2025 02 10 todo dhwani

Todov- Dhwani - Feb 10, 2025

Move code to fastapi

Create basic ux / single page for deno

Deploy with hf spaces ?

Learn about about docket container orchestration.

---

Use batch streaming for speaker audio. To get same speaker voice.

Combine all generation for each voice at once.

https://github.com/huggingface/parler-tts/blob/main/INFERENCE.md#batch-generation

---

Balance between creating demo video and the demo itself has become nerve-wracking. Last week, it was just an idea. In 7 days, poc has transformed into production grade system. UX is the last stretch. Building an Immersive Audiobook Generator with Text/Speech prompts. slabstech.com/dhwani

--

Usecase

Blind and illiterate users can talk to books and learn From a lot of material that is not available for them.

Braille books can be replaced by speech to Speech. They can interact with Voice and learn new things.

Cost of access will be reduced significantly

---

Meta Ai - hypersclaer

https://www.cs.cmu.edu/~dskarlat/publications/xfaas_sosp23.pdf

https://cacm.acm.org/research/metas-hyperscale-infrastructure-overview-and-insights/

---

Hackathon- print qr code. Scan project idea

Pitch presentation