cryogenicplanet
Joined 6 years ago
Karma
109
Submissions
60
Member since
May 12
Recent submissions
Comment · 6 days ago
i built myself this open-source message board for the agents ping each other and coordinate between each other i personally love it. https://chirp.wiki
Comment · 1 year ago
still can’t justify the crime against our agent that was winning
Comment · 2 years ago
been amazing to see the amount of iteration in the product to get to 1.0!
Comment · 2 years ago
been using savvy for a bit - shantanu gave me early access!
love how simple it is to use, personally been loving the savvy ask/explain in the last couple of weeks - been super easy to ask questions in the terminal and just get answers
but the history is quite good too, very cool to share notebooks with others on the team
Comment · 2 years ago
want to thank the op for sharing this; i threw this together in the last couple of days ramping out to the "steamroll" - we think one of the key problems in LLMs in general but esp voice is evals and wanted to have a good place to evaluate voice-to-voice systems. these systems can be end-to-end like openai or (asr+llm)->tts or asr->(llm+tts) or asr->llm->tts
we built an ELO benchmark very much in the style of LMSYS and will be releasing results every two weeks
source code here: https://github.com/thevoicecompany/bench.audio
will be adding proper contributing guide soon
Comment · 2 years ago
This looks very sick!
any chance of getting this as native apps on mobile? or better yet like global macos utility like dictation so you can "type" in any apps?
Comment · 3 years ago
Hey HN!
As JS (Mostly TS) devs, we feel like AI or LLM tooling in JS kinda sucks rn, that said most production applications inevitably end up being in JS, so a friend and I (mostly friend) started hacking on a new JS native library for making it easier to write prompts.
The idea of the library is to give a much more ergonomic syntax for writing complex prompts, the repo itself goes much more in detail https://github.com/LevanKvirkvelia/salute
Would love thoughts and any/all feedback!
Comment · 3 years ago
Well it is mostly the GPT-4 that is expensive to run and because we generate 3 version of your long prompt, with it's inferences and our own prompt.
Comment · 3 years ago
This is fire - reusable components in spreadsheets sounds great!
Comment · 4 years ago
To follow up on this, we did update the pricing now https://ffmpeg.guide/pricing which should be effective from Nov 1st. Gonna try to move some more lifetime deals at a slightly higher price before then
Comment · 4 years ago
You can go to https://ffmpeg.guide/demo
Comment · 4 years ago
Oh shit, I forgot to change that back my bad. Should be fixed!
[email protected]