Explorar skills

Search academic literature and generate research hypotheses

Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.

Draft a LaTeX research paper from all previous stage outputs

Produce an experiment blueprint from a research hypothesis

Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization

Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or batch jobs across multiple clouds, leverage spot instances with auto-recovery, or optimize GPU costs across providers.

Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.

Técnicas de Engenharia Reversa para CTF, aplicadas na análise de binários desconhecidos, quebra de verificações de jogos, desofuscação de código e interpretação de VMs customizadas. Abrange análise estática/dinâmica, bypass de anti-depuração e engenharia reversa multiplataforma para WASM, .NET, APK, bytecode Python, Go e Rust.

Uma CLI com foco em agentes de IA para BOSS Zhipin, que permite pesquisa de vagas, filtragem de benefícios, fluxo de trabalho para recrutadores, ferramentas MCP e otimização de currículos por IA.

Push architecture docs to oh-my-mermaid cloud. Handles login, link, and push workflow with error guidance. Use when the user says "omm push", "push to cloud", "deploy architecture", or "share architecture".

Scan codebase architecture and generate/update .omm/ documentation. Use when the user says "omm scan", "scan architecture", "update architecture", "refresh diagrams".