Dokyoon

leeloolee

·

Eruly

AI & ML interests

ai

Recent Activity

upvoted a collection 7 days ago

reacted to medmekk's post with 🚀 13 days ago

🚀 Introducing Halo 1.0 Today, we are open-sourcing Halo, the training framework we use to train every model at White Circle. It comes with: 🧠 Full post-training stack: SFT, DPO/KTO/SMPO, reward modeling, GRPO, distillation 🤖 Async multi-turn RL with vLLM/SGLang rollouts and sandboxed tool use ⚡ ~2.8× TRL throughput on 8× B300 (EP+FSDPv2, FA4, fp8/fp4) 🤗 Dense HF models + 15 MoE families (Qwen, GLM, Mistral, DeepSeek-V4…) 🛠️ One halo command, prebuilt Docker images, and docs for humans and agents 💻 https://github.com/whitecircle/halo Try it and tell us what you're training

liked a model 14 days ago

yandex/AliceAI-Foundation-80B-A3B-Base

View all activity

Organizations

leeloolee 's datasets 2

leeloolee/arxiv-cs-metadata-enriched

Viewer • Updated May 19 • 423k • 113

leeloolee/mdpo

Updated Oct 23, 2024 • 65 • 1