SGLang-native diffusion transformer overrides converted with NVIDIA ModelOpt.
AI & ML interests
LLM, distributed systems
Recent Activity
View all activity
Organization Card
The large model systems organization (LMSYS) develops large models and systems that are open accessible and scalable.
Learn more about us at https://lmsys.org.
A collection of production-grade draft models for speculative decoding
-
lmsys/SGLang-EAGLE3-Llama-3.3-70B-Instruct-SpecForge
1B • Updated • 43 -
lmsys/SGLang-EAGLE3-Llama-3.1-8B-Instruct-SpecForge
0.4B • Updated • 2.48k -
lmsys/SGLang-EAGLE3-Qwen3-30B-A3B-Instruct-2507-SpecForge-Nex
0.2B • Updated • 896 • 3 -
lmsys/SGLang-EAGLE3-Llama-4-Scout-17B-16E-Instruct-SpecForge
0.8B • Updated • 139
SGLang-native diffusion transformer overrides converted with NVIDIA ModelOpt.
A collection of production-grade draft models for speculative decoding
-
lmsys/SGLang-EAGLE3-Llama-3.3-70B-Instruct-SpecForge
1B • Updated • 43 -
lmsys/SGLang-EAGLE3-Llama-3.1-8B-Instruct-SpecForge
0.4B • Updated • 2.48k -
lmsys/SGLang-EAGLE3-Qwen3-30B-A3B-Instruct-2507-SpecForge-Nex
0.2B • Updated • 896 • 3 -
lmsys/SGLang-EAGLE3-Llama-4-Scout-17B-16E-Instruct-SpecForge
0.8B • Updated • 139
models 55
lmsys/SGLang-EAGLE3-Step-3.5-Flash-SpecForge-RadixArk
0.4B • Updated
lmsys/qwen-image-edit-2511-modelopt-nvfp4-sglang
Image-to-Image • Updated
lmsys/qwen-image-edit-modelopt-nvfp4-sglang
Image-to-Image • Updated
lmsys/qwen-image-2512-modelopt-nvfp4-sglang
Text-to-Image • Updated
lmsys/qwen-image-modelopt-nvfp4-sglang
Text-to-Image • Updated
lmsys/Qwen3.5-397B-A17B-DFlash
Text Generation • 1B • Updated • 299 • 1
lmsys/hunyuanvideo-modelopt-fp8-sglang-transformer
Updated • 11 • 1
lmsys/qwen-image-edit-modelopt-fp8-sglang-transformer
Updated • 17 • 1
lmsys/qwen-image-modelopt-fp8-sglang-transformer
Updated • 8
lmsys/wan22-t2v-a14b-modelopt-nvfp4-sglang-transformer
Updated • 8 • 1