NX-AI/KronLSTM-1.5B-100B
Text Generation • 1B • Updated • 15
The language models of the paper "Kronecker Delta Rule for Long Context Linear Attention" pretrained on 100B tokens.
Note 24 KronLSTM layers, 8K context
Note 24 KronLSTM layers, continued at 32K context
Note 20 KronLSTM + 4 sliding-window attention layers, 8K context
Note 20 KronLSTM + 4 sliding-window attention layers, continued at 32K context