Ascend/msagent / skills/modeling/msmodeling-optix-param-recommend
msmodeling-optix-param-recommend
Verified当首次使用 msmodeling optix 的用户需要根据硬件、模型、负载和优化目标推荐 MindIE/vLLM 寻优参数、搜索范围、benchmark 侧字段或 config.toml 片段时使用。
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 16.2 KB |
| references/historical-configs.md | markdown | 11.3 KB |
| references/input-schema.md | markdown | 2.2 KB |
| references/parameter-rules.md | markdown | 13.3 KB |
| scripts/auto_config.py | script | 43.8 KB |
| scripts/recommend_params.py | script | 57.1 KB |
| scripts/test_recommend_params.py | script | 10.2 KB |
references/input-schema.md
View on GitHub# 参数推荐输入 Schema
收集首次寻优信息时使用本参考文档。
## 必需字段
```json
{
"engine": "mindie | vllm",
"hardware": {
"single_card_mem_gb": 64,
"world_size": 8,
"num_per_nodes": 8,
"num_nodes": 1
},
"model": {
"config_path": "/path/to/model/config.json"
},
"workload": {
"input_len_avg": 1024,
"input_len_max": 4096,
"output_len_avg": 256,
"output_len_max": 512
},
"target": "throughput | ttft | tpot | balanced"
}
```
`benchmark_policy` 是可选字段。缺省时 vLLM 使用 `vllm_benchmark`,MindIE 使用 `ais_bench`。
`discovery` 是可选字段。缺省时不做 `--help` 参数发现。
## 没有 config.json 时的模型字段
如果用户无法提供模型 `config.json`,需要收集:
- `hidden_size`
- `intermediate_size`
- `num_hidden_layers`
- `num_attention_heads`
- `num_key_value_heads`
- `torch_dtype`
- `max_position_embeddings`
- `vocab_size`,如用户知道
可选字段:
- `model_weight_gb`
- `num_parameters_billion`
- `model_type`
- `is_moe`
- `num_experts`
- `num_local_experts`
- `n_routed_experts`
- `is_multimodal`
- `is_quantized`
## 可选 discovery 字段
```json
{
"discovery": {
"enabled": true,
"vllm_help_text_path": "/tmp/vllm-serve-help.txt",
"vllm_help_text": "也可以直接放入 vllm serve --help 的输出"
}
}
```
使用建议:
- `enabled` 为 `false` 或缺省时,只输出基础推荐参数。
- `vllm_help_text_path` 适合 agent 先执行 `vllm serve --help > /tmp/vllm-serve-help.txt` 后再调用脚本。
- `vllm_help_text` 适合测试或用户直接粘贴 help 输出。
- 当前 discovery 只分析 vLLM help;MindIE 仍优先读取 config。
- discovery 会结合 `model.is_multimodal`、`model.is_moe` 和业务输入长度决定是否追加参数。
## 问题顺序
每轮只问一个主题:
1. 硬件信息。
2. 模型路径或模型结构。
3. 业务 token 长度。
4. 优化目标。
5. 可选:启动命令或 MindIE config 路径。
6. 可选:如果用户同意执行本地只读命令,收集 `vllm serve --help` 输出用于 discovery。
如果 `recommend_params.py` 返回 `need_more_info`,优先询问返回的 `next_question`。
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |