List your company

Language Modeling on LM1B GPT-2 small model size equivalent (test)

20.53PerplexityVADD
18.830.477542.15553.8325May 23, 2025

Evaluation results

MethodLinks
VADDModel Category=Diffusion, #Iterations=1M2025.0520.53
TransformerModel Category=Auto-Regressive, #Iterations=5M, Source=Reported in Sahoo et al. (2024)2025.0520.86
OmniNetTModel Category=Auto-Regressive, #Iterations=-2025.0521.5
TransformerModel Category=Auto-Regressive, #Iterations=0.5M, Source=Reported in Sahoo et al. (2024)2025.0522.32
MDLMModel Category=Diffusion, #Iterations=10M, Source=Reported in Sahoo et al. (2024)2025.0523
Transformer-XLModel Category=Auto-Regressive, #Iterations=-2025.0523.5
MDLMModel Category=Diffusion, #Iterations=1M, Source=Reported in Sahoo et al. (2024)2025.0527.04
MDLMModel Category=Diffusion, #Iterations=1M, Source=Reproduced by current authors2025.0527.7
SEDD-AbsorbModel Category=Diffusion, #Iterations=1M2025.0532.79
DiffusionBertModel Category=Diffusion, #Iterations=1.9M2025.0563.78