Roy Huang

EECS @ Berkeley / ai researcher / Agentica

[Roy Huang @_royh021]

@overleaf Even the status page is down 😂

[Roy Huang @_royh021]

@ceeeengo @overleaf HOW???????

[Roy Huang @_royh021]

@overleaf pls fix i need to submit my paperrrrrrr

[Roy Huang @_royh021]

@ManasSharma07 @overleaf its down for me too

[Roy Huang @_royh021]

@SpotifyStatus its backk!!!!

[Agentica Project @Agentica_]

We're trending on @huggingface models today! 🔥

Huge thanks to our amazing community for your support. 🙏

[Roy Huang @_royh021]

@StringChaos @1jaskiratsingh @slimshetty_ @istoica05 Wow, this is amazing work!

[ollama @ollama]

ollama run deepcoder

🫡 let’s go open source!!

[Roy Huang @_royh021]

@ollama Thank you for featuring us!

[Rohan Paul @rohanpaul_ai]

DeepCoder: A Fully Open-Source 14B Coder at O3-mini Level

From @togethercompute

[Roy Huang @_royh021]

@Agentica_ 🔥🔥🔥

[Roy Huang @_royh021]

@michaelzluo deepcoder might be the goat 🐐

[Roy Huang @_royh021]

@sijun_tan @sama deepcoder 🔛🔝

[Roy Huang @_royh021]

😱14B Fully Open Source o3-mini level model?? 😱

DeepCoder is a small but mighty reasoning model that is o3-mini and o1 level on coding and math that runs at a tiny fraction of the cost. Another W by open source. Go check it out!!

[Agentica Project @Agentica_]

Introducing DeepCoder-14B-Preview - our fully open-sourced reasoning model reaching o1 and o3-mini level on coding and math.

The best part is, we’re releasing everything: not just the model, but the dataset, code, and training recipe—so you can train it yourself!🔥

Links below:

[Roy Huang @_royh021]

@Agentica_ Another win for open source!

[Agentica Project @Agentica_]

✨RL magic is in the air! Introducing DeepScaleR-1.5B-Preview—a fully open-source, 1.5B-parameter model trained with RL to surpass o1-preview for general math reasoning.

📜Blog: https://t.co/eHqApwRfnH
💻Github: https://t.co/tRsDN7xV4M

[Shishir Patil @shishirpatil_]

📢Excited to release GoEx⚡️a runtime for LLM-generated actions like code, API calls, and more. Featuring "post-facto validation" for assessing LLM actions after execution 🔍 Key to our approach is "undo" 🔄 and "damage confinement" abstractions to manage unintended actions & risks. This paves the way for fully autonomous LLM agents, enhancing interaction between apps & services with human-out-of-loop🚀
Blog: https://t.co/vZVW7LQ1aB
Code: https://t.co/sq05KMrlw9
Paper: https://t.co/Dw28GCmKm0

[Joey Gonzalez @profjoeyg]

My students @shishirpatil_ and @tianjun_zhang and their undergrads @_royh021 and @fanjia_yan just presented some exciting new features for #GorillaLLM at Sky Camp https://t.co/rYnFKQnKcd.

Now you can LoRA fine-tune and get open function calling from one place.

[Roy Huang @_royh021]

@0xazusa i miss being little and going sledding and being cold resistant

[Roy Huang @_royh021]

@0xazusa I got so confused for a hot second cuz I tried to read it literally lol