AI firm Sesame has launched the bottom mannequin that powers Maya, the impressively reasonable voice assistant.
The mannequin, which is 1 billion parameters in dimension (“parameters” referring to particular person elements of the mannequin), is beneath an Apache 2.0 license, which means it may be used commercially with few restrictions. Referred to as CSM-1B, the mannequin generates “RVQ audio codes” from textual content and audio inputs, based on Sesame’s description on the AI dev platform Hugging Face.
RVQ refers to “residual vector quantization,” a method for encoding audio into discrete tokens referred to as codes. RVQ is utilized in plenty of current AI audio applied sciences, together with Google’s SoundStream and Meta’s Encodec.
CSM-1B makes use of a mannequin from Meta’s Llama household as its spine paired with an audio “decoder” part. A fine-tuned variant of CSM powers Maya, Sesame says.
“The mannequin open-sourced here’s a base technology mannequin,” Sesame writes in CSM-1B’s Hugging Face and GitHub repositories. “It’s able to producing quite a lot of voices, but it surely has not been fine-tuned on any particular voice […] The mannequin has some capability for non-English languages as a result of knowledge contamination within the coaching knowledge, but it surely possible gained’t do properly.”
It’s unclear what knowledge Sesame used to coach CSM-1B. The corporate didn’t say.
It’s value noting the mannequin has no actual safeguards to talk of. Sesame has an honor system and merely urges builders and customers to not use the mannequin to imitate an individual’s voice with out their consent, create deceptive content material like faux information, or have interaction in “dangerous” or “malicious” actions.
I attempted the demo on Hugging Face, and cloning my voice took lower than a minute. From there, it was straightforward to generate speech to my coronary heart’s need, together with on controversial subjects just like the election and Russian propaganda.
Shopper Reviews not too long ago warned that many fashionable AI-powered voice cloning instruments available on the market don’t have “significant” safeguards to forestall fraud or abuse.
Sesame, co-founded by Oculus co-creator Brendan Iribe, went viral in late February for its assistant tech, which comes near clearing uncanny valley territory. Maya and Sesame’s different assistant, Miles, take breaths and converse with disfluencies, and might be interrupted whereas talking, very similar to OpenAI’s Voice Mode.
Sesame has raised an undisclosed quantity of capital from Andreessen Horowitz, Spark Capital, and Matrix Companions. Along with constructing voice assistant tech, the corporate says it’s prototyping AI glasses “designed to be worn all day” that’ll be outfitted with its customized fashions.
