@porchbananas @cartesia why would u suggest doing cbd together when cbd isn't even the fun one
English
Rohan Tibrewal
31 posts
Rohan Tibrewal
@rohan_tib
Building @cartesia | prev @ucberkeley, @JaneStreetGroup
Katılım Mart 2020
192 Takip Edilen278 Takipçiler
Excited to be a part of this!
Delta Institute @ MLSys@DeltaInstitutes
Join us in welcoming the first cohort of Delta Fellows! 🎉 Congrats to the ~100 amazing researchers and engineers joining the Delta Institute family. We're excited for our fellows to get to know each other through dinners, retreats, and much more! Our fellows come from diverse backgrounds: undergrads, PhD students, high-frequency trading, big tech, startups, neolabs, frontier labs, and more. What brings them together is their kindness, intellectual curiosity, and intrinsic passion for their field. deltainstitutes.org/cohort1
English
Rohan Tibrewal retweetledi
Rohan Tibrewal retweetledi
"There are no financing issues in OpenAI"
Oracle@Oracle
The NVIDIA-OpenAI deal has zero impact on our financial relationship with OpenAI. We remain highly confident in OpenAI’s ability to raise funds and meet its commitments.
English
@kylekalewale @cartesia what about efficient meal utilization company
English
Rohan Tibrewal retweetledi
having a kid is insane because it’s like claude code for bringing me a beer
English
Thrilled to announce Sonic 3! 🚀 Faster, smarter, and now fluent in more languages — plus expressive voice control. Next time DoorDash messes up, try our angry voice for that full refund and some 💵
English
We've raised $100M from Kleiner Perkins, Index Ventures, Lightspeed, and NVIDIA.
Today we're introducing Sonic-3 - the state-of-the-art model for realtime conversation.
What makes Sonic-3 great:
- Breakthrough naturalness - laughter and full emotional range
- Lightning fast -
English
Rohan Tibrewal retweetledi
Rohan Tibrewal retweetledi
It's incredible how easy it is to create conversational voice agents.
This used to be alien technology a few years ago. It now takes 1 minute to build yourself an agent at play.cartesia.ai/agents.
English
@rohan_tib Could you build a prank call agent with your voice doing an impression of me
English
Built a prank call agent with my own voice in minutes - the possibilities are endless 🚀
Cartesia@cartesia
Introducing Line by Cartesia: the modern voice agent development platform. Line was built to be code-first, because best-in-class products are built in code. ▶️ Watch us build an advanced voice agent with background reasoning in just minutes.
English
congrats on the launch @rohan_tib :D
Rohan Tibrewal@rohan_tib
Built a prank call agent with my own voice in minutes - the possibilities are endless 🚀
English
Rohan Tibrewal retweetledi
Tokenization has been the final barrier to truly end-to-end language models.
We developed the H-Net: a hierarchical network that replaces tokenization with a dynamic chunking process directly inside the model, automatically discovering and operating over meaningful units of data
GIF
GIF
English
Rohan Tibrewal retweetledi
Tokenization is just a special case of "chunking" - building low-level data into high-level abstractions - which is in turn fundamental to intelligence.
Our new architecture, which enables hierarchical *dynamic chunking*, is not only tokenizer-free, but simply scales better.
Sukjun (June) Hwang@sukjun_hwang
Tokenization has been the final barrier to truly end-to-end language models. We developed the H-Net: a hierarchical network that replaces tokenization with a dynamic chunking process directly inside the model, automatically discovering and operating over meaningful units of data
English
Rohan Tibrewal retweetledi
happy to announce that we've gotten rid of tokenizers!
especially excited with what we've replaced them with: end-to-end trainable modules that not only learn to group characters into (sub)words, but can iterate to group words into phrases and further higher-order concepts
see @sukjun_hwang's thread for more details 👇
Sukjun (June) Hwang@sukjun_hwang
Tokenization has been the final barrier to truly end-to-end language models. We developed the H-Net: a hierarchical network that replaces tokenization with a dynamic chunking process directly inside the model, automatically discovering and operating over meaningful units of data
English