[HN Gopher] DeepSpeech Is Discontinued (2020)
___________________________________________________________________
DeepSpeech Is Discontinued (2020)
Author : LorenDB
Score : 37 points
Date : 2025-06-25 17:15 UTC (5 hours ago)
(HTM) web link (github.com)
(TXT) w3m dump (github.com)
| dabinat wrote:
| It was discontinued 5 years ago - I'm not sure why it took so
| long to archive the repo.
|
| https://discourse.mozilla.org/t/future-of-deepspeech-stt-aft...
| HelloUsername wrote:
| Perhaps OP URL can be changed to
| https://www.phoronix.com/news/Mozilla-DeepSpeech-Discontinue...
| LorenDB wrote:
| I'd change it, except this submission is past the edit
| window. Perhaps dang or tomhow will see this and change it
| for me :)
| altairprime wrote:
| You should email them about that!
| HelloUsername wrote:
| Then comments should be moved hither:
| https://news.ycombinator.com/item?id=44380572
| altairprime wrote:
| If you email that to the mods, they'll do so :)
| xeonmc wrote:
| Imagine an alternate timeline where Mozilla had named this
| project "FreeSpeech" instead as a free and open TTS solution.
| echelon wrote:
| My personal little conspiracy theory is that Google pays
| Mozilla as an antitrust shield.
|
| A lot of us pretty much assume that much, but I think it goes
| much deeper.
|
| I think Google pays and maintains a working relationship with
| the CEO of Mozilla (current and former) to purposely keep the
| organization rudderless, uncompetitive, and shrinking.
|
| Mozilla spends its money building a 3D VR metaverse here, a
| bunch of AI models it later scraps over there, a web3 /
| distributed social program, etc. It scraps Rust, doesn't invest
| into Firefox. Just silly toys and experiments.
|
| That nice CEO salary is hush money.
|
| Just a fun little pet theory, totally not based on evidence.
| Teever wrote:
| I don't know if there's a direct quid pro quo relationship
| between the CEO of Mozilla and Google but I feel quite
| confident that Google absolutely influences the organization
| in ways beyond just the cash injection to make it rudderless,
| uncompetitive and shrinking as you say.
|
| It could be as simple as ex-Google employees at lower levels
| than CEO who are paid by Google take positions at Mozilla, or
| more subtle things like guiding the direction of the
| organization through standards boards.
|
| It would be really fascinating to look at the org charts of
| Mozilla past and present and try and build the network
| between people who worked at Google or Google related
| organizations before, during, and after their time at
| Mozilla.
|
| Because you're absolutely right that the organization is so
| absolutely dysfunctional that it can't just be incompetence,
| it has to be absolute malice.
| Devorlon wrote:
| Yes simple things such as becoming employed at Mozilla to
| perform corporate espionage for Google.
| CamouflagedKiwi wrote:
| Has a nice sound to it but Hanlon's razor says: never
| attribute to malice that which can be adequately explained by
| incompetence. I'd be impressed if this was the case but I
| expect neither side is sufficiently competent or malicious to
| explain it.
| ipsum2 wrote:
| I've been using Nvidia's parakeet model, it's been better than
| Whisper v3 large and smaller. Only supports English.
|
| https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2
| nico wrote:
| Does it need a newer GPU? Or can it run on just CPU?
|
| Would it run on a raspberry pi?
| GaggiX wrote:
| Look up for faster whisper or distilled whisper models,
| smaller models run quite nicely but perform poorly outside of
| English, if you are interested in a different language it's
| better to finetune it (HuggingFace has a huge amount of
| finetuned Whisper models).
| ipsum2 wrote:
| If you want real-time, it requires a GPU, but can be
| underpowered. CPU is a little slower but works fine.
| lupusreal wrote:
| Best CPU TTS that can run on something like a raspberry pi is
| Piper. It can do real time synthesis on a raspberry pi and on
| a real computer it runs several times faster with negligible
| performance cost. I use it for 'reading' ebooks when my eyes
| get tired. The quality is roughly on par with where Mac OS's
| TTS was ~10 years ago (the last time I used it.) You can tell
| it's TTS, but it's good enough that you can become accustomed
| to it fairly easily.
|
| https://github.com/rhasspy/piper
| GaggiX wrote:
| They are talking about STT, not TTS, but as a TTS piper is
| very good and works nicely on a raspberry pi, I agree.
| 100xlong wrote:
| are there any linux/mac apps that allow people to use parakeet
| for daily dictation like SuperWhisper?
| msgodel wrote:
| I still prefer festival, it's fast, it's in all the package
| repos, and I don't like automations having realistic voices.
| i80and wrote:
| They're opposites: DeepSpeech is speech to text, where Festival
| is TTS
| quentindanjou wrote:
| It seems that the team that used to work on DeepSpeech then
| worked on coqui-ai STT https://github.com/coqui-ai/STT and now
| recommends using OpenAI Whisper
| (https://github.com/openai/whisper)
___________________________________________________________________
(page generated 2025-06-25 23:01 UTC)