@curlysaarthak well everyone has a different philosophy of life. we all optimizing for different objective functions. even if 2 people end up at the same minima the gradient steps they take can be entirely different so i hope we both get there buddy.
Data hunger is governed less by parameter count alone and more by the inductive biases baked into the model, training objective, and input representation.
I just realized something. There are many people these days who intentionally post racist stuff about Indians calling them all kinds of names. People think its just personal and that its their personal hate towards Indians but there are bigger things at play here than personal hate. It has nothing to do with personal hate.
These people took “no publicity is bad publicity” seriously. They know that if they post hateful, racist stuff against Indians they will get a lot of impressions whether its bad comments, threats or even some people agreeing with them. They know they will get all this because there are a lot of Indians on the internet these days.
They know they will initially get hate but like anything on the internet it settles down. After all the hate and the storm what remains is a lot of impressions on that account and a lot of followers as well and priority seat in the algorithm. Apparently brands start doing deals with them because some brands dont give a fuck about the image of the creator. They care about how many eyes that creator can bring. Its a classic replica of the tech bubble crash. In the aftermath of the crash companies might die but the technology from the bubble remains like the .com boom and it keeps happening. And its not just with people who spread hate against Indians. Its also with people in India who create controversies and you would be surprised to know that a lot of these X accounts abusing Indians are just some dudes sitting in India.
Its just funny how gaining a few thousand followers suddenly makes people have the notion that they have figured everything out. every day i open x and theres a new "thought leader" handing out universal truths.
most philosophers of their time werent celebrated. many were ignored, ridiculed or understood only long after they were gone. popularity has never been a reliable measure of wisdom. fame alwys follows stupidity because the average audience IQ is low so lot of people can relate to it and thats exactly why it becomes so popular.
silk mulberry 1.5 is built to be versatile
with a simple voice description and utterance, it can generate natural speech across 20+ languages, with accents and emotions
this thread is a collection of sample prompts to help you explore what this looks like in practice.
@maharshii i mean u might be right but like idk if its entirely a bad thing because like every engineering students thinks the courses outside their specialisation are useless. then they graduate and discover the world isnt just their branch.
mulberry can:
- respond in 162ms, the fastest ttfb on the market
- switch between two languages mid-sentence, naturally
- speak in different languages with their actual native accents, not a translated voice
launching silk mulberry 1.5
one of the fastest multilingual voice models in the world
it matches the best voice models in quality benchmarks (MOS)
all this at more than 95% lower cost ₹0.40/min (~$0.0046/min)
try now 👇
launching silk mulberry 1.5
one of the fastest multilingual voice models in the world
it matches the best voice models in quality benchmarks (MOS)
all this at more than 95% lower cost ₹0.40/min (~$0.0046/min)
try now 👇
45K Followers 3K FollowingChoose disfavour where obedience does not bring honour.
I do math. And was once asked by R. Morris Sr. : "For whom?"
@[email protected]
2K Followers 949 FollowingCofounder and Managing Partner at Grep VC, investing in AI and Robotics. Led WHO COVID-19 App. Led engineering on Google Earth and Enterprise Gmail.
215 Followers 4K FollowingSenior Data Scientist, ServiceNow. Ex Microsoft.Mechanical Engineer.Blogger. Avid Reader. Movie buff.Eternal Dreamer. Finds joy in those small things.
7K Followers 171 FollowingScientist at @GoogleDeepMind, leading Blueshift team doing science and reasoning at Gemini. Previously physics @Stanford. Directionally correct.
841 Followers 409 FollowingI love science, technology and engineering, I'm fascinated by the speed and power of innovation and i can't wait for the future! ML Researcher
4K Followers 127 FollowingCEO & Co-Founder of EquiLibre Technologies. Previously Senior Research Scientist at Google DeepMind. Co-author of DeepStack and Player of Games.
9K Followers 409 FollowingPhilosopher in tech, currently at @MistralAI. Doctor of talking machines, now teaching them good behavior and multiple languages.
6K Followers 239 FollowingAI Researcher: From theory to practice (and back)
Postdoc @MetaAI with @ylecun
PhD @RiceUniversity with @rbaraniuk
Masters @ENS_Ulm @Paris_Sorbonne