Llasa: Llama-Based Speech Synthesis
llasatts.github.io
Llasa: Llama-Based Speech Synthesis
1–10 of 23 posts
Re: Llasa: Llama-Based Speech Synthesis
#2LLaSA is a simple framework for speech synthesis that employs a single-layer vector quantizer (VQ) codec and a single Transformer architecture to fully align with standard LLMs such as LLaMA.
Re: Llasa: Llama-Based Speech Synthesis
#3LLaSA is a simple framework for speech synthesis that employs a single-layer vector quantizer (VQ) codec and a single Transformer architecture to fully align with standard LLMs such as LLaMA.
Probably the title should have the correct capitalization then. Cause I was fully expecting a speech synthesis tool that sounded like llamas talking human language and now I'm bummed out!
Re: Llasa: Llama-Based Speech Synthesis
#4I can't wait see this integrated into Open WebUI! These sound amazing.
Re: Llasa: Llama-Based Speech Synthesis
#5the long 'uuuuhhhhhhh' from some of the lesser models is killing me.
Re: Llasa: Llama-Based Speech Synthesis
#6the long 'uuuuhhhhhhh' from some of the lesser models is killing me.
based on the samples, it really seams like anything smaller than 3B is pretty useless.
Re: Llasa: Llama-Based Speech Synthesis
#7> employs a single-layer vector quantizer (VQ) codec and a single Transformer architecture to fully align
I really wish when new models were released that they would draw a diagram of all the layers and the tensor input and output sizes at each layer, with zoom in/out capabilities if needed using D3.js or whatever visualization framework if needed. Every single layer should be on there with its input and output sizes.
These one-sentence descriptions, and approximate block diagrams with arrows pointing at each other are never enough to understand how something is actually implemented.
Re: Llasa: Llama-Based Speech Synthesis
#8Odd that the page doesn't seem to link to either,
Re: Llasa: Llama-Based Speech Synthesis
#9the long 'uuuuhhhhhhh' from some of the lesser models is killing me.
based on the samples, it really seams like anything smaller than 3B is pretty useless.
If you're doing a home lab voice assistant 1B is nice, because on a 12gb gpu you can run a moderately competent 7b LLM and two 1b models; 1 for speech to text and also text to speech, plus some for the wake word monitor. Maybe in a couple of years we can combine all this into a single ~8b model that runs efficiently on 12gb gpu. Nvidia doesn't seem very incentivized right now to sell consumer GPUs that can run all this on a single consumer grade chip when they're making so much money selling commercial grade 48gb cards.
Re: Llasa: Llama-Based Speech Synthesis
#10> employs a single-layer vector quantizer (VQ) codec and a single Transformer architecture to fully align I really wish when new models were released that they would draw a diagram of all the layers and the tensor input and output sizes at each layer, with zoom in/out capabilities if needed using D3.js or whatever visualization framework if needed. Every single layer should be on there with its input and output sizes…
Sounds like a solid SaaS business plan!