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

📝 docs: add XAttention BSA Policy design documentation

Browse Source 
- Create docs/xattn_bsa_policy_design.md with:
  - Algorithm overview and data flow diagram
  - select_blocks implementation details
  - GQA-aware aggregation and majority voting
  - compute_chunked_prefill ring buffer pipeline
  - Parameter configuration and usage examples
  - Performance characteristics and limitations
- Update CLAUDE.md documentation index

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Zijie Tian
2026-01-23 08:36:56 +08:00
parent 4d1e40152d
commit ad361c2c3b
2 changed files with 295 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
CLAUDE.md
View File

@@ -17,6 +17,7 @@ Nano-vLLM is a lightweight vLLM implementation (~1,200 lines) for fast offline L
| [`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/xattn_bsa_policy_design.md`](docs/xattn_bsa_policy_design.md) | XAttention BSA Policy 设计: select_blocks 算法、majority voting、compute_chunked_prefill |
| [`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 |
| [`docs/optimization_guide.md`](docs/optimization_guide.md) | Performance optimizations: sgDMA (15x), Triton merge (4.3x), N-way pipeline (2x) |