Skip to content

Commit

Permalink
feat: update to pydantic 2.0, update error handling, fix voice settin…
Browse files Browse the repository at this point in the history 
…gs/fetching, fix voice design/clone
  • Loading branch information
@flavioschneider
flavioschneider committed Sep 8, 2023
1 parent 3d1f80a commit 1af6a16
Show file tree
Hide file tree
Showing 11 changed files with 131 additions and 507 deletions.
58 changes: 7 additions & 51 deletions API.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -168,7 +168,7 @@ Models(

### `Voices`

The `Voices` API is used to get a list of all available voices for the authenticated user. The `Voice` contains all the info for a voice, such as the voice's stability and similarity_boost, and can be passed to the `generate` function as the `voice` argument to select the voice. The voice settings can be changed to control the voice behaviour.
The `Voices` API is used to get a list of all available voices for the authenticated user. The `Voice` contains all the info for a voice, and can be passed to the `generate` function as the `voice` argument to select the voice. The voice settings can be changed to control the voice behaviour.

```py
from elevenlabs.api import Voices
Expand All @@ -184,7 +184,7 @@ Voice(
voice_id='21m00Tcm4TlvDq8ikWAM',
name='Rachel',
category='premade',
settings=VoiceSettings(stability=0.75, similarity_boost=0.75)
settings=None
)
```

Expand All @@ -195,72 +195,27 @@ Voices(
voice_id='21m00Tcm4TlvDq8ikWAM',
name='Rachel',
category='premade',
settings=VoiceSettings(stability=0.75, similarity_boost=0.75)
settings=None
),
Voice(
voice_id='AZnzlk1XvdvUeBnXmlld',
name='Domi',
category='premade',
settings=VoiceSettings(stability=0.1, similarity_boost=0.75)
settings=None
),
Voice(
voice_id='EXAVITQu4vr4xnSDxMaL',
name='Bella',
category='premade',
settings=VoiceSettings(stability=0.245, similarity_boost=0.75)
settings=None
),
Voice(
voice_id='ErXwobaYiN019PkySvjV',
name='Antoni',
category='premade',
settings=VoiceSettings(stability=0.195, similarity_boost=0.75)
),
Voice(
voice_id='MF3mGyEYCl7XYWbV9V6O',
name='Elli',
category='premade',
settings=VoiceSettings(stability=0.755, similarity_boost=0.75)
),
Voice(
voice_id='TxGEqnHWrfWFTfGW9XjX',
name='Josh',
category='premade',
settings=VoiceSettings(stability=0.15, similarity_boost=0.51)
),
Voice(
voice_id='VR6AewLTigWG4xSOukaG',
name='Arnold',
category='premade',
settings=VoiceSettings(stability=0.15, similarity_boost=0.75)
),
Voice(
voice_id='pNInz6obpgDQGcFmaJgB',
name='Adam',
category='premade',
settings=VoiceSettings(stability=0.2, similarity_boost=0.75)
),
Voice(
voice_id='yoZ06aMxZJJ28mfd3POQ',
name='Sam',
category='premade',
settings=VoiceSettings(stability=0.25, similarity_boost=0.75)
)
...
]
)
```

</details>

An example of changing the `Voice` object to control the voice behaviour:

```py
my_voice = voices[0]
my_voice.settings.stability = 0.1
my_voice.settings.similarity_boost = 0.75

text = f""" Hi! My name is {my_voice.name}, nice to meet you! """
audio = generate(text, voice=my_voice)
```

### `VoiceDesign`

Expand All @@ -272,6 +227,7 @@ from elevenlabs import Voice, VoiceDesign, Gender, Age, Accent, play
design = VoiceDesign(
name='Lexa',
text="Hello, my name is Lexa. I'm your personal assistant, I can help you with your daily tasks and I can also read you the news.",
voice_description="Calm and soft with a slight British accent.",
gender=Gender.female,
age=Age.young,
accent=Accent.british,
Expand Down
153 changes: 26 additions & 127 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -18,14 +18,15 @@ pip install elevenlabs
[![Open in Spaces](https://img.shields.io/badge/🤗-Open%20in%20Spaces-blue.svg)](https://huggingface.co/spaces/elevenlabs/tts)
[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/gist/flavioschneider/49468d728a816c6538fd2f56b3b50b96/elevenlabs-python.ipynb)

We support two main models: the newest `eleven_multilingual_v2`, a single foundational model supporting 28 languages including English, Chinese, Spanish, Hindi, Portuguese, French, German, Japanese, Arabic, Korean, Indonesian, Italian, Dutch, Turkish, Polish, Swedish, Filipino, Malay, Romanian, Ukrainian, Greek, Czech, Danish, Finnish, Bulgarian, Croatian, Slovak, and Tamil; and `eleven_monolingual_v1`, a low-latency model specifically trained for English speech.

```py
from elevenlabs import generate, play

audio = generate(
text="Hi! My name is Bella, nice to meet you!",
text="Hello! 你好! Hola! नमस्ते! Bonjour! こんにちは! مرحبا! 안녕하세요! Ciao! Cześć! Привіт! வணக்கம்!",
voice="Bella",
model="eleven_monolingual_v1"
model="eleven_multilingual_v2"
)

play(audio)
Expand All @@ -35,78 +36,17 @@ play(audio)

<i> Don't forget to unmute the player! </i>

<b>Adam</b>

[Adam.webm](https://user-images.githubusercontent.com/12028621/232730475-4babdd1b-6078-47d0-811a-68678d009918.webm)

<b>Antoni</b>

[Antoni.webm](https://user-images.githubusercontent.com/12028621/232730870-164e2423-26d2-4423-89ff-36e78483e4e2.webm)

<b>Arnold</b>

[Arnold.webm](https://user-images.githubusercontent.com/12028621/232731257-f7cccbf0-e4d3-49de-9bc8-280e54a29e88.webm)

<b>Bella</b>

[Bella.webm](https://user-images.githubusercontent.com/12028621/232731276-00a8e665-5f7c-4fe7-adcf-47ac0d634874.webm)

<b>Domi</b>

[Domi.webm](https://user-images.githubusercontent.com/12028621/232731299-ca33fdab-fa79-4343-afad-ece0d4363ffe.webm)

<b>Elli</b>

[Elli.webm](https://user-images.githubusercontent.com/12028621/232731318-a1debbd9-ce06-4e71-8199-119cddb2f19c.webm)

<b>Josh</b>

[Josh.webm](https://user-images.githubusercontent.com/12028621/232731374-f81bcc7c-d30c-4958-8086-2271274d6f12.webm)

<b>Rachel</b>

[Rachel.webm](https://user-images.githubusercontent.com/12028621/232731393-9ccdcf54-a957-44ac-b882-67a95e95d7d0.webm)

<b>Sam</b>

[Sam.webm](https://user-images.githubusercontent.com/12028621/232731428-18bca274-6b84-42e4-b4d8-819b0bd0a19a.webm)

</details>

## 🌎 Multilingual

The `eleven_multilingual_v1` model supports multiple languages, including English, German, Polish, Spanish, Italian, French, Portuguese, and Hindi.
```py
from elevenlabs import generate, play

audio = generate(
text="¡Hola! Mi nombre es Arnold, encantado de conocerte!",
voice="Arnold",
model='eleven_multilingual_v1'
)

play(audio)
```

<details> <summary> Play </summary>

<i> Don't forget to unmute the player! </i>

[hola.webm](https://user-images.githubusercontent.com/12028621/235474694-584f7103-dab2-4c39-bb9a-8e5f00be85da.webm)

</details>


## 🗣️ Voices

List all your available voices with `voices()`.
```py
from elevenlabs import voices, generate

voices = voices()

audio = generate(text="Hello there!", voice=voices[0])

print(voices)
```

Expand All @@ -119,80 +59,39 @@ Voices(
voice_id='21m00Tcm4TlvDq8ikWAM',
name='Rachel',
category='premade',
settings=VoiceSettings(stability=0.75, similarity_boost=0.75)
settings=None,
),
Voice(
voice_id='AZnzlk1XvdvUeBnXmlld',
name='Domi',
category='premade',
settings=VoiceSettings(stability=0.1, similarity_boost=0.75)
settings=None,
),
Voice(
voice_id='EXAVITQu4vr4xnSDxMaL',
name='Bella',
category='premade',
settings=VoiceSettings(stability=0.245, similarity_boost=0.75)
),
Voice(
voice_id='ErXwobaYiN019PkySvjV',
name='Antoni',
category='premade',
settings=VoiceSettings(stability=0.195, similarity_boost=0.75)
),
Voice(
voice_id='MF3mGyEYCl7XYWbV9V6O',
name='Elli',
category='premade',
settings=VoiceSettings(stability=0.755, similarity_boost=0.75)
),
Voice(
voice_id='TxGEqnHWrfWFTfGW9XjX',
name='Josh',
category='premade',
settings=VoiceSettings(stability=0.15, similarity_boost=0.51)
),
Voice(
voice_id='VR6AewLTigWG4xSOukaG',
name='Arnold',
category='premade',
settings=VoiceSettings(stability=0.15, similarity_boost=0.75)
),
Voice(
voice_id='pNInz6obpgDQGcFmaJgB',
name='Adam',
category='premade',
settings=VoiceSettings(stability=0.2, similarity_boost=0.75)
),
Voice(
voice_id='yoZ06aMxZJJ28mfd3POQ',
name='Sam',
category='premade',
settings=VoiceSettings(stability=0.25, similarity_boost=0.75)
),
Voice(
voice_id='3KehPe3gxEYqOFSGDzGM',
name='test',
category='cloned',
settings=VoiceSettings(stability=0.75, similarity_boost=0.75)
),
Voice(
voice_id='duorpit9NOULscUJ2OAp',
name='test',
category='cloned',
settings=VoiceSettings(stability=0.75, similarity_boost=0.75)
),
Voice(
voice_id='h2rNV1Iql95D2QSSuvLY',
name='test',
category='cloned',
settings=VoiceSettings(stability=0.75, similarity_boost=0.75)
)
...
]
)
```

</details>

Build a voice object with custom settings to personalize the voice style, or call `voice.fetch_settings()` to get the default settings for the voice.

```py
from elevenlabs import Voice, VoiceSettings, generate

audio = generate(
text="Hello! My name is Bella.",
voice=Voice(
voice_id='EXAVITQu4vr4xnSDxMaL',
settings=VoiceSettings(stability=0.71, similarity_boost=0.5, style=0.0, use_speaker_boost=True)
)
)

play(audio)
```

</details>


### Clone Voice

Expand Down Expand Up @@ -228,7 +127,7 @@ stream(audio_stream)
```

### Input streaming
Stream text chunks into audio as it's being generated. Note chunks must end with a space or punctuation (" ", ".", "?", "!").
Stream text chunks into audio as it's being generated, with <1s latency. Note: if chunks don't end with space or punctuation (" ", ".", "?", "!"), the stream will wait for more text.
```py
from elevenlabs import generate, stream

Expand All @@ -248,7 +147,7 @@ stream(audio_stream)

## 🔑 API Key

The basic API has a limited number of characters. To increase this limit, you can get a free API key from [Elevenlabs](https://elevenlabs.io/) ([step-by-step guide](https://docs.elevenlabs.io/authentication/01-xi-api-key)) and set is as environment variable `ELEVEN_API_KEY`. Alternatively you can provide the `api_key` string argument to the `generate` function, or set it globally in code with:
The basic API has a limited number of characters. To increase this limit, you can get a free API key from [Elevenlabs](https://elevenlabs.io/) ([step-by-step guide](https://docs.elevenlabs.io/authentication/01-xi-api-key)) and set is as environment variable `ELEVEN_API_KEY`. Alternatively, you can provide the `api_key` string argument to the `generate` function, or set it globally in code with:

```py
from elevenlabs import set_api_key
Expand Down
31 changes: 21 additions & 10 deletions elevenlabs/api/base.py
View file
Original file line number Diff line number Diff line change
@@ -1,3 +1,4 @@
import json
import os
from typing import Optional, Sequence

Expand All @@ -7,7 +8,6 @@
from .error import (
APIError,
AuthorizationError,
HTTPError,
RateLimitError,
UnauthenticatedRateLimitError,
)
Expand All @@ -16,12 +16,12 @@


class API(BaseModel):
class Config:
class ConfigDict:
# Parse enum to strings when converting to dict
use_enum_values = True
# Validate fields when setting manually
validate_assignment = True
#
# Allows having a field called `model_id` in the class
protected_namespaces = ()

@staticmethod
Expand All @@ -43,18 +43,29 @@ def request(url: str, method: str, api_key: Optional[str] = None, **kwargs):
if status_code == 200:
return response

error = HTTPError(response)
# Parse the error message and status
error = json.loads(response.text)
message, status = "", ""
if "detail" in error:
detail = error["detail"]
if isinstance(error, dict):
message = detail.get("message", "")
status = detail.get("status", "")
else:
message = str(error)
status = str(response.status_code)

# Raise the appropriate error
if status_code == 401:
if error.status == "quota_exceeded":
if status == "quota_exceeded":
if api_key is None:
raise UnauthenticatedRateLimitError(error)
raise UnauthenticatedRateLimitError(message)
else:
raise RateLimitError(error)
elif error.status == "needs_authorization":
raise AuthorizationError(error)
raise RateLimitError(message)
elif status == "needs_authorization":
raise AuthorizationError(message)

raise APIError(error)
raise APIError(message, status)

@staticmethod
def get(url: str, *args, **kwargs):
Expand Down
Loading

0 comments on commit 1af6a16

Please sign in to comment.