Paged Flash-Attention Prefill (RDNA4/gfx1201)
Paged/chunked flash-attention prefill (+fp8) over a block-table KV cache.
Built with kernel-builder for AMD RDNA4 (gfx1201).
Load with the kernels library:
from kernels import get_kernel
kernel = get_kernel("pat883/attn-prefill-paged")
Requires a ROCm PyTorch build (torch 2.10 / ROCm 7.x) on an RDNA4 card. Built variants:
torch210-cxx11-rocm70andtorch210-cxx11-rocm71.
- Downloads last month
- 27
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support