Hugging Face
Models
Datasets
Spaces
Buckets
new
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
5.2
TFLOPS
Denis Shilov
mixeden
1
2
37
Follow
dmitriihook's profile picture
lenacircle's profile picture
Igralino's profile picture
12 followers
ยท
16 following
mixedenn
mixeden
AI & ML interests
None yet
Recent Activity
reacted
to
medmekk
's
post
with ๐
4 days ago
๐ 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: ๐ง Full post-training stack: SFT, DPO/KTO/SMPO, reward modeling, GRPO, distillation ๐ค Async multi-turn RL with vLLM/SGLang rollouts and sandboxed tool use โก ~2.8ร TRL throughput on 8ร B300 (EP+FSDPv2, FA4, fp8/fp4) ๐ค Dense HF models + 15 MoE families (Qwen, GLM, Mistral, DeepSeek-V4โฆ) ๐ ๏ธ One halo command, prebuilt Docker images, and docs for humans and agents ๐ป https://github.com/whitecircle/halo Try it and tell us what you're training
liked
a model
10 days ago
whitecircle/GLM-4.7-Flash-Coder
upvoted
an
article
2 months ago
Security incident disclosure โ July 2026
View all activity
Organizations
mixeden
's models
None public yet