experimental/cuda-ubi9/: flash-attn versions
Latest version on stage is: 2.5.7
Flash Attention: Fast and Memory-Efficient Exact Attention
Index | Version | Documentation |
---|---|---|
experimental/cuda-ubi9 | 2.5.7 |
Latest version on stage is: 2.5.7
Flash Attention: Fast and Memory-Efficient Exact Attention
Index | Version | Documentation |
---|---|---|
experimental/cuda-ubi9 | 2.5.7 |