sustcsonglin/flash-linear-attention Issues
RWKV6 backward issue
Closed 15更新后的rwkv6,loss会nan
Closed 16illegal memory access error
Closed 2'RebasedFeatureMap' is missing?
Closed 4Mistakes in the GLA paper
Closed 4
Efficient implementations of state-of-the-art linear attention models in Pytorch and Triton