List your company

Language Modeling on WikiText (Accuracy and LIS)

17.2826AccuracyRaw
17.2066817.2263917.246117.26581Feb 23, 2026

Evaluation results

MethodLinks
RawModel architecture=Switch Transformer (8 experts), Strategy=Unmodified model2026.0217.28262.09
R&QModel architecture=Switch Transformer (8 experts), Strategy=Replicate-and-Quantize2026.0217.20961.6235