Ramchand Kumaresan

This account is verified. Learn more

@Mechramc

Exploring ML/AI and local AI research. Author of Under the Hood (Building LLMs) - love தமிழ் and building தமிழ் models!

Austin, TX ramchandk.com Joined November 2020

1.2K Following 4.1K Followers 1.1K Posts

Latest posts 5

Pinned post

Ramchand Kumaresan @Mechramc Oct 28

Agaram 2.5B is out. It's a Tamil-first language model we trained from scratch at @MuraiLabsHQ , our small non-profit lab. As far as we know, it's the first Tamil model trained from scratch that can call tools.

It chats in Tamil, Tanglish and English, translates, and can call tools. We're releasing the base model, chat model and GGUF files under Apache-2.0. You can download it and run it offline on your own laptop.

The first chat model had a four-day public preview. Seeing where it fell short with real users helped us decide what to work on. We rebuilt the fine-tuning around what we saw, using 222,662 records.

Thank you to everyone who spent time with that early version.

The video walks through the model. You can try it here:
hf.co/spaces/mechram…

நன்றி 🙏

2:02

Ramchand Kumaresan @Mechramc Nov 6

Agaram 2.5B has been out about a week. If you ran it (Ollama, GGUF, or the chat demo), what is the one failure that still blocks you? Reply with the fail and your setup. I will use the replies to prioritize the next fixes. Thank you again for your support!

Ramchand Kumaresan @Mechramc Nov 5

@0xAbhi__ Thank you for this instagram.com/reel/DfWAUeNTL…

I have been down sick and have not had a chance to do more to promote Agaram! But this was really cool! Thank you!

www.instagram.com Abhi (@0xabhi) • Instagram reel

Ramchand Kumaresan @Mechramc Oct 29

A small fix for anyone who ran Agaram 2.5B in Ollama and got nonsense: our GGUF repo held the base and chat files under the same names, so the default command loaded the base model, which can't chat.
Now fixed. If you pulled it before today, remove it and pull again:
ollama run hf.co/murai-labs/aga…
Thanks to @TVKFailsTN for reporting it.
Chat GGUF (llama.cpp, Ollama): huggingface.co/murai-labs/aga…
Base GGUF (new repo): huggingface.co/murai-labs/aga…
Chat model: huggingface.co/murai-labs/aga…
Demo: huggingface.co/spaces/mechram…
Paper: arxiv.org/abs/2610.18420
Code: github.com/Murai-Labs/Aga…

Ramchand Kumaresan @Mechramc Oct 29

Thank you so much for all the love and support! It has been a wonderful 24 hours and I am truly humbled. And to celebrate that - here is our launch video - now in Tamil.

Try it: huggingface.co/spaces/mechram…
Chat model: huggingface.co/murai-labs/aga…
Base model: huggingface.co/murai-labs/aga…
GGUF (llama.cpp / Ollama): huggingface.co/murai-labs/aga…
Paper: arxiv.org/abs/2610.18420
Code: github.com/Murai-Labs/Aga…

Apache-2.0. Feedback in Tamil or Tanglish welcome. Original launch post: x.com/Mechramc/statu…

2:08

Ramchand Kumaresan Verified account @Mechramc Oct 28

Agaram 2.5B is out. It's a Tamil-first language model we trained from scratch at @MuraiLabsHQ , our small non-profit lab. As far as we know, it's the first Tamil model trained from scratch that can call tools.

It chats in Tamil, Tanglish and English, translates, and can call tools. We're releasing the base model, chat model and GGUF files under Apache-2.0. You can download it and run it offline on your own laptop.

The first chat model had a four-day public preview. Seeing where it fell short with real users helped us decide what to work on. We rebuilt the fine-tuning around what we saw, using 222,662 records.

Thank you to everyone who spent time with that early version.

The video walks through the model. You can try it here:
hf.co/spaces/mechram…

நன்றி 🙏

View quoted post on X