Sesame, the AI firm behind the impressively realistic voice assistant Maya, has launched the bottom AI mannequin powering Maya, because it not too long ago promised.
The mannequin, which is 1 billion parameters in measurement (“parameters” referring to particular person elements of the mannequin), is below 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, in keeping with Sesame’s description on the AI dev platform Hugging Face.
RVQ refers to “residual vector quantization,” a method for encoding audio into discrete tokens known as codes. RVQ is used in a number of recent AI audio technologies, 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” element. A fine-tuned variant of CSM powers Maya, Sesame says.
“The mannequin open-sourced here’s a base era mannequin,” Sesame writes in CSM-1B’s Hugging Face and GitHub repositories. “It’s able to producing a wide range of voices, nevertheless it has not been fine-tuned on any particular voice […] The mannequin has some capability for non-English languages resulting from knowledge contamination within the coaching knowledge, nevertheless it seemingly gained’t do effectively.”
It’s unclear what knowledge Sesame used to coach CSM-1B. The corporate didn’t say.
The mannequin has no actual safeguards to talk of, it’s value noting. It’s an “honor system” scenario. Sesame is merely urging 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 interact 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 want, together with on controversial matters just like the election and Russian propaganda:
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 could 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.