Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Dia, an open-weights TTS model for generating realistic dialogue

github.com

Post reply on HN
Print view View original

181–190 of 202 posts

Re: Show HN: Dia, an open-weights TTS model for generating realistic dialogue

#181

Post by xdfgh1112 »

throwaway-alpha wrote: post #162
sebstefan wrote: post #157

I inserted the non-verbal command "(pauses)" in the middle of a sentence and I think I caused it to have an aneurysm. https://i.horizon.pics/4sEVXh8GpI (27s) It starts with an intro, too. Really strange

I have a hunch they're pulling data from radio shows to give it that "high quality" vibe. Tried running it through this script and hit some weird bugs too: [S1] It really sounds as if they've started using NPR to source TTS models [S2] Yeah... yeah... it's kind of disturbing (laughs dejectedly). [S3] I really wish, that they would just Stop with this. https://i.horizon.pics/Tx2PrPTRM3

Just noticed he says dejectedly too

Re: Show HN: Dia, an open-weights TTS model for generating realistic dialogue

#182

Post by genewitch »

Havoc wrote: post #73
yjftsjthsd-h wrote: post #71

Earlier quoted context omitted.

> Does mean it pulls the model every time on startup though which isn't ideal Surely it just downloads to a directory that can be volume mapped?

Yep. I just didn't spend the time to track down the location tbh. Plus huggingface usually does links to a cache folder that I don't recall the location of Literally got cuda containers working earlier today so haven't spent a huge amount of time figuring things out

Its in a dot folder in your home dir on Linux and in %appdata% on windows.

Re: Show HN: Dia, an open-weights TTS model for generating realistic dialogue

#183

Post by genewitch »

jelling wrote: post #77
dindindin wrote: post #75

Was this trained on Planet Money / NPR podcasts? The last audio (continuation of prompt) sounds eerily like Planet Money, I had to double check if my Spotify had accidentally started playing.

NPR voice is a thing. It started with Ira Glass voice and now the default voice is someone that sounds like they're not certain they should be saying the very banal thing they are about to say, followed by a hand-shake protocol of nervous laughter.

Thank goodness for Scott Simon!

Re: Show HN: Dia, an open-weights TTS model for generating realistic dialogue

#184

Post by xdfgh1112 »

dangoodmanUT wrote: post #93
toebee wrote: post #2

Hey HN! We’re Toby and Jay, creators of Dia. Dia is 1.6B parameter open-weights model that generates dialogue directly from a transcript. Unlike TTS models that generate each speaker turn and stitch them together, Dia generates the entire conversation in a single pass. This makes it faster, more natural, and easier to use for dialogue generation. It also supports audio prompts — you can condition the output on a spec…

I know it’s taboo to ask, but I must: where’s the dataset from? Very eager to play around with audio models myself, but I find existing datasets limiting

I suspect podcasts, as you have a huge amount of transcribed data with good diction and mic quality. The voices sound like podcast voices to me.

Re: Show HN: Dia, an open-weights TTS model for generating realistic dialogue

#186

Post by ivape »

toebee wrote: post #108
ivape wrote: post #16

Earlier quoted context omitted.

No doubt, these TTS models locally are what I'm looking for because I'm so done typing and reading :)

You can try it now on https://huggingface.co/spaces/nari-labs/Dia-1.6B !!!

Woah, you guys are on it!

Re: Show HN: Dia, an open-weights TTS model for generating realistic dialogue

#187

Post by devnen »

This is really impressive work and the dialogue quality is fantastic.

For anyone wanting a quick way to spin this up locally with a web UI and API access, I put together a FastAPI server wrapper around the model: https://github.com/devnen/Dia-TTS-Server

The setup is just a standard pip install -r requirements.txt (works on Linux/Windows). It pulls the model from HF automatically – defaulting to the faster BF16 safetensors (ttj/dia-1.6b-safetensors), but that's configurable in the .env. You get an OpenAI-compatible API endpoint (/v1/audio/speech) for easy integration, plus a custom one (/tts) to control all the Dia parameters. The web UI gives you a simple way to type text, adjust sliders, and test voice cloning. It'll use your CUDA GPU if you have one configured, otherwise, it runs on the CPU.

Might be a useful starting point or testing tool for someone. Feedback is welcome!

Re: Show HN: Dia, an open-weights TTS model for generating realistic dialogue

#188

Post by amp-lifier »

toebee wrote: post #2

Hey HN! We’re Toby and Jay, creators of Dia. Dia is 1.6B parameter open-weights model that generates dialogue directly from a transcript. Unlike TTS models that generate each speaker turn and stitch them together, Dia generates the entire conversation in a single pass. This makes it faster, more natural, and easier to use for dialogue generation. It also supports audio prompts — you can condition the output on a spec…

In terms of guiding voice and expression, audio prompts are promising but I believe text instructions serve different experiences as well. Will there be support for that as well?

Re: Show HN: Dia, an open-weights TTS model for generating realistic dialogue

#189

Post by sebstefan »

degosuke wrote: post #179
throwaway-alpha wrote: post #162

Earlier quoted context omitted.

I have a hunch they're pulling data from radio shows to give it that "high quality" vibe. Tried running it through this script and hit some weird bugs too: [S1] It really sounds as if they've started using NPR to source TTS models [S2] Yeah... yeah... it's kind of disturbing (laughs dejectedly). [S3] I really wish, that they would just Stop with this. https://i.horizon.pics/Tx2PrPTRM3

It even added an extra f-word at the end. Still veeery impressive

Oh brother imagine using this and noticing 4 months in that it randomly throws f-words for flavor at your users
Post reply on HN