sarvamai 0.1.23a3__py3-none-any.whl → 0.1.23a5__py3-none-any.whl

sarvamai/speech_to_text_job/client.py

@@ -12,6 +12,7 @@ from ..types.files_upload_response import FilesUploadResponse
 from ..types.job_status_v_1_response import JobStatusV1Response
 from ..types.speech_to_text_model import SpeechToTextModel
 from ..types.speech_to_text_language import SpeechToTextLanguage
+from ..types.mode import Mode
 from .raw_client import AsyncRawSpeechToTextJobClient, RawSpeechToTextJobClient
 from .job import AsyncSpeechToTextJob, SpeechToTextJob
 
@@ -72,7 +73,9 @@ class SpeechToTextJobClient:
         )
         """
         _response = self._raw_client.initialise(
-            job_parameters=job_parameters, callback=callback, request_options=request_options
+            job_parameters=job_parameters,
+            callback=callback,
+            request_options=request_options,
         )
         return _response.data
 
@@ -145,11 +148,17 @@ class SpeechToTextJobClient:
             job_id="job_id",
         )
         """
-        _response = self._raw_client.start(job_id, ptu_id=ptu_id, request_options=request_options)
+        _response = self._raw_client.start(
+            job_id, ptu_id=ptu_id, request_options=request_options
+        )
         return _response.data
 
     def get_upload_links(
-        self, *, job_id: str, files: typing.Sequence[str], request_options: typing.Optional[RequestOptions] = None
+        self,
+        *,
+        job_id: str,
+        files: typing.Sequence[str],
+        request_options: typing.Optional[RequestOptions] = None,
     ) -> FilesUploadResponse:
         """
         Start a speech to text bulk job V1
@@ -180,11 +189,17 @@ class SpeechToTextJobClient:
             files=["files"],
         )
         """
-        _response = self._raw_client.get_upload_links(job_id=job_id, files=files, request_options=request_options)
+        _response = self._raw_client.get_upload_links(
+            job_id=job_id, files=files, request_options=request_options
+        )
         return _response.data
 
     def get_download_links(
-        self, *, job_id: str, files: typing.Sequence[str], request_options: typing.Optional[RequestOptions] = None
+        self,
+        *,
+        job_id: str,
+        files: typing.Sequence[str],
+        request_options: typing.Optional[RequestOptions] = None,
     ) -> FilesDownloadResponse:
         """
         Start a speech to text bulk job V1
@@ -215,12 +230,15 @@ class SpeechToTextJobClient:
             files=["files"],
         )
         """
-        _response = self._raw_client.get_download_links(job_id=job_id, files=files, request_options=request_options)
+        _response = self._raw_client.get_download_links(
+            job_id=job_id, files=files, request_options=request_options
+        )
         return _response.data
 
     def create_job(
         self,
         model: SpeechToTextModel = "saarika:v2.5",
+        mode: typing.Optional[Mode] = None,
         with_diarization: bool = False,
         with_timestamps: bool = False,
         language_code: typing.Optional[SpeechToTextLanguage] = None,
@@ -236,6 +254,10 @@ class SpeechToTextJobClient:
         model : SpeechToTextModel, default="saarika:v2.5"
             The model to use for transcription.
 
+        mode : typing.Optional[Mode], default=None
+            Mode of operation. Only applicable for saaras:v3 model.
+            Options: transcribe, translate, indic-en, verbatim, translit, codemix
+
         with_diarization : typing.Optional[bool], default=False
             Whether to enable speaker diarization (distinguishing who said what).
 
@@ -244,7 +266,7 @@ class SpeechToTextJobClient:
 
         language_code : typing.Optional[SpeechToTextLanguage], default=None
             The language code of the input audio (e.g., "hi-IN", "bn-IN").
-            
+
         num_speakers : typing.Optional[int], default=None
             The number of distinct speakers in the audio, if known.
 
@@ -263,6 +285,7 @@ class SpeechToTextJobClient:
             job_parameters=SpeechToTextJobParametersParams(
                 language_code=language_code,
                 model=model,
+                mode=mode,  # type: ignore[typeddict-item]
                 num_speakers=num_speakers,  # type: ignore[typeddict-item]
                 with_diarization=with_diarization,
                 with_timestamps=with_timestamps,
@@ -350,7 +373,9 @@ class AsyncSpeechToTextJobClient:
         asyncio.run(main())
         """
         _response = await self._raw_client.initialise(
-            job_parameters=job_parameters, callback=callback, request_options=request_options
+            job_parameters=job_parameters,
+            callback=callback,
+            request_options=request_options,
         )
         return _response.data
 
@@ -392,7 +417,9 @@ class AsyncSpeechToTextJobClient:
 
         asyncio.run(main())
         """
-        _response = await self._raw_client.get_status(job_id, request_options=request_options)
+        _response = await self._raw_client.get_status(
+            job_id, request_options=request_options
+        )
         return _response.data
 
     async def start(
@@ -439,11 +466,17 @@ class AsyncSpeechToTextJobClient:
 
         asyncio.run(main())
         """
-        _response = await self._raw_client.start(job_id, ptu_id=ptu_id, request_options=request_options)
+        _response = await self._raw_client.start(
+            job_id, ptu_id=ptu_id, request_options=request_options
+        )
         return _response.data
 
     async def get_upload_links(
-        self, *, job_id: str, files: typing.Sequence[str], request_options: typing.Optional[RequestOptions] = None
+        self,
+        *,
+        job_id: str,
+        files: typing.Sequence[str],
+        request_options: typing.Optional[RequestOptions] = None,
     ) -> FilesUploadResponse:
         """
         Start a speech to text bulk job V1
@@ -482,11 +515,17 @@ class AsyncSpeechToTextJobClient:
 
         asyncio.run(main())
         """
-        _response = await self._raw_client.get_upload_links(job_id=job_id, files=files, request_options=request_options)
+        _response = await self._raw_client.get_upload_links(
+            job_id=job_id, files=files, request_options=request_options
+        )
         return _response.data
 
     async def get_download_links(
-        self, *, job_id: str, files: typing.Sequence[str], request_options: typing.Optional[RequestOptions] = None
+        self,
+        *,
+        job_id: str,
+        files: typing.Sequence[str],
+        request_options: typing.Optional[RequestOptions] = None,
     ) -> FilesDownloadResponse:
         """
         Start a speech to text bulk job V1
@@ -533,6 +572,7 @@ class AsyncSpeechToTextJobClient:
     async def create_job(
         self,
         model: SpeechToTextModel = "saarika:v2.5",
+        mode: typing.Optional[Mode] = None,
         with_diarization: bool = False,
         with_timestamps: bool = False,
         language_code: typing.Optional[SpeechToTextLanguage] = None,
@@ -548,6 +588,10 @@ class AsyncSpeechToTextJobClient:
         model : SpeechToTextModel, default="saarika:v2.5"
             The model to use for transcription.
 
+        mode : typing.Optional[Mode], default=None
+            Mode of operation. Only applicable for saaras:v3 model.
+            Options: transcribe, translate, indic-en, verbatim, translit, codemix
+
         with_diarization : typing.Optional[bool], default=False
             Whether to enable speaker diarization (distinguishing who said what).
 
@@ -556,8 +600,8 @@ class AsyncSpeechToTextJobClient:
 
         language_code : typing.Optional[SpeechToTextLanguage], default=None
             The language code of the input audio (e.g., "hi-IN", "bn-IN").
-            
-        num_speakers : typing.Optional[int], default=None
+
+        num_speakers : typing.Optional[int] = None
             The number of distinct speakers in the audio, if known.
 
         callback : typing.Optional[BulkJobCallbackParams], default=OMIT
@@ -575,6 +619,7 @@ class AsyncSpeechToTextJobClient:
             job_parameters=SpeechToTextJobParametersParams(
                 language_code=language_code,
                 model=model,
+                mode=mode,  # type: ignore[typeddict-item]
                 with_diarization=with_diarization,
                 with_timestamps=with_timestamps,
                 num_speakers=num_speakers,  # type: ignore[typeddict-item]