PyTorch agent skills, page 6

基于 PyTorch 框架的昇腾 NPU 模型推理适配与部署基线技能。从 HF 链接或本地模型代码出发，按 cann-recipes-infer 仓库规范适配到 ModelRunner 推理框架，输出可运行的标准模型目录和性能基线数据。触发场景：新模型适配到昇腾 NPU 推理框架、已有模型的部署基线采集、模型迁移和初始跑通验证。

ascend-ai-coding/awesome-ascend-skills

从用户 PyTorch/Python 代码中提取算子实现，构建为算子任务格式的标准化 任务文件。支持两种模式：单 case（单一自包含 .py，getinputs 返回单组）和 多 case（.py + 同名 .json 配对，getinputgroups 返回多组）。

ascend-ai-coding/awesome-ascend-skills

根据设计文档生成 AscendC 算子完整代码实现并完成框架适配。TRIGGER when: 设计文档已完成，需要生成 ophost/opkernel 代码、注册到 PyTorch 框架、编译测试。关键词：代码生成、ophost、opkernel、tiling、kernel、框架适配、算子注册。

ascend-ai-coding/awesome-ascend-skills

面向 Ascend PyTorch Profiler / msprof DB（如 ascendpytorchprofiler.db、msprof.db）的 SQL 分析技能。将自然语言问题（算子耗时、通信、下发、调度、schema/table 查询）转为安全可执行 SQL，并按需从官方文档提取表结构详情。

ascend-ai-coding/awesome-ascend-skills

MindSpeed-MM multimodal model suite environment setup guide for Huawei Ascend NPU.

ascend-ai-coding/awesome-ascend-skills

Directly launch Docker containers on Ascend NPU servers for users.

ascend-ai-coding/awesome-ascend-skills

Implement machine learning solutions including model architectures, training pipelines, optimization strategies, and performance improvements.

aiskillstore/marketplace

PyTorch model training skill with custom training loops, gradient management, and GPU optimization.

majiayu000/claude-skill-registry

A skill your agent uses when training or debugging a neural net in PyTorch — the forward/loss/backward/step loop and its silent bugs, mixed precision (AMP), AdamW/LR schedules, DDP/FSDP/ZeRO…

ericrisco/rsc-harness

A skill your agent uses when predicting a column from rows of tabular features with classic models — scikit-learn pipelines, RandomForest, XGBoost/LightGBM, leak-free cross-validation, metrics for…

ericrisco/rsc-harness

Guide for building Graph Neural Networks with PyTorch Geometric (PyG).

BioTender-max/awesome-bio-agent-skills

Triton language skill for Python GPU kernel authoring. An agent skill from mohitmishra786/low-level-dev-skills.

mohitmishra786/low-level-dev-skills

Route text-labeling requests across LDA topic modeling, sklearn baselines, pretrained transformer models, and OpenAI-compatible LLM labeling.

Drchronx/ai-agent-research-starter-kit

AI 模型文件逆向与静态分析：ONNX/PyTorch/Safetensors/TFLite 格式解析、 网络结构还原、权重提取、文件级水印分析（权重 pattern/metadata/tensor hash/embedding 异常）。

dslsdzc/rev-skills

使用本地 BiRefNet GGUF 模型完成图片或视频抠图、人物抠图、主体分割和背景移除，并输出透明 PNG、MOV 或 WebM。适用于用户提到图片抠图、照片去背景、人像透明图、视频抠图、透明视频、BiRefNet、JPG/PNG/BMP/WebP 图片，或 MP4/MOV/WebM 视频的场景；无需 Python、PyTorch 或 CUDA。

aiskillstore/marketplace

A full ML pipeline where an agent team collaborates to perform data preparation, model design, training, evaluation, and deployment readiness.

revfactory/harness-100

Expert AI/ML Engineer with deep MLOps expertise. An agent skill from theneoai/awesome-skills.

theneoai/awesome-skills

W&B expert: experiment tracking, hyperparameter search, artifact management, sweep, team dashboards, performance visualization.

theneoai/awesome-skills

Computer vision implementation covering image classification, object detection (YOLO), image segmentation, OCR (Tesseract), face detection, image preprocessing, data augmentation, transfer learning…

FerroxLabs/wayland