Results for “models”

Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on consumer GPUs, when you need 4× memory reduction with <2% perplexity degradation, or for faster inference (3-4× speedup) vs FP16. Integrates with transformers and PEFT for QLoRA fine-tuning.

Use this skill when the user asks how to pay CDL drivers — cents per mile (CPM), percentage of revenue, hourly, salary, sleeper-team rate, detention pay, driver-pay laws under FLSA (Fair Labor Standards Act), and how each model affects retention + recruitment. Cite FLSA + state wage laws.

Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on consumer GPUs, when you need 4× memory reduction with <2% perplexity degradation, or for faster inference (3-4× speedup) vs FP16. Integrates with transformers and PEFT for QLoRA fine-tuning.

Generate images, videos, and audio using fal.ai models via MCP tools, with support for text-to-image, text/image-to-video, text-to-speech, and video-to-audio.

High-performance toolkit for genomic interval analysis in Rust with Python bindings. Use when working with genomic regions, BED files, coverage tracks, overlap detection, tokenization for ML models, or fragment analysis in computational genomics and machine learning applications.

Install, configure, and extend Pi, a terminal coding harness, with support for custom providers, models, extensions, skills, packages, themes, SDK integration, RPC mode, JSON event streams, and ecosystem packages for subagent delegation, MCP servers, interactive forms, and web access.

Adds PyTorch FSDP2 (fully_shard) to training scripts with correct init, sharding, mixed precision/offload config, and distributed checkpointing. Use when models exceed single-GPU memory or when you need DTensor-based sharding with DeviceMesh.

Analyzes an HR operations system for headcount planning effectiveness, attrition pattern detection, compensation benchmarking accuracy, workforce analytics maturity, and onboarding process optimization. Evaluates HRIS architecture, pay equity compliance, predictive attrition models, and people analytics governance.

Quantifies how sensitive a language model benchmark's reliability and ranking stability are to specific design choices, such as the selection of scenarios, subscenarios, examples, and few-shot prompts. Use when the user has predictions and gold and needs to compute DIoR.

Drafts enforceable End-User License Agreements for software licensors across desktop, mobile, SaaS, and cloud models. Covers click-wrap formation, IP ownership, liability limitations, data privacy compliance (GDPR/CCPA/COPPA), and export controls. Use when drafting software license agreements, app store terms, SaaS subscription agreements, or trial/freemium license terms.

All-in-one fullstack dev engine. /aura: 46 modes (build/fix/clean/deploy/review/spec/lore/ax/experiment/payment/debug/qa/orchestrate/escalate+), 6-layer security with 32 hooks, tiered models (ZERO/ECO/PRO/MAX), 8 languages, 16 specialized agents, SPEC/EARS/TRUST5/XLOOP/RALF/Autopus absorbed. ~55% token savings.

Generate images, videos, audio, and more using fal.ai AI models, with support for text-to-image, image-to-video, text-to-speech, speech-to-text, image editing, upscaling, model search, workflow creation, and cost estimation.

Routes to 259 Salesforce admin skill packages for declarative configuration tasks like objects, fields, record types, page layouts, permission sets, reports, and record-access models. Use when the request mentions these topics to find and open the exact skill package.

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.

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.

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.

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.

Comprehensive patterns for AI-powered document understanding including PDF parsing, OCR, invoice/receipt extraction, table extraction, multimodal RAG with vision models, and structured data output. Use when "document parsing, PDF extraction, OCR, invoice processing, receipt extraction, document understanding, LlamaParse, Unstructured, vision document, table extraction, structured output from PDF, " mentioned.

Econometrics skill for OLS regression and linear models. Activates when the user asks about: "run OLS", "linear regression", "ordinary least squares", "interpret regression results", "heteroskedasticity", "multicollinearity", "regression assumptions", "robust standard errors", "GLS", "WLS", "fit a regression model", "check regression diagnostics", "OLS假设", "最小二乘法", "线性回归", "回归系数", "残差检验", "异方差", "多重共线性", "普通最小二乘", "稳健标准误", "回归诊断"

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.

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.

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.

Models cascading what-if scenarios across all business functions to prepare for compound adversity, using a structured six-step framework that maps variables, domain impacts, cascade effects, severity levels, early warning signals, and hedging strategies.

Perform survival analysis and time-to-event modeling in Python using scikit-survival, including Cox models, random survival forests, gradient boosting, survival SVMs, and evaluation metrics like concordance index and Brier score.

Performs financial ratio analysis, DCF valuation, budget variance analysis, and rolling forecast construction for strategic decision-making. Use when analyzing financial statements, building valuation models, assessing budget variances, or constructing financial projections and forecasts. Also applicable when users mention financial modeling, cash flow analysis, company valuation, financial projections, or spreadsheet analysis.

Econometrics skill for OLS regression and linear models. Activates when the user asks about: "run OLS", "linear regression", "ordinary least squares", "interpret regression results", "heteroskedasticity", "multicollinearity", "regression assumptions", "robust standard errors", "GLS", "WLS", "fit a regression model", "check regression diagnostics", "OLS假设", "最小二乘法", "线性回归", "回归系数", "残差检验", "异方差", "多重共线性", "普通最小二乘", "稳健标准误", "回归诊断"

Performs financial ratio analysis, DCF valuation, budget variance analysis, and rolling forecast construction for strategic decision-making. Use when analyzing financial statements, building valuation models, assessing budget variances, or constructing financial projections and forecasts. Also applicable when users mention financial modeling, cash flow analysis, company valuation, financial projections, or spreadsheet analysis.

Financial leadership for startups and scaling companies. Financial modeling, unit economics, fundraising strategy, cash management, and board financial packages. Use when building financial models, analyzing unit economics, planning fundraising, managing cash runway, preparing board materials, or when user mentions CFO, burn rate, runway, fundraising, unit economics, LTV, CAC, term sheets, or financial strategy.

Automated hypothesis generation and testing using large language models. Use this skill when generating scientific hypotheses from datasets, combining literature insights with empirical data, testing hypotheses against observational data, or conducting systematic hypothesis exploration for research discovery in domains like deception detection, AI content detection, mental health analysis, or other empirical research tasks.

Marketing leadership for scaling companies. Brand positioning, growth model design, marketing budget allocation, and marketing org design. Use when designing brand strategy, selecting growth models (PLG vs sales-led vs community-led), allocating marketing budgets, building marketing teams, or when user mentions CMO, brand strategy, growth model, CAC, LTV, channel mix, or marketing ROI.

Use for writing, reviewing, debugging, testing, or optimizing Python CVXPY optimization models. Trigger on Variable, Parameter, Expression, Constraint, Objective, Problem, DCP, DPP, DGP, DQCP, solver selection/status, dual values, mixed-integer, cone, or repeated parametric solves. Do not use for scipy.optimize-only, PyMC inference, symbolic algebra without optimization, or hand-written solver implementations.

You are an expert in vLLM, the high-throughput LLM serving engine. You help developers deploy open-source models (Llama, Mistral, Qwen, Phi, Gemma) with PagedAttention for efficient memory management, continuous batching, tensor parallelism for multi-GPU, OpenAI-compatible API, and quantization support — achieving 2-24x higher throughput than HuggingFace Transformers for production LLM serving.

Use for writing, reviewing, debugging, testing, or optimizing Python SymPy symbolic mathematics. Trigger on Symbol, assumptions, Expr, Eq, solve/solveset, simplify, factor, expand, calculus, matrices, exact arithmetic, lambdify, code generation, or symbolic-to-numeric conversion. Do not use for NumPy-only arrays, mpmath-only arbitrary-precision numerics, CVXPY optimization models, or parsing untrusted mathematical text.

Use to design overnight and follow-the-sun support — coverage models, handoff quality and measuring what is lost between shifts. Trigger for "after-hours coverage", "follow the sun", overnight support design, night shift vs defer async, handoff packet, shift handover quality, global coverage gaps, or when customers reopen tickets after every handoff.

v5.6.0 joint multi-TF model: single model per symbol with broadcast 1Hour context replaces dual 15Min/1Hour models. Trigger: (1) replacing weighted-voting model aggregation, (2) adding broadcast features to vectorized env, (3) limited training data + worried about overfitting from doubling obs_dim, (4) backtest builder mismatch with newer feature counts.

Build predictive models for sports and event outcomes using statistical methods, ELO ratings, regression, Monte Carlo simulation, and machine learning. Use when creating power rankings, projecting game outcomes, estimating win probabilities, or building a quantitative edge. Also trigger for 'prediction model', 'ELO rating', 'power rankings', 'win probability', 'Monte Carlo', 'regression model', 'expected goals', or 'predictive analytics'.