cartesia 2.0.13__tar.gz → 2.0.15__tar.gz

{cartesia-2.0.13 → cartesia-2.0.15}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cartesia
-Version: 2.0.13
+Version: 2.0.15
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{cartesia-2.0.13 → cartesia-2.0.15}/pyproject.toml

@@ -3,7 +3,7 @@ name = "cartesia"
 
 [tool.poetry]
 name = "cartesia"
-version = "2.0.13"
+version = "2.0.15"
 description = ""
 readme = "README.md"
 authors = []

{cartesia-2.0.13 → cartesia-2.0.15}/src/cartesia/core/client_wrapper.py

@@ -16,7 +16,7 @@ class BaseClientWrapper:
         headers: typing.Dict[str, str] = {
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "cartesia",
-            "X-Fern-SDK-Version": "2.0.13",
+            "X-Fern-SDK-Version": "2.0.15",
         }
         headers["X-API-Key"] = self.api_key
         headers["Cartesia-Version"] = "2024-11-13"

{cartesia-2.0.13 → cartesia-2.0.15}/src/cartesia/tts/_async_websocket.py

@@ -8,7 +8,7 @@ from typing import Any, AsyncGenerator, Callable, Dict, List, Optional, Union
 
 import aiohttp
 
-from cartesia.tts.requests import TtsRequestVoiceSpecifierParams
+from cartesia.tts.requests import GenerationConfigParams, TtsRequestVoiceSpecifierParams
 from cartesia.tts.requests.output_format import OutputFormatParams
 from cartesia.tts.types import (
     WebSocketResponse,
@@ -61,6 +61,7 @@ class _AsyncTTSContext:
         model_id: str,
         transcript: str,
         output_format: OutputFormatParams,
+        generation_config: Optional[GenerationConfigParams] = None,
         voice: TtsRequestVoiceSpecifierParams,
         context_id: Optional[str] = None,
         duration: Optional[int] = None,
@@ -116,6 +117,12 @@ class _AsyncTTSContext:
         if flush:
             request_body["flush"] = flush
 
+        if generation_config is not None:
+            if isinstance(generation_config, dict):
+                request_body["generation_config"] = generation_config
+            else:
+                request_body["generation_config"] = generation_config.dict()
+
         if (
             "context_id" in request_body
             and request_body["context_id"] is not None
@@ -315,10 +322,10 @@ class AsyncTtsWebsocket(TtsWebsocket):
                 # Extract status code if available
                 status_code = None
                 error_message = str(e)
-                
+
                 if hasattr(e, 'status') and e.status is not None:
                     status_code = e.status
-                
+
                     # Create a meaningful error message based on status code
                     if status_code == 402:
                         error_message = "Payment required. Your API key may have insufficient credits or permissions."
@@ -328,7 +335,7 @@ class AsyncTtsWebsocket(TtsWebsocket):
                         error_message = "Forbidden. You don't have permission to access this resource."
                     elif status_code == 404:
                         error_message = "Not found. The requested resource doesn't exist."
-                    
+
                     raise RuntimeError(f"Failed to connect to WebSocket.\nStatus: {status_code}. Error message: {error_message}")
                 else:
                     raise RuntimeError(f"Failed to connect to WebSocket at {url}. {e}")
@@ -369,6 +376,7 @@ class AsyncTtsWebsocket(TtsWebsocket):
         output_format: OutputFormatParams,
         voice: TtsRequestVoiceSpecifierParams,
         context_id: Optional[str] = None,
+        generation_config: Optional[GenerationConfigParams] = None,
         duration: Optional[int] = None,
         language: Optional[str] = None,
         stream: bool = True,
@@ -388,6 +396,7 @@ class AsyncTtsWebsocket(TtsWebsocket):
             output_format=output_format,
             voice=voice,
             context_id=context_id,
+            generation_config=generation_config,
             duration=duration,
             language=language,
             continue_=False,

{cartesia-2.0.13 → cartesia-2.0.15}/src/cartesia/tts/_websocket.py

@@ -14,7 +14,7 @@ except ImportError:
 
 from iterators import TimeoutIterator  # type: ignore
 
-from cartesia.tts.requests import TtsRequestVoiceSpecifierParams
+from cartesia.tts.requests import GenerationConfigParams, TtsRequestVoiceSpecifierParams
 from cartesia.tts.requests.output_format import OutputFormatParams
 from cartesia.tts.types import (
     WebSocketResponse,
@@ -60,6 +60,7 @@ class _TTSContext:
         model_id: str,
         transcript: typing.Generator[str, None, None],
         output_format: OutputFormatParams,
+        generation_config: Optional[GenerationConfigParams] = None,
         voice: TtsRequestVoiceSpecifierParams,
         context_id: Optional[str] = None,
         max_buffer_delay_ms: Optional[int] = None,
@@ -111,6 +112,11 @@ class _TTSContext:
         if max_buffer_delay_ms:
             request_body["max_buffer_delay_ms"] = max_buffer_delay_ms
 
+        if generation_config is not None:
+            if isinstance(generation_config, dict):
+                request_body["generation_config"] = generation_config
+            else:
+                request_body["generation_config"] = generation_config.dict()
         if (
             "context_id" in request_body
             and request_body["context_id"] is not None
@@ -293,10 +299,10 @@ class TtsWebsocket:
                 # Extract status code if available
                 status_code = None
                 error_message = str(e)
-                
+
                 if hasattr(e, 'status') and e.status is not None:
                     status_code = e.status
-                
+
                     # Create a meaningful error message based on status code
                     if status_code == 402:
                         error_message = "Payment required. Your API key may have insufficient credits or permissions."
@@ -306,7 +312,7 @@ class TtsWebsocket:
                         error_message = "Forbidden. You don't have permission to access this resource."
                     elif status_code == 404:
                         error_message = "Not found. The requested resource doesn't exist."
-                    
+
                     raise RuntimeError(f"Failed to connect to WebSocket.\nStatus: {status_code}. Error message: {error_message}")
                 else:
                     raise RuntimeError(f"Failed to connect to WebSocket. {e}")
@@ -356,6 +362,7 @@ class TtsWebsocket:
         transcript: str,
         output_format: OutputFormatParams,
         voice: TtsRequestVoiceSpecifierParams,
+        generation_config: Optional[GenerationConfigParams] = None,
         context_id: Optional[str] = None,
         duration: Optional[int] = None,
         language: Optional[str] = None,
@@ -388,6 +395,7 @@ class TtsWebsocket:
             "output_format": output_format,
             "voice": voice,
             "context_id": context_id,
+            "generation_config": generation_config,
             "duration": duration,
             "language": language,
             "stream": stream,