Solopreneur | indie hacker
Kochi, Kerala
Joined April 2010
- Tweets6K
- Following1.4K
- Followers310
- Likes5.1K
Pinned Tweet
"What is an ocean, but a multitude of drops"
youtube.com/watch?v=lxumQYpB…
Krishna Sangeeth KS retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
Wow. There is a coreml model. Definitely worth trying. Decently expressive voice for the model size.
github.com/gradium-ai/xn-ptt…
A new frontier for on-device speech.
We're open-sourcing Phonon, a suite of TTS models under 50m parameters, setting a new standard for naturalness and accuracy across languages, with the broadest inference support to date.
Phonon-31m is available under Apache 2.0.
gradium.ai/phonon🧵
Krishna Sangeeth KS retweeted
Most of these tricks are not specific to speech and should carry over to other transformers. The full recipe, with ablations, is in the blog post: kyutai.org/blog/2026-10-20-p…
Krishna Sangeeth KS retweeted
Pocket TTS has been getting updates those last few days. While it was ~100M parameters, we release smaller versions, down to 1.9M: an 18.2M model with the quality of the original, a 6.9M model, and a 1.9M model that runs 103x faster than real time on one CPU core. All with voice cloning.
GemmaEmbedding2 model is a really good backbone to build other models.
I built a Malayalam ASR model which is really fast. The current bottleneck is amount of data and once I can scale it further, I am sure WER would come down.
huggingface.co/whiletruelear…
Great read on the topic.
IndiaAI Mission: 9 model builders get Rs 1,058 cr of govt’s compute subsidy, researchers only Rs 10 cr
Vrinda Tulsian @vrinda_tulsian reports for ThePrint
Illustration: Shruti Naithani @shrru_art
theprint.in/tech/indiaai-mis…
Imagine this developer workflow:
1. Register your existing UPI ID, such as myname@upibank.
2. Authorize access through your bank.
Register dummy-app.com/webhook.
3. Receive signed events whenever someone sends money to that UPI ID.
No payment gateway onboarding, no dependency on a specific checkout flow, and no SMS scraping needed then. There is clearly demand for this workflow as most payments go through UPI now.
Charge a small amount for this and use this money to subsidise UPI infra for everyone else. @NPCI_NPCI
pls checkout Rea + Hopper Disassembler. Hopper comes with a demo mode that lasts a very short time. It's crazy what's possible.
Krishna Sangeeth KS retweeted
Over the past month, we completely switched to frontier LLMs to control a robot in a customer deployment.
By doing this, we have a classic System 1/System 2 architecture, but on steroids.
For the high-level reasoning layer, we use an agent based on Opus and Astra, which are many generations more capable than any VLA. For the fast execution layer, we use computer vision code written by the model, which is fast, editable, and reliable.
As a bonus, we got in-context learning, which is just a basic feature of LLMs, and continual learning, since whenever a new situation arises, the large model steps in and adapts the pipeline.
And it just feels better to be on the right side of the Bitter Lesson, benefiting from the billions being invested in frontier models around the world.
Awesome 🙌
We at @ai4bharat, with @NVIDIA, are releasing open data and evals to build frontier LLMs for all 22 Indian languages.
- 2.8T tokens of synthetic Indic pretraining data
- Audited, native-speaker-verified benchmarks
- Open recipes
📖 Tech Report: tinyurl.com/4rpt383x
(1/6)
"I don't think we will have it in just one giant brain"
This is a realistic picture of AGI instead of one entity that can do everything. There is more recognition that there are going to be specialised tools or models that are extremely good at at a narrow task and the larger general purpose models would do tool calling.
Demis Hassabis on choosing what’s worth spending your life building in a world changing this fast:
“Going after hard problems and deep problems, it’s no more difficult in some ways than going after a shallower, simpler, more superficial problem. They’re just differently difficult.”
“Given life’s very short and you only have so much time and energy, you might as well put your life force into something that will really make a difference.”
“If you start off on a deep tech journey today... you’re talking about a 10-year journey for true deep tech. So now you have to consider AGI appearing in the middle of that journey.”
“Imagine what that world would look like and then build something that would be useful if that comes in halfway through.”