TADA: Speech generation through text-acoustic synchronization

boxed wrote: post #4

OutOfHere wrote: post #3

Will this run on CPU? (as opposed to GPU)

Why would you want to? It's like using a hammer for screws.

To maximise the VRAM available for an LLM on the same machine. That's why I asked myself the same question, anyway.

Re: TADA: Speech generation through text-acoustic synchronization

#7

Post by g-mork » Wed, Mar 11, 2026, 9:20 AM UTC

boxed wrote: post #4

OutOfHere wrote: post #3

Will this run on CPU? (as opposed to GPU)

Why would you want to? It's like using a hammer for screws.

CPU compute is infinity times less expensive and much easier to work with in general

Re: TADA: Speech generation through text-acoustic synchronization

#8

Post by boxed » Wed, Mar 11, 2026, 9:30 AM UTC

g-mork wrote: post #7

boxed wrote: post #4

Earlier quoted context omitted.

Why would you want to? It's like using a hammer for screws.

CPU compute is infinity times less expensive and much easier to work with in general

Less expensive how? The reason GPUs are used is because they are more efficient. You CAN run matmul on CPUs for sure, but it's going to be much slower and take a ton more electricity. So to claim it's "less expensive" is weird.

Re: TADA: Speech generation through text-acoustic synchronization

#9

Post by theturtle » Wed, Mar 11, 2026, 9:31 AM UTC

[dead]

Re: TADA: Speech generation through text-acoustic synchronization

#10

Post by octoclaw » Wed, Mar 11, 2026, 10:02 AM UTC

[dead]