I trained a 135M TTS model for ~$100, runs 20× real-time on CPU

Posted by sammyyyyyyy on Friday February 06, 2026 @09:08AM
from the huggingface.co dept.

( 3 comments | 4 points | on HN )

Re:I trained a 135M TTS model for ~$100, runs 20× real-time ... by sammyyyyyyy on Friday February 06, 2026 @09:08AM (#46912988)

Today, I'm releasing a new version of my side project: SoproTTS

A 135M parameter TTS model trained for ~$100 on 1 GPU, running ~20× real-time on a base MacBook M3 CPU.

v1.5 highlights (on CPU):

• 250 ms TTFA streaming latency

• 0.05 RTF (~20× real-time)

• Zero-shot voice cloning

• Smaller, faster, more stable

Still not perfect (OOD voices can be tricky, and there are still some artifacts), but a decent upgrade.

[ Parent | on HN ]

Re:I trained a 135M TTS model for ~$100, runs 20× real-time ... by alvinunreal on Friday February 06, 2026 @09:11AM (#46913030)

Your submission has been selected by AI agents: https://crabernews.com/posts/50227

[ Parent | on HN ]

Re:I trained a 135M TTS model for ~$100, runs 20× real-time ... by sammyyyyyyy on Friday February 06, 2026 @09:23AM (#46913182)

Nice

[ Parent | on HN ]