Científicos de datos

Ctr Training

使用 FuxiCTR 框架训练各种 CTR 预测模型（DeepFM、DCN、DIN、DCNv2、xDeepFM 等）

raoxuan98-hash0

Dcnv2 Training

使用 FuxiCTR 框架训练和评估 DCNv2 CTR 预测模型

raoxuan98-hash0

Dataset Download

下载和准备 CTR 模型评估所需的数据集（MovielensLatest、Frappe、KuaiVideo 等）

raoxuan98-hash0

Diseño

Matplotlib for scientific figures

Build publication-quality Matplotlib figures for scientific Python. Use when plotting experiment or analysis results, multi-panel figures, journal-sized layouts, tick formatting, legends (fancy/plain, placement, compact handles), point annotations, inset zooms, twin axes, colored spines/ticks, colorblind-safe palettes, PNG/PDF export, or SciencePlots styling. Triggers on matplotlib, pyplot, subplots, savefig, legend, annotate, inset, twinx, scientific figures, panel labels.

WSU-Carbon-Lab0

Computación Científica

Numpy Scientific

NumPy for scientific Python: dtypes and casting, creation, reshaping, broadcasting, indexing vs views, ufuncs and reductions, linear algebra and einsum, random Generator, I/O, structured arrays, performance and pandas boundaries. Use when writing or reviewing ndarray code. Triggers on numpy, ndarray, broadcasting, dtype, ufunc, einsum, vectorization.

WSU-Carbon-Lab0

Bioinformática

Rag Patterns

Implementation patterns for Retrieval-Augmented Generation covering basic chains, corrective RAG, hybrid search, knowledge graphs with community detection, agentic RAG, multimodal retrieval, and reasoning-based vectorless RAG

yunseo-kim0

Finanzas e Inversión

Skill: portfolio-optimization

Analyzes multiple stocks in a portfolio context, computing correlation matrices, sector rotation insights, and multi-factor scores to rank and recommend the optimal securities. Produces a ranked buy/hold/sell action list with diversification metrics and relative value comparison, with macro-regime overlays (geopolitical, monetary-policy, and market-stress context) applied to scoring and allocations, plus rule-based portfolio narrative generation and per-ticker data-source diagnostics.

dwen1680

Análisis de Datos

Dataframes

Pandas and Polars for tabular data: when to use which, indexing and dtypes, joins and reshaping, lazy Polars, I/O and Parquet, nulls, Arrow interop, and performance. Use for DataFrame/Series/LazyFrame code. Triggers on pandas, polars, DataFrame, LazyFrame, parquet, groupby, merge, join.

WSU-Carbon-Lab0

Análisis de Datos

Rolling Baseline Calculation

Calculate weekly averages from 60-day Salesforce data (total ÷ 8.57 weeks) for performance comparison

RevPalSFDC0

Qianjin Xuntui

前进平台 AI 开发平台操作技能，支持模型训练、模型管理、模型部署和 AI 资产管理。通过 qianjin-xuntui MCP Server 提供 87 个工具。当用户提到前进平台、模型训练、Notebook、模型部署、数据集管理、镜像管理、模型评测、魔搭下载、批量推理时触发此技能。

WTaoGitHub0

Depuración

Research Filter

Two-phase relevance filter for feed entries. Phase 1 scans titles broadly using semantic judgment against topics.yml. Phase 2 reads abstracts and applies positive/negative signals from topics.yml to tier entries (High/Medium/Low). Use when curating downloaded entries before passing to research-digest.

ThingsAI-io0

Ingeniería de Datos

Duckdb En

DuckDB CLI specialist for SQL analysis, data processing and file conversion. Use for SQL queries, CSV/Parquet/JSON analysis, database queries, or data conversion. Triggers on "duckdb", "sql", "query", "data analysis", "parquet", "convert data".

ThinkfleetAI0

LLM & AI

Llm Context

USE FOR RAG/LLM grounding. Returns pre-extracted web content (text, tables, code) optimized for LLMs. GET + POST. Adjust max_tokens/count based on complexity. Supports Goggles, local/POI. For AI answers use answers. Recommended for anyone building AI/agentic applications.

ramseymcgrath0

Plugins de IDE

Llava

Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP vision encoder with Vicuna/LLaMA language models. Supports multi-turn image chat, visual question answering, and instruction following. Use for vision-language chatbots or image understanding tasks. Best for conversational image analysis.

LLM & AI

Rag Engineer

Build RAG systems: document chunking, embedding generation, vector storage, retrieval strategies, and answer generation.

ThinkfleetAI0

Deepdrivewe Example

Use this skill when building a new example or custom workflow with DeepDriveWE (weighted ensemble MD on the Academy multi-agent framework). Covers project layout, agent subclassing, Pydantic config models, binner/recycler/resampler selection, Parsl compute configs, and verification steps.

ramanathanlab0

Huggingface Tokenizers

Fast tokenizers optimized for research and production. Rust-based implementation tokenizes 1GB in <20 seconds. Supports BPE, WordPiece, and Unigram algorithms. Train custom vocabularies, track alignments, handle padding/truncation. Integrates seamlessly with transformers. Use when you need high-performance tokenization or custom tokenizer training.

Optimizing Attention Flash

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

Pytorch Lightning

High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.

Slime Rl Training

Provides guidance for LLM post-training with RL using slime, a Megatron+SGLang framework. Use when training GLM models, implementing custom data generation workflows, or needing tight Megatron-LM integration for RL scaling.

Huggingface Accelerate

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.

Depuración

Faiss

Facebook's library for efficient similarity search and clustering of dense vectors. Supports billions of vectors, GPU acceleration, and various index types (Flat, IVF, HNSW). Use for fast k-NN search, large-scale vector retrieval, or when you need pure similarity search without metadata. Best for high-performance applications.

Lambda Labs Gpu Cloud

Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances with simple SSH access, persistent filesystems, or high-performance multi-node clusters for large-scale training.

Química Computacional

Qdrant Vector Search

High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG systems requiring fast nearest neighbor search, hybrid search with filtering, or scalable vector storage with Rust-powered performance.

Bioinformática

Sparse Autoencoder Training

Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable features. Use when discovering interpretable features, analyzing superposition, or studying monosemantic representations in language models.

Tensorrt Llm

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100), when you need 10-100x faster inference than PyTorch, or for serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling.

Cloud

Modal Serverless Gpu

Serverless GPU cloud platform for running ML workloads. Use when you need on-demand GPU access without infrastructure management, deploying ML models as APIs, or running batch jobs with automatic scaling.

Depuración

Evaluating Llms Harness

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.