Initial commit: obsidian to gitea

This commit is contained in:
2026-05-07 15:04:41 +08:00
commit a57afa86b4
323 changed files with 42569 additions and 0 deletions

View File

@@ -0,0 +1,89 @@
### prefill
对于 attention
对一个序列长度为 S 的 transformer attention多头、按 d = H * head_dim
1. **Q/K/V 投影**(一次性做 3 个线性变换):
FLOPsQKV_proj≈2×S×d×(3d)=6Sd2FLOPsQKV_proj≈2×S×d×(3d)=6Sd2
矩阵乘法的常用近似2·m·n·k
2. **Attention 矩阵乘Q·K^T**
FLOPsQK≈2×H×S×dhead×S=2S2dFLOPsQK≈2×H×S×dhead×S=2S2d
因为 Hdhead=dHdhead=d。
3. **Attention·V权重与 V 相乘)**
FLOPsAV≈2S2dFLOPsAV≈2S2d
4. **输出投影**(从 heads 拼回 d 再线性变换):
FLOPsout≈2Sd2FLOPsout≈2Sd2
总 FLOPs$8Sd^2 + 4S^2d$
$$
T_{\text{comp}} = \frac{\text{FLOPs}_{\text{per\_GPU}}}{\text{peak\_flops\_per\_GPU} \times \text{compute\_utils}}
$$
总 memory$\text{bytes}_\text{prefill} \approx N \cdot \alpha \cdot BLd \cdot \text{elem\_bytes}$$\alpha \sim 6$
$$
T_{\text{mem}} = \frac{\text{bytes}_{\text{per\_GPU}}}{\text{bandwidth\_per\_GPU} \times \text{mem\_utils}}
$$
### decode
总 FLOPs$8d^2 + 4dL$
总 memory$\text{bytes}_\text{decode} \approx N \cdot \beta \cdot BLd \cdot \text{elem\_bytes}$$\beta \sim 4$
$\text{output} = \text{SiLU}(xW_1)W_2$
TP 下,每 token $T$ 激活的 expert $E$ 的通信为:
1. 输入 $x$ AllGather 到所有 TP 节点通信量hidden_size * (TP - 1)
2. 每个 TP 节点独立计算 $xW_1'$
3. AllGather 后每个节点得到完整的 $xW_1$通信量moe_intermediate_size / TP * (TP - 1)
4. 每个节点计算 SiLU 和 $IW2'$AllReduce 每个节点得到完整的 output通信量hidden_size * (TP - 1)
EP 下dispatch+combine
2 * hidden_size * (EP - 1) / EP (假设负载均衡)
---
With batch_size=2000, seq_len=2048, EP=8
Qwen-235B:
attention comp time 0.06944874306412531
moe combine comm time 0.00028672
moe comp time 0.0004069259060131379
moe comm time with TP 0.00045056
Qwen-30B:
attention comp time 0.01736218682573421
moe combine comm time 0.00014336
moe comp time 0.00010173147650328448
moe comm time with TP 0.00022528
EP=64:
Qwen-235B:
attention comp time 0.06944874306412531
moe combine comm time 0.00032256
moe comp time 0.0004069259060131379
moe comm time with TP 0.00045056