☆ Yσɠƚԋσʂ ☆@lemmy.ml to Technology@lemmy.mlEnglish · 9 hours agoQwen-Audio-3.0-TTSfunaudiollm.github.ioexternal-linkmessage-square3fedilinkarrow-up19arrow-down10
arrow-up19arrow-down1external-linkQwen-Audio-3.0-TTSfunaudiollm.github.io☆ Yσɠƚԋσʂ ☆@lemmy.ml to Technology@lemmy.mlEnglish · 9 hours agomessage-square3fedilink
minus-squareloathsome dongeater@lemmygrad.mllinkfedilinkEnglisharrow-up1·7 hours agoSorry but what kind of specs are needed to run something like this? Is it more or less demanding than a 8B moderately quantised LLM?
minus-square☆ Yσɠƚԋσʂ ☆@lemmy.mlOPlinkfedilinkarrow-up3·7 hours agoyeah the models are tiny, just 2b https://huggingface.co/collections/Qwen/qwen3-tts
Sorry but what kind of specs are needed to run something like this? Is it more or less demanding than a 8B moderately quantised LLM?
Try out CrispASR.
yeah the models are tiny, just 2b https://huggingface.co/collections/Qwen/qwen3-tts