📝 docs: add CUDA Graph optimization plan for offload mode decode

- Update task_plan.md with 6-phase segmented graph implementation plan
- Add findings.md documenting 7 key discoveries about current implementation
- Add progress.md for tracking implementation progress
- Add test_chunk_attention_graph_reuse.py validating 2-graph reuse strategy

Key architecture decision: Split transformer layer into 3 segments:
- PRE-ATTENTION GRAPH: norm → qkv_proj → rotary (1 graph, reused)
- CHUNKED ATTENTION: H2D (eager) + flash_attn (2 graphs) + merge (eager)
- POST-ATTENTION GRAPH: o_proj → norm → FFN (1 graph, reused)

Total: 4 graphs serving all layers via copy_() tensor updates.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Zijie Tian
2026-01-22 02:12:24 +08:00
parent d808970f2f
commit a5307fb124
4 changed files with 651 additions and 64 deletions

55
progress.md Normal file
View File

@@ -0,0 +1,55 @@
# Progress: CUDA Graph for Offload Mode
## Session: 2026-01-22
### 调研阶段 ✅ 完成
**完成的调研**:
1. ✅ 分析 `model_runner.py` 中的 CUDA Graph 实现
- `capture_cudagraph()`: 为不同 batch size 捕获完整 model forward
- `run_model()`: 通过 `is_chunked_prefill` 决定 eager/graph
2. ✅ 分析 offload decode 流程
- `run_chunked_offload_decode()` 设置 `is_chunked_prefill=True`
- 导致永远使用 eager mode
3. ✅ 分析 ring buffer pipeline
- `_decode_ring_buffer_pipeline()` 包含 H2D 传输 + attention 计算
- H2D 不能 graphattention 可以 graph
4. ✅ 验证 graph 复用策略
- 创建 `test_chunk_attention_graph_reuse.py`
- 确认 2 个 graph 可复用于所有层
### 计划编写 ✅ 完成
- ✅ 创建 `task_plan.md`
- ✅ 创建 `findings.md`
- ✅ 创建 `progress.md`
### 下一步: 实现
**Phase 1**: 添加 graph 捕获到 OffloadEngine
- [ ]`offload_engine.py` 添加 `capture_attention_graphs()`
- [ ] 添加 `attention_graph_causal``attention_graph_non_causal` 属性
**Phase 2**: 修改 ring buffer pipeline
- [ ]`_decode_ring_buffer_pipeline()` 使用 graph replay
- [ ] 保持 H2D 和 merge 为 eager
**Phase 3**: 测试
- [ ] 运行 needle test 验证正确性
- [ ] 对比性能
---
## 文件清单
| 文件 | 状态 | 说明 |
|------|------|------|
| `tests/test_chunk_attention_graph.py` | ✅ 已提交 | 预分配 chunk pair graphs 测试 |
| `tests/test_chunk_attention_graph_reuse.py` | 待提交 | Graph 复用验证 |
| `task_plan.md` | ✅ 创建 | 实现计划 |
| `findings.md` | ✅ 创建 | 调研发现 |
| `progress.md` | ✅ 创建 | 进度日志 |