ __all__ = [
     "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingInputAudioCodec",
     "SpeechToTextStreamingLanguageCode",
-    "SpeechToTextStreamingModel",
     "SpeechToTextStreamingVadSignals",
 ]

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/speech_to_text_translate_streaming/__init__.py

@@ -6,7 +6,6 @@ from .types import (
     SpeechToTextTranslateStreamingFlushSignal,
     SpeechToTextTranslateStreamingHighVadSensitivity,
     SpeechToTextTranslateStreamingInputAudioCodec,
-    SpeechToTextTranslateStreamingModel,
     SpeechToTextTranslateStreamingVadSignals,
 )
 
@@ -14,6 +13,5 @@ __all__ = [
     "SpeechToTextTranslateStreamingFlushSignal",
     "SpeechToTextTranslateStreamingHighVadSensitivity",
     "SpeechToTextTranslateStreamingInputAudioCodec",
-    "SpeechToTextTranslateStreamingModel",
     "SpeechToTextTranslateStreamingVadSignals",
 ]

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/speech_to_text_translate_streaming/client.py

@@ -16,7 +16,6 @@ from .types.speech_to_text_translate_streaming_high_vad_sensitivity import (
     SpeechToTextTranslateStreamingHighVadSensitivity,
 )
 from .types.speech_to_text_translate_streaming_input_audio_codec import SpeechToTextTranslateStreamingInputAudioCodec
-from .types.speech_to_text_translate_streaming_model import SpeechToTextTranslateStreamingModel
 from .types.speech_to_text_translate_streaming_vad_signals import SpeechToTextTranslateStreamingVadSignals
 
 try:
@@ -44,7 +43,7 @@ class SpeechToTextTranslateStreamingClient:
     def connect(
         self,
         *,
-        model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
+        model: typing.Optional[typing.Literal["saaras:v2.5"]] = None,
         input_audio_codec: typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
@@ -58,7 +57,7 @@ class SpeechToTextTranslateStreamingClient:
 
         Parameters
         ----------
-        model : typing.Optional[SpeechToTextTranslateStreamingModel]
+        model : typing.Optional[typing.Literal["saaras:v2.5"]]
             Speech to text model to use (defaults to "saaras:v2.5" if not specified)
 
         input_audio_codec : typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec]
@@ -143,7 +142,7 @@ class AsyncSpeechToTextTranslateStreamingClient:
     async def connect(
         self,
         *,
-        model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
+        model: typing.Optional[typing.Literal["saaras:v2.5"]] = None,
         input_audio_codec: typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
@@ -157,7 +156,7 @@ class AsyncSpeechToTextTranslateStreamingClient:
 
         Parameters
         ----------
-        model : typing.Optional[SpeechToTextTranslateStreamingModel]
+        model : typing.Optional[typing.Literal["saaras:v2.5"]]
             Speech to text model to use (defaults to "saaras:v2.5" if not specified)
 
         input_audio_codec : typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec]

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/speech_to_text_translate_streaming/raw_client.py

@@ -15,7 +15,6 @@ from .types.speech_to_text_translate_streaming_high_vad_sensitivity import (
     SpeechToTextTranslateStreamingHighVadSensitivity,
 )
 from .types.speech_to_text_translate_streaming_input_audio_codec import SpeechToTextTranslateStreamingInputAudioCodec
-from .types.speech_to_text_translate_streaming_model import SpeechToTextTranslateStreamingModel
 from .types.speech_to_text_translate_streaming_vad_signals import SpeechToTextTranslateStreamingVadSignals
 
 try:
@@ -32,7 +31,7 @@ class RawSpeechToTextTranslateStreamingClient:
     def connect(
         self,
         *,
-        model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
+        model: typing.Optional[typing.Literal["saaras:v2.5"]] = None,
         input_audio_codec: typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
@@ -46,7 +45,7 @@ class RawSpeechToTextTranslateStreamingClient:
 
         Parameters
         ----------
-        model : typing.Optional[SpeechToTextTranslateStreamingModel]
+        model : typing.Optional[typing.Literal["saaras:v2.5"]]
             Speech to text model to use (defaults to "saaras:v2.5" if not specified)
 
         input_audio_codec : typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec]
@@ -120,7 +119,7 @@ class AsyncRawSpeechToTextTranslateStreamingClient:
     async def connect(
         self,
         *,
-        model: typing.Optional[SpeechToTextTranslateStreamingModel] = None,
+        model: typing.Optional[typing.Literal["saaras:v2.5"]] = None,
         input_audio_codec: typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
@@ -134,7 +133,7 @@ class AsyncRawSpeechToTextTranslateStreamingClient:
 
         Parameters
         ----------
-        model : typing.Optional[SpeechToTextTranslateStreamingModel]
+        model : typing.Optional[typing.Literal["saaras:v2.5"]]
             Speech to text model to use (defaults to "saaras:v2.5" if not specified)
 
         input_audio_codec : typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec]

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/speech_to_text_translate_streaming/types/__init__.py

@@ -5,13 +5,11 @@
 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_input_audio_codec import SpeechToTextTranslateStreamingInputAudioCodec
-from .speech_to_text_translate_streaming_model import SpeechToTextTranslateStreamingModel
 from .speech_to_text_translate_streaming_vad_signals import SpeechToTextTranslateStreamingVadSignals
 
 __all__ = [
     "SpeechToTextTranslateStreamingFlushSignal",
     "SpeechToTextTranslateStreamingHighVadSensitivity",
     "SpeechToTextTranslateStreamingInputAudioCodec",
-    "SpeechToTextTranslateStreamingModel",
     "SpeechToTextTranslateStreamingVadSignals",
 ]

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/types/audio_output_data.py

@@ -17,6 +17,11 @@ class AudioOutputData(UniversalBaseModel):
     Base64-encoded audio data ready for playback or download
     """
 
+    request_id: typing.Optional[str] = pydantic.Field(default=None)
+    """
+    Unique identifier for the request
+    """
+
     if IS_PYDANTIC_V2:
         model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2
     else:

{sarvamai-0.1.20a2 → sarvamai-0.1.21}/src/sarvamai/types/error_response_data.py

@@ -18,6 +18,11 @@ class ErrorResponseData(UniversalBaseModel):
     Additional error details and context information
     """
 
+    request_id: typing.Optional[str] = pydantic.Field(default=None)
+    """
+    Unique identifier for the request
+    """
+
     if IS_PYDANTIC_V2:
         model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2
     else:

sarvamai-0.1.21/src/sarvamai/types/speech_to_text_model.py

@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+SpeechToTextModel = typing.Literal["saarika:v2.5"]

sarvamai-0.1.21/src/sarvamai/types/speech_to_text_translate_model.py

@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+SpeechToTextTranslateModel = typing.Literal["saaras:v2.5"]

sarvamai-0.1.20a2/src/sarvamai/play.py

@@ -1,72 +0,0 @@
-import base64
-import shutil
-import subprocess
-import wave
-from .types.text_to_speech_response import TextToSpeechResponse
-
-
-def is_installed(lib_name: str) -> bool:
-    lib = shutil.which(lib_name)
-    if lib is None:
-        return False
-    return True
-
-
-def play(
-    audio: TextToSpeechResponse, 
-    notebook: bool = False, 
-    use_ffmpeg: bool = True
-) -> None:
-    af_str = "".join(audio.audios)
-    af_bytes = base64.b64decode(af_str)
-    if notebook:
-        try:
-            from IPython.display import Audio, display  # type: ignore
-        except ModuleNotFoundError:
-            message = (
-                "`pip install ipython` required when `notebook=False` "
-            )
-            raise ValueError(message)
-
-        display(Audio(af_bytes, rate=22050, autoplay=True))
-    elif use_ffmpeg:
-        if not is_installed("ffplay"):
-            message = (
-                "ffplay from ffmpeg not found, necessary to play audio. "
-                "On mac you can install it with 'brew install ffmpeg'. "
-                "On linux and windows you can install it from "
-                "https://ffmpeg.org/"
-            )
-            raise ValueError(message)
-        args = ["ffplay", "-autoexit", "-", "-nodisp"]
-        proc = subprocess.Popen(
-            args=args,
-            stdout=subprocess.PIPE,
-            stdin=subprocess.PIPE,
-            stderr=subprocess.PIPE,
-        )
-        out, err = proc.communicate(input=af_bytes)
-        proc.poll()
-    else:
-        try:
-            import io
-
-            import sounddevice as sd  # type: ignore
-            import soundfile as sf  # type: ignore
-        except ModuleNotFoundError:
-            message = (
-                "`pip install sounddevice soundfile` required when "
-                "`use_ffmpeg=False` "
-            )
-            raise ValueError(message)
-        sd.play(*sf.read(io.BytesIO(af_bytes)))
-        sd.wait()
-
-
-def save(audio: TextToSpeechResponse, filename: str) -> None:
-    if isinstance(audio.audios, list):
-        combined_audio = "".join(audio.audios)
-        b64_file = base64.b64decode(combined_audio)
-
-        with open(filename, "wb") as f:
-            f.write(b64_file)

sarvamai-0.1.20a2/src/sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_model.py

@@ -1,5 +0,0 @@
-# This file was auto-generated by Fern from our API Definition.
-
-import typing
-
-SpeechToTextStreamingModel = typing.Union[typing.Literal["saarika:v2.5", "saarika:v2"], typing.Any]

sarvamai-0.1.20a2/src/sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_model.py

@@ -1,5 +0,0 @@
-# This file was auto-generated by Fern from our API Definition.
-
-import typing
-
-SpeechToTextTranslateStreamingModel = typing.Union[typing.Literal["saaras:v2.5", "saaras:v2"], typing.Any]

sarvamai-0.1.20a2/src/sarvamai/types/speech_to_text_model.py

@@ -1,7 +0,0 @@
-# This file was auto-generated by Fern from our API Definition.
-
-import typing
-
-SpeechToTextModel = typing.Union[
-    typing.Literal["saarika:v1", "saarika:v2", "saarika:v2.5", "saarika:flash"], typing.Any
-]

sarvamai-0.1.20a2/src/sarvamai/types/speech_to_text_translate_model.py

@@ -1,7 +0,0 @@
-# This file was auto-generated by Fern from our API Definition.
-
-import typing
-
-SpeechToTextTranslateModel = typing.Union[
-    typing.Literal["saaras:v1", "saaras:v2", "saaras:v2.5", "saaras:flash"], typing.Any
-]