kali20gakki/msAgent / skills/quantizer/quantization-expert-experience-tuning-rules
quantization-expert-experience-tuning-rules
Verified量化专家经验调优库:按「L1 通用专家调优意见 → L2 结构化专家调优意见 → L3 模型专属专家调优意见」三级递进, 输出量化精度调优的完整手段(离群值抑制 / 量化方法·粒度·对称性选型 / 校准集调整 / 敏感层回退 / 模型专属策略), 每条结论附专家原因分析与专家意见可信度等级。 本 Skill 是专家经验库,负责说明可选离群值抑制算法并询问用户选择哪些算法,以及回答「怎么调、哪些层需要回退及为什么」;不运行离群值抑制 processor,不修改 YAML、不执行量化、不做 EP 检查 / 服务化 / 任务评测。
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 10.8 KB |
| cross-model-pitfalls.md | markdown | 8.3 KB |
| models/README.md | markdown | 1.3 KB |
| models/deepseek/deepseek-v3.md | markdown | 2.9 KB |
| models/deepseek/deepseek-v4.md | markdown | 2.5 KB |
| models/glm/glm-4-moe.md | markdown | 1.3 KB |
| models/glm/glm-5.md | markdown | 1.4 KB |
| models/minimax/minimax-m.md | markdown | 1.6 KB |
| scripts/README.md | markdown | 3.1 KB |
| scripts/apply_rollback.py | script | 10.1 KB |
| scripts/inspect_model_structure.py | script | 7.1 KB |
| scripts/yaml_utils.py | script | 4.2 KB |
| structure-family-pitfalls.md | markdown | 14.2 KB |
models/README.md
View on GitHub# L3 模型专属专家调优意见经验(models/)
> **定位与阅读规则**
> - 本目录是**各 `<vendor>/<model>.md` 个案**,只放**该模型/该模型类独有、不可迁移**的调优坑与优化手段。
> - 通用手段(量化格式/离群值抑制/方法/粒度/校准集)见 **L1**;结构模型类共性(MoE/MLA/混合 attention/VLM/DiT)见 **L2**。L3 只在 L1+L2 基础上补充「只有这个模型才这样」的内容。
> - 每条给出专家原因与专家意见可信度;没有独立经验的模型不建假条目。
## 已收录模型索引
| vendor | 模型/模型类 | 文件 | 独有特征速览 |
|---|---|---|---|
| deepseek | DeepSeek-V3 / R1 / V3.1 / V3.2 | `deepseek/deepseek-v3.md` | MLA 低秩投影排除、fa3_quant per-head、`model.layers.61` 尾部敏感层 |
| deepseek | DeepSeek-V4-Pro / Flash | `deepseek/deepseek-v4.md` | `ffn` 命名(非 mlp)、`indexer`/`compressor` 投影排除、QuaRot block_size=32 |
| glm | GLM-4-MoE | `glm/glm-4-moe.md` | 末层 `model.layers.92.*` 排除 |
| glm | GLM-5.1 | `glm/glm-5.md` | W4A4 mxfp4、`part_file_size: 8` |
| minimax | MiniMax-M2.7 / M3 | `minimax/minimax-m.md` | `block_sparse_moe.experts`、iter_smooth、mxfp8 |
未列出的模型:先走 L1 + L2(按触发信号判定结构模型类),有独立实践 YAML 时再补 L3 条目。Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |