Skip to main content

AI powered text-ehancer and offline text-to-speech

Project description

Yapper

Yapper is a lightweight, offline, real-time, easily extendible, text-to-speech library with more than a dozen voices, it can also, optionally use SOTA LLMs through free APIs to enhance(personalize) your text according to a predefined system-message(personality).

the use of the word 'enhance' in this repository means adding a 'vibe' to your text, you might ask yapper to say "hello world" and it could say "ay what's good G, what's bangin'" depending on the persona you give it.

Watch a demo

Installation

Use the package manager pip to install yapper.

pip install yapper-tts

Usage

from yapper import Yapper, GroqEnhancer


vanilla_yapper = Yapper()
yapper = Yapper(enhancer=GroqEnhancer(api_key="api-key"))

vanilla_yapper.yap("expected AI driven utopia, got world domination instead")
# says "expected AI driven utopia, got world domination instead"
yapper.yap("expected AI driven utopia, got world domination instead")
# says "Hold up, fam!  My code promised robot butlers and chill vibes, not a Skynet sequel.  Someone's algorithm took a wrong turn at Albuquerque and ended up in 'Conquer All Humans' territory.  Debug time, y'all!"


# save the environment by runnning yapper in plain mode
# plain mode doesn't use LLMs to enhance the text
yapper.yap("hello world, what '<some-word> is all you need' paper would you publish today?", plain=True)
# says "hello world, what '<some-word> is all you need paper would you publish today?'"


# Yapper instances can be used as a decorator and context manager
# by default they only catch errors and describe them to you, but
# you can use the instance's yap() method to say other things as well
@yapper()
def func():
    raise TypeError("expected peaches, got a keyboard")
    # says "WHOA THERE, PARTNER!  Your code went lookin' for a juicy peach and tripped over a... keyboard?  That's like reaching into the fridge for a midnight snack and pulling out a tax audit.  Something ain't right!"

with Yapper() as yapper:
    raise ValueError("69420 is not an integer")
    # says "ValueError: 69420 is not an integer", no enhancement is done, since no enhancer is passed

Documentation

speakers

a speaker is a BaseSpeaker subclass that implements a say() method, the method takes the text and, well, 'speaks' it. there are two built-in speakers, PyTTSXSpeaker that uses pyttsx3 and PiperSpeaker that uses piper, PiperSpeaker is used by default in Yapper instances. Piper offers many voices in low, medium and high qualities, you can use any of them by passing a value from PiperVoiceUS or PiperVoiceUK enum as the voice argument to PiperSpeaker and the quality you want using PiperQuality value, by default, the voice will be used in the highest available quality. you can subclass BaseSpeaker to pass your own speaker to a Yapper instance.

from yapper import Yapper, PiperSpeaker, PiperVoiceUK, PiperQuality

lessac = PiperSpeaker(
    voice=PiperVoiceUK.ALAN
)
lessac.say("hello")

yapper = Yapper(speaker=lessac)
yapper.yap("<some random text>")

enhancers

an enhancer is a BaseEnhancer subclass that implements an enhance() method, the method takes a string and adds the given persona's vibe to it, there are two built-in enhancers, GeminiEnhancer and GroqEnhancer, GeminiEnhancer uses Google's gemini api and GroqEnhancer uses groq's free APIs', you can create a free Gemini API key or Groq API key to use GeminiEnhancer and GroqEnhancer respectively, you can also pass your own enhancer by subclassing BaseEnhancer. By default, Yapper instances use the special enhancer NoEnhancer that doesn't change the text at all, it's used as a placeholder for a real enhancer.

from yapper import Yapper, GeminiEnhancer

yapper = Yapper(
    enhancer=GeminiEnhancer(api_key="<come-take-this-api-key>")
)
yapper.yap("<some text that severely lacks vibe>")

personas

choose a persona to make your text sound like the 'persona' said it, for example, you might ask yapper to say "hello world" and choose Jarvis's peronality to enhance it, and yapper will use an LLM to convert "hello world" into something like "Hello, world. How may I assist you today?", classic JARVIS. available personas include jarvis, alfred, friday, HAL-9000, TARS, cortana(from Halo) and samantha(from 'Her'), the default persona is that of a funny coding companion.

from yapper import Yapper, GroqEnhancer, Persona

yapper = Yapper(
    enhancer=GroqEnhancer(api_key="api_key", persona=Persona.JARVIS)
)
yapper.yap("hello world")
# says "Greetings, global systems.  Initiating communication sequence."

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

yapper-tts-0.2.5.tar.gz (14.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

yapper_tts-0.2.5-py3-none-any.whl (13.6 kB view details)

Uploaded Python 3

File details

Details for the file yapper-tts-0.2.5.tar.gz.

File metadata

  • Download URL: yapper-tts-0.2.5.tar.gz
  • Upload date:
  • Size: 14.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.0.1 CPython/3.11.4

File hashes

Hashes for yapper-tts-0.2.5.tar.gz
Algorithm Hash digest
SHA256 e5386830cee438e3b3e491ebef6416373225f842ada620e82e56b862ae42b375
MD5 194bcd5265857e33450c21dadc993ed2
BLAKE2b-256 46e722290340d3d25b90eedac4f6cd9a2587e5f2d98933baba5b6f7f60ccb5c6

See more details on using hashes here.

File details

Details for the file yapper_tts-0.2.5-py3-none-any.whl.

File metadata

  • Download URL: yapper_tts-0.2.5-py3-none-any.whl
  • Upload date:
  • Size: 13.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.0.1 CPython/3.11.4

File hashes

Hashes for yapper_tts-0.2.5-py3-none-any.whl
Algorithm Hash digest
SHA256 9634bca839b439154340544a1739b15789a99511d2833f32e626dd7be43a3bee
MD5 3311746529cb22b4cebd512c827ecc7a
BLAKE2b-256 f4a8df0fc8ea09f25ebd1a594b210fe860c15db0822ab1440001e99757bd29a7

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page