Vendor-agnostic orchestration for training, inference and agentic workloads across NVIDIA, AMD, TPU, and Tenstorrent on clouds, Kubernetes, and bare metal.

BertViz: Visualize Attention in Transformer Models

Use when reviewing a PR, API, IPC channel, endpoint, parameter, type, config, or architectural extension point that adds or expands shared surface area, especially when consumers are absent, exports are unused or speculative, existing consumers are hack-heavy, forward compatibility is claimed, or multiple similar APIs may express one demand.

Automated Cherry Studio review for local branches, PRs, commits, files, architecture docs, and repository skills. Use for code or documentation reviews that need project-specific naming, main/renderer/shared placement and dependency rules, IpcApi and DataApi boundaries, lifecycle/service ownership, renderer hooks, React/UI conventions, and tests. Supports single-agent review with interactive fix selection or multi-agent reviewer-verifier review with risk-based auto-fix. To diagnose gaps in the skill after a review session, run `/gh-pr-review diag`.

[ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.

[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.

Autonomously improve a real artifact (code, training recipe, agent harness, data pipeline, prompt) against an objective and an evaluator, using Hypothesis Tree Refinement (HTR) from the Arbor paper. Use this whenever someone wants to iteratively optimize something over many experiments without overfitting — e.g. "get my model's eval score up", "improve this agent/harness", "tune this pipeline", "beat the baseline on this benchmark", "run a search over approaches and keep the best", "do an MLE-bench / Kaggle-style optimization", or any long-horizon "make this artifact better and don't just memorize the dev set" task. Trigger it even when the user doesn't say "Arbor" or "hypothesis tree" but describes repeated experiment-and-evaluate loops, branching exploration of competing ideas, or worries about a dev/test gap. Runs Claude itself as the coordinator with subagent executors in isolated git worktrees; for the standalone `arbor` CLI tool see references/arbor-upstream.md.

When the user needs to generate, iterate, or scale ad creative for paid advertising. Use when they say 'write ad copy,' 'generate headlines,' 'create ad variations,' 'bulk creative,' 'iterate on ads,' 'ad copy validation,' 'RSA headlines,' 'Meta ad copy,' 'LinkedIn ad,' or 'creative testing.' This is pure creative production — distinct from paid-ads (campaign strategy). Use ad-creative when you need the copy, not the campaign plan.

Go实现的Trojan代理，支持多路复用/路由功能/CDN中转/Shadowsocks混淆插件，多平台，无依赖。A Trojan proxy written in Go. An unidentifiable mechanism that helps you bypass GFW. https://p4gefau1t.github.io/trojan-go/

DiffSinger: Singing Voice Synthesis via Shallow Diffusion Mechanism (SVS & TTS); AAAI 2022; Official code

It is open source ebook about TensorFlow kernel and implementation mechanism.

中文长文本分类、短句子分类、多标签分类、两句子相似度（Chinese Text Classification of Keras NLP, multi-label classify, or sentence classify, long or short），字词句向量嵌入层（embeddings）和网络层（graph）构建基类，FastText，TextCNN，CharCNN，TextRNN, RCNN, DCNN, DPCNN, VDCNN, CRNN, Bert, Xlnet, Albert, Attention, DeepMoji, HAN, 胶囊网络-CapsuleNet, Transformer-encode, Seq2seq, SWEM, LEAM, TextGCN