E-MoE: Enhanced Mixture-of-Experts for Non-Factorized Diffusion Language Models. LM1B checkpoints: E-MoE and the MDLM / MDLM-MoE baselines.
Arseny Ivanov
ArsenyIvanov
AI & ML interests
NLP, LLM, RecSys, GNN
Recent Activity
published a model about 14 hours ago
ArsenyIvanov/mdlm-lm1b published a model about 14 hours ago
ArsenyIvanov/mdlm-moe-lm1b published a model about 14 hours ago
ArsenyIvanov/emoe-lm1bOrganizations
None yet
models 5
ArsenyIvanov/mdlm-moe-lm1b
Text Generation • 0.5B • Updated
ArsenyIvanov/emoe-lm1b
Text Generation • 0.5B • Updated
ArsenyIvanov/mdlm-lm1b
Text Generation • 0.1B • Updated
ArsenyIvanov/toolace-halu-qwen-lora
Text Generation • Updated • 11 • 1
ArsenyIvanov/toolace-halu-modernbert-large
Token Classification • 0.4B • Updated • 12 • 2
datasets 0
None public yet