zijie-tian/nano-vllm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: add XAttention kernels guide

Browse Source 
- Document flat_group_gemm_fuse_reshape and softmax_fuse_block_sum kernels
- Explain anti-diagonal sum principle and stride sampling
- Add GPU-specific BLOCK_M/N constraints (RTX 3090 vs A100)
- Show Q/K can have different lengths (chunked prefill support)
- Update CLAUDE.md with doc reference

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Zijie Tian
2026-01-23 03:22:25 +08:00
parent 999858e82f
commit edc006463b
2 changed files with 199 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
CLAUDE.md
View File

@@ -15,6 +15,7 @@ Nano-vLLM is a lightweight vLLM implementation (~1,200 lines) for fast offline L
| [`docs/sparse_policy_implementation_guide.md`](docs/sparse_policy_implementation_guide.md) | How to implement custom SparsePolicy: required methods, hooks, ring buffer pipeline pattern |
| [`docs/sparse_attention_guide.md`](docs/sparse_attention_guide.md) | Block sparse attention methods (XAttention, FlexPrefill, MInference, AvgPool, Quest), computation flow, algorithms |
| [`docs/xattention_algorithm_guide.md`](docs/xattention_algorithm_guide.md) | XAttention 算法详解: stride reshape、Triton kernels、BSA 依赖、块选择算法 |
| [`docs/xattn_kernels_guide.md`](docs/xattn_kernels_guide.md) | XAttention Triton kernels: flat_group_gemm (反对角线求和)、softmax_fuse_block_sum (block 聚合) |
| [`docs/xattn_chunked_prefill.md`](docs/xattn_chunked_prefill.md) | XAttention chunked prefill: API、使用方式、一致性要求 |
| [`docs/block_sparse_attn_interface.md`](docs/block_sparse_attn_interface.md) | BSA (Block Sparse Attention) 接口文档: 函数签名、使用示例、约束条件 |
| [`docs/debugging_guide.md`](docs/debugging_guide.md) | PyTorch hooks for debugging, hook positions, tensor comparison, memory profiling |