Hugging Face
Models
Datasets
Spaces
Posts
Docs
Enterprise
Pricing
Log In
Sign Up
moonshotai
/
Moonlight-16B-A3B
like
74
Follow
Moonshot AI
200
Text Generation
Transformers
Safetensors
deepseek_v3
conversational
custom_code
arxiv:
2502.16982
License:
mit
Model card
Files
Files and versions
Community
5
Train
Use this model
main
Moonlight-16B-A3B
/
figures
4 contributors
History:
1 commit
liushaowei
add figures
b13722c
20 days ago
banner.png
Safe
48.8 kB
add figures
20 days ago
banner_short.png
Safe
26.9 kB
add figures
20 days ago
chinlaw_8k_flops_ratio.png
Safe
145 kB
add figures
20 days ago
fig_MMLU_performance.png
Safe
225 kB
add figures
20 days ago
fig_weight_decay.png
Safe
416 kB
add figures
20 days ago
logo.png
Safe
13.1 kB
add figures
20 days ago
megatron.png
Safe
1.99 kB
add figures
20 days ago
scaling.png
Safe
224 kB
add figures
20 days ago