开发敏感数据github未标认证来源可访问许可证需确认审计通过

wan-ascend-adaptation（万提升适应）

Agent Skill

wan-ascend-adaptation 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息，适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态，以及是否会触发联网、命令执行或文件读写。

总安装

247

周安装

GitHub Stars

下载量

CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息，不托管安装包，也不代用户安装。

GitHub

来源数

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手，先确认命令和权限，再让它执行。

请帮我安装这个 Agent Skill：wan-ascend-adaptation（万提升适应）
来源仓库：https://github.com/ascend-ai-coding/awesome-ascend-skills
仓库路径：skills/wan-ascend-adaptation
安装命令：
npx skills add https://github.com/ascend-ai-coding/awesome-ascend-skills --skill wan-ascend-adaptation
安装前请先检查当前环境是否支持对应 CLI，并向我确认将要执行的命令、安装目录、联网范围和文件读写权限；确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill；本站只展示命令，不托管安装包，也不自动执行。

skills.shnpx skills

npx skills add https://github.com/ascend-ai-coding/awesome-ascend-skills --skill wan-ascend-adaptation

简介

用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息。

适合在 Codex、Claude、Cursor、Gemini CLI 中围绕仓库状态、代码变更或协作事项进行整理。
通过 npx skills add 命令从指定仓库安装并使用。
安装前需确认权限范围、维护状态，注意是否触发联网、命令执行或文件读写。
wan-ascend-adaptation 属于开发类 Skill，可作为该场景下的辅助能力补充。

SKILL.md

Wan-Series Model Ascend NPU Adaptation Skill

Purpose

Provide a systematic, step-by-step guide for adapting Wan-series (and similar DiT-based) video generation models from NVIDIA CUDA/GPU to Huawei Ascend NPU. The skill encodes 9 major adaptation domains covering every layer of the inference stack, from device initialization to distributed parallelism.

When to Use

Porting a Wan-series (Wan2.1 / Wan2.2) model from CUDA to Ascend NPU
Adapting any DiT-based video diffusion model for Ascend hardware
Optimizing inference performance on Ascend NPU (attention, quantization, VAE parallel)
Setting up multi-card distributed inference on Atlas 800 series hardware
Integrating MindIE SD acceleration library into a PyTorch video generation pipeline

Adaptation Domains Overview

The adaptation work is organized into 9 domains. Each domain has a dedicated reference file under references/ with detailed instructions, code patterns, and pitfalls.

#	Domain	Reference File	Priority
1	Device Layer Adaptation	`references/01-device-layer.md`	P0 — Must
2	Operator Replacement	`references/02-operator-replacement.md`	P0 — Must
3	Precision Strategy	`references/03-precision-strategy.md`	P0 — Must
4	Attention Mechanism	`references/04-attention-mechanism.md`	P1 — Critical
5	Distributed Parallelism	`references/05-distributed-parallelism.md`	P1 — Critical
6	VAE Patch Parallel	`references/06-vae-patch-parallel.md`	P2 — Important
7	Model Quantization	`references/07-model-quantization.md`	P2 — Important
8	Sparse Attention (RainFusion)	`references/08-sparse-attention.md`	P2 — Important
9	Inference Pipeline Integration	`references/09-pipeline-integration.md`	P1 — Critical

Workflow

To adapt a Wan-series model to Ascend, follow these steps in order:

Step 1: Device Layer Adaptation (Domain 1)

Read references/01-device-layer.md for complete guidance.

Key actions:

Import torch_npu and transfer_to_npu at the entry point
Configure NPU compile mode and internal format settings
Replace dist.init_process_group(backend="nccl") with backend="hccl"
Replace all torch.amp.autocast('cuda',...) with autocast('npu',...)
Replace device type checks from 'cuda' to 'npu'

Step 2: Operator Replacement (Domain 2)

Read references/02-operator-replacement.md for complete guidance.

Key actions:

Replace RMSNorm with torch_npu.npu_rms_norm()
Replace LayerNorm forward to remove .float() type casting
Replace RoPE with mindiesd.rotary_position_embedding() fused operator
Optionally enable mindiesd.fast_layernorm via FAST_LAYERNORM env var
Replace Flash Attention with mindiesd.attention_forward() multi-backend dispatch

Step 3: Precision Strategy (Domain 3)

Read references/03-precision-strategy.md for complete guidance.

Key actions:

Lower sinusoidal embedding from float64 to float32
Lower RoPE frequency from complex128 to complex64
Change autocast dtype from float32 to bfloat16
Remove .float() type conversions in normalization layers
Use PRECISION env var to control random number device for cross-platform reproducibility

Step 4: Attention Mechanism Adaptation (Domain 4)

Read references/04-attention-mechanism.md for complete guidance.

Key actions:

Implement multi-backend attention dispatch via ALGO env var (0/1/3)
Create xFuserLongContextAttention combining Ulysses + Ring Attention
Integrate Attention Cache via mindiesd.CacheAgent
Add sub-head splitting support via USE_SUB_HEAD env var

Step 5: Distributed Parallelism Refactoring (Domain 5)

Read references/05-distributed-parallelism.md for complete guidance.

Key actions:

Implement ParallelConfig with 4D parallelism: TP × SP × CFG
Create RankGenerator for orthogonal process group assignment
Create GroupCoordinator with dual-channel communication (HCCL + Gloo)
Implement TensorParallelApplicator for automatic model sharding
Implement CFG parallel to halve sampling loop forward passes

Step 6: VAE Patch Parallel (Domain 6)

Read references/06-vae-patch-parallel.md for complete guidance.

Key actions:

Implement spatial H×W slicing across NPUs
Monkey-patch F.conv3d, F.conv2d, F.interpolate, F.pad for boundary exchange
Use P2P communication for neighbor boundary data exchange
Adjust VAE CausalConv3d padding strategy for compatibility

Step 7: Model Quantization (Domain 7)

Read references/07-model-quantization.md for complete guidance.

Key actions:

Use msmodelslim for W8A8 dynamic quantization
Integrate mindiesd.quantize() for runtime quantization loading
Handle FSDP + float8 compatibility via patch_cast_buffers_for_float8()

Step 8: Sparse Attention — RainFusion (Domain 8)

Read references/08-sparse-attention.md for complete guidance.

Key actions:

Implement RainFusion v1 (window-based Local/Global adaptive)
Implement RainFusion v2 (blockwise Top-K sparse)
Configure skip_timesteps for quality-speed tradeoff

Step 9: Pipeline Integration (Domain 9)

Read references/09-pipeline-integration.md for complete guidance.

Key actions:

Add warm-up generation steps for NPU operator compilation
Configure T5_LOAD_CPU for flexible T5 loading strategy
Add RoPE frequency cache (freqs_list) lifecycle management
Implement multi-resolution VAE decode condition (rank < 8)
Add performance timing with stream.synchronize()

Key Environment Variables

Variable	Default	Description
`ALGO`	`0`	Attention algorithm: 0=fused_attn_score, 1=ascend_laser_attention, 3=npu_fused_infer
`FAST_LAYERNORM`	`0`	Enable mindiesd fast LayerNorm
`USE_SUB_HEAD`	`0`	Sub-head group size for attention splitting
`T5_LOAD_CPU`	`0`	Load T5 model on CPU to save NPU memory
`PRECISION`	`0`	Generate random numbers on CPU for cross-platform reproducibility
`OVERLAP`	`0`	Enable FA-AllToAll communication overlap
`PYTORCH_NPU_ALLOC_CONF`	-	NPU memory allocation strategy
`TASK_QUEUE_ENABLE`	-	NPU task queue optimization
`CPU_AFFINITY_CONF`	-	CPU affinity configuration

Key Dependencies

Library	Purpose
`torch_npu`	PyTorch Ascend NPU backend
`mindiesd`	MindIE Stable Diffusion acceleration (FA, RoPE, LayerNorm, quantize)
`msmodelslim`	Huawei model compression toolkit (W8A8 quantization)
`yunchang`	Sequence parallel framework (Ulysses + Ring Attention)
`torch_atb`	Ascend Transformer Boost operators
`atb_ops`	ATB fused matmul-allreduce operators

Notes

This skill is derived from comparing Wan2.2-Original (CUDA) and Wan2.2-Ascend (NPU) codebases
The Ascend version removes S2V (Speech-to-Video) and Animate tasks, focusing on T2V, I2V, and TI2V
Hardware target: Atlas 800I A2 / Atlas 800T A2 with 8×64G NPU
All adaptation patterns are applicable to similar DiT-based video diffusion architectures

适合场景

用户想查找某类 Agent Skill 时

需要根据任务场景推荐可安装能力包时

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明，方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用；具体调用方式请以来源页面和 README 为准。

平台分布

Codex

36.23%

按下载量换算28

Claude

32.02%

按下载量换算25

Cursor

16.5%

按下载量换算13

Gemini CLI

9.91%

按下载量换算8

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置，应进入高风险复核队列，默认不自动发布。

安装前确认

本站仅展示第三方公开信息，不托管安装包，不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源，正式发布前建议补源仓库或其他目录站核验。

来源信息

GitHub 仓库

github.com

/ascend-ai-coding/awesome-ascend-skills

SKILL.md 来源

github.com

/ascend-ai-coding/awesome-ascend-skills/tree/main/skills/wan-ascend-adaptation

skills-sh

skills.sh

/ascend-ai-coding/awesome-ascend-skills/wan-ascend-adaptation

继续浏览同类 Skills

更多开发 Skills（21,746）返回 Skills 目录

上一个 Skill

wan-ai-leading-ai-video-generation-model（万爱领先的 AI 视频生成模型）

用于辅助视频生成、动画合成、脚本化剪辑或 Remotion 等视频项目开发。它适合让 Agent 组织镜头、生成素材说明、维护合成代码或排查渲染问题。使用时需要确认分辨率、时长、素材路径和导出格式；涉及外部素材、人物肖像或商业发布时，应先核对版权授权和内容审核要求。

下一个 Skill

wandb-experiment-logger（wandb 实验记录器）

wandb-experiment-logger 用于记录任务执行中的错误、用户纠正、经验和能力缺口，适合在 Codex、Claude、Cursor、Gemini CLI 中希望让 Agent 持续沉淀问题、修正和最佳实践时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态，以及是否会触发联网、命令执行或文件读写。