cartesia 1.1.0.dev0__tar.gz → 1.2.0__tar.gz

{cartesia-1.1.0.dev0 → cartesia-1.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cartesia
-Version: 1.1.0.dev0
+Version: 1.2.0
 Summary: The official Python library for the Cartesia API.
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown
@@ -642,8 +642,6 @@ display(audio)
 
 You can use the `client.tts.get_output_format` method to convert string-based output format names into the `output_format` dictionary which is expected by the `output_format` parameter. You can see the `OutputFormatMapping` class in `cartesia._types` for the currently supported output format names. You can also view the currently supported `output_format`s in our [API Reference](https://docs.cartesia.ai/reference/api-reference/rest/stream-speech-server-sent-events).
 
-The previously used `output_format` strings are now deprecated and will be removed in v1.2.0. These are listed in the `DeprecatedOutputFormatMapping` class in `cartesia._types`.
-
 ```python
 # Get the output format dictionary from string name
 output_format = client.tts.get_output_format("raw_pcm_f32le_44100")

{cartesia-1.1.0.dev0 → cartesia-1.2.0}/README.md

@@ -629,8 +629,6 @@ display(audio)
 
 You can use the `client.tts.get_output_format` method to convert string-based output format names into the `output_format` dictionary which is expected by the `output_format` parameter. You can see the `OutputFormatMapping` class in `cartesia._types` for the currently supported output format names. You can also view the currently supported `output_format`s in our [API Reference](https://docs.cartesia.ai/reference/api-reference/rest/stream-speech-server-sent-events).
 
-The previously used `output_format` strings are now deprecated and will be removed in v1.2.0. These are listed in the `DeprecatedOutputFormatMapping` class in `cartesia._types`.
-
 ```python
 # Get the output format dictionary from string name
 output_format = client.tts.get_output_format("raw_pcm_f32le_44100")

{cartesia-1.1.0.dev0 → cartesia-1.2.0}/cartesia/_types.py

@@ -27,39 +27,6 @@ class OutputFormatMapping:
             raise ValueError(f"Unsupported format: {format_name}")
 
 
-class DeprecatedOutputFormatMapping:
-    """Deprecated formats as of v1.0.1. These will be removed in v1.2.0. Use :class:`OutputFormatMapping` instead."""
-
-    _format_mapping = {
-        "fp32": {"container": "raw", "encoding": "pcm_f32le", "sample_rate": 44100},
-        "pcm": {"container": "raw", "encoding": "pcm_s16le", "sample_rate": 44100},
-        "fp32_8000": {"container": "raw", "encoding": "pcm_f32le", "sample_rate": 8000},
-        "fp32_16000": {"container": "raw", "encoding": "pcm_f32le", "sample_rate": 16000},
-        "fp32_22050": {"container": "raw", "encoding": "pcm_f32le", "sample_rate": 22050},
-        "fp32_24000": {"container": "raw", "encoding": "pcm_f32le", "sample_rate": 24000},
-        "fp32_44100": {"container": "raw", "encoding": "pcm_f32le", "sample_rate": 44100},
-        "pcm_8000": {"container": "raw", "encoding": "pcm_s16le", "sample_rate": 8000},
-        "pcm_16000": {"container": "raw", "encoding": "pcm_s16le", "sample_rate": 16000},
-        "pcm_22050": {"container": "raw", "encoding": "pcm_s16le", "sample_rate": 22050},
-        "pcm_24000": {"container": "raw", "encoding": "pcm_s16le", "sample_rate": 24000},
-        "pcm_44100": {"container": "raw", "encoding": "pcm_s16le", "sample_rate": 44100},
-        "mulaw_8000": {"container": "raw", "encoding": "pcm_mulaw", "sample_rate": 8000},
-        "alaw_8000": {"container": "raw", "encoding": "pcm_alaw", "sample_rate": 8000},
-    }
-
-    @classmethod
-    @deprecated(
-        vdeprecated="1.0.1",
-        vremove="1.2.0",
-        reason="Old output format names are being deprecated in favor of names aligned with the Cartesia API. Use names from `OutputFormatMapping` instead.",
-    )
-    def get_format_deprecated(cls, format_name):
-        if format_name in cls._format_mapping:
-            return cls._format_mapping[format_name]
-        else:
-            raise ValueError(f"Unsupported format: {format_name}")
-
-
 class VoiceMetadata(TypedDict):
     id: str
     name: str

{cartesia-1.1.0.dev0 → cartesia-1.2.0}/cartesia/tts.py

@@ -4,7 +4,6 @@ import httpx
 
 from cartesia._sse import _SSE
 from cartesia._types import (
-    DeprecatedOutputFormatMapping,
     OutputFormat,
     OutputFormatMapping,
     VoiceControls,
@@ -86,10 +85,6 @@ class TTS(Resource):
         """
         if output_format_name in OutputFormatMapping._format_mapping:
             output_format_obj = OutputFormatMapping.get_format(output_format_name)
-        elif output_format_name in DeprecatedOutputFormatMapping._format_mapping:
-            output_format_obj = DeprecatedOutputFormatMapping.get_format_deprecated(
-                output_format_name
-            )
         else:
             raise ValueError(f"Unsupported format: {output_format_name}")
 
@@ -114,10 +109,6 @@ class TTS(Resource):
         """
         if output_format_name in OutputFormatMapping._format_mapping:
             output_format_obj = OutputFormatMapping.get_format(output_format_name)
-        elif output_format_name in DeprecatedOutputFormatMapping._format_mapping:
-            output_format_obj = DeprecatedOutputFormatMapping.get_format_deprecated(
-                output_format_name
-            )
         else:
             raise ValueError(f"Unsupported format: {output_format_name}")
 

cartesia-1.2.0/cartesia/version.py

@@ -0,0 +1 @@
+__version__ = "1.2.0"

{cartesia-1.1.0.dev0 → cartesia-1.2.0}/cartesia/voices.py

@@ -58,29 +58,65 @@ class Voices(Resource):
 
         return response.json()
 
-    def clone(self, filepath: Optional[str] = None, enhance: str = True) -> List[float]:
+    def clone(
+        self,
+        filepath: Optional[str] = None,
+        enhance: str = True,
+        mode: str = "clip",
+        language: str = "en",
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        transcript: Optional[str] = None,
+    ) -> Union[List[float], VoiceMetadata]:
         """Clone a voice from a clip.
 
         Args:
             filepath: The path to the clip file.
             enhance: Whether to enhance the clip before cloning the voice (highly recommended). Defaults to True.
+            mode: The mode to use for cloning. Either "similarity" or "stability".
+            language: The language code of the language spoken in the clip. Defaults to "en".
+            name: The name of the cloned voice.
+            description: The description of the cloned voice.
+            transcript: The transcript of the clip. Only used if mode is "similarity".
 
         Returns:
             The embedding of the cloned voice as a list of floats.
         """
         if not filepath:
             raise ValueError("Filepath must be specified.")
-        url = f"{self._http_url()}/voices/clone/clip"
+        headers = self.headers.copy()
+        headers.pop("Content-Type", None)
+
         with open(filepath, "rb") as file:
             files = {"clip": file}
-            files["enhance"] = str(enhance).lower()
-            headers = self.headers.copy()
-            headers.pop("Content-Type", None)
-            response = httpx.post(url, headers=headers, files=files, timeout=self.timeout)
-            if not response.is_success:
-                raise ValueError(f"Failed to clone voice from clip. Error: {response.text}")
-
-        return response.json()["embedding"]
+            data = {
+                "enhance": str(enhance).lower(),
+                "mode": mode,
+            }
+            if mode == "clip":
+                url = f"{self._http_url()}/voices/clone/clip"
+                response = httpx.post(
+                    url, headers=headers, files=files, data=data, timeout=self.timeout
+                )
+                if not response.is_success:
+                    raise ValueError(f"Failed to clone voice from clip. Error: {response.text}")
+                return response.json()["embedding"]
+            else:
+                data["name"] = name
+                data["description"] = description
+                data["language"] = language
+                if mode == "similarity" and transcript:
+                    data["transcript"] = transcript
+                url = f"{self._http_url()}/voices/clone"
+                response = httpx.post(
+                    url, headers=headers, files=files, data=data, timeout=self.timeout
+                )
+                if not response.is_success:
+                    raise ValueError(
+                        f"Failed to clone voice. Status Code: {response.status_code}\n"
+                        f"Error: {response.text}"
+                    )
+                return response.json()
 
     def create(
         self,
@@ -88,6 +124,7 @@ class Voices(Resource):
         description: str,
         embedding: List[float],
         base_voice_id: Optional[str] = None,
+        language: str = "en",
     ) -> VoiceMetadata:
         """Create a new voice.
 
@@ -108,6 +145,7 @@ class Voices(Resource):
                 "description": description,
                 "embedding": embedding,
                 "base_voice_id": base_voice_id,
+                "language": language,
             },
             timeout=self.timeout,
         )

{cartesia-1.1.0.dev0 → cartesia-1.2.0}/cartesia.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cartesia
-Version: 1.1.0.dev0
+Version: 1.2.0
 Summary: The official Python library for the Cartesia API.
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown
@@ -642,8 +642,6 @@ display(audio)
 
 You can use the `client.tts.get_output_format` method to convert string-based output format names into the `output_format` dictionary which is expected by the `output_format` parameter. You can see the `OutputFormatMapping` class in `cartesia._types` for the currently supported output format names. You can also view the currently supported `output_format`s in our [API Reference](https://docs.cartesia.ai/reference/api-reference/rest/stream-speech-server-sent-events).
 
-The previously used `output_format` strings are now deprecated and will be removed in v1.2.0. These are listed in the `DeprecatedOutputFormatMapping` class in `cartesia._types`.
-
 ```python
 # Get the output format dictionary from string name
 output_format = client.tts.get_output_format("raw_pcm_f32le_44100")

{cartesia-1.1.0.dev0 → cartesia-1.2.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "cartesia"
-version = "1.1.0-dev0"
+version = "1.2.0"
 description = "The official Python library for the Cartesia API."
 readme = "README.md"
 requires-python = ">=3.9"

{cartesia-1.1.0.dev0 → cartesia-1.2.0}/tests/test_tts.py

@@ -1093,17 +1093,6 @@ def test_output_formats(resources: _Resources, output_format_name: str):
     assert output_format["encoding"] is not None, "Output format encoding is None"
     assert output_format["sample_rate"] is not None, "Output format sample rate is None"
 
-
-@pytest.mark.parametrize("output_format_name", deprecated_output_format_names)
-def test_deprecated_output_formats(resources: _Resources, output_format_name: str):
-    logger.info(f"Testing deprecated output format: {output_format_name}")
-    output_format = resources.client.tts.get_output_format(output_format_name)
-    assert isinstance(output_format, dict), "Output is not of type dict"
-    assert output_format["container"] is not None, "Output format container is None"
-    assert output_format["encoding"] is not None, "Output format encoding is None"
-    assert output_format["sample_rate"] is not None, "Output format sample rate is None"
-
-
 def test_invalid_output_format(resources: _Resources):
     logger.info("Testing invalid output format")
     with pytest.raises(ValueError):

cartesia-1.1.0.dev0/cartesia/version.py

@@ -1 +0,0 @@
-__version__ = "1.1.0-dev0"