migration-prep
Phase 1 of 1st-gen to 2nd-gen component migration. Use to understand the component, plan breaking changes, and define scope before any refactoring begins.
1.52k repo starsObserved in 1 repos
Development
Development skills
Browse reusable Agent Skills, each with a clear purpose and practical guidance.
migration-setup
Phase 2 of 1st-gen to 2nd-gen component migration. Use to create the 2nd-gen file and folder structure, wire up exports, and confirm the build passes before implementation begins.
1.52k repo starsObserved in 1 repos
Development
migration-styling
Phase 5 of 1st-gen to 2nd-gen component migration. Use to migrate CSS to the 2nd-gen structure, apply Spectrum 2 tokens, and ensure stylelint passes.
1.52k repo starsObserved in 1 repos
Development
adapt-new-diffusion-model
Adapt AutoRound to support a new diffusion model architecture (DiT, UNet, hybrid AR+DiT). Use when a new diffusion model fails quantization, needs custom output configs, requires a custom pipeline function, or is a hybrid architecture with both autoregressive and diffusion components.
1.52k repo starsObserved in 1 repos
Development
adapt-new-llm
Adapt AutoRound to support a new LLM architecture that doesn't work out-of-the-box. Use when quantization fails for a new model type, block detection doesn't find layers, MoE models need unfusing, custom forward passes are needed, or non-standard linear layer types need handling.
1.52k repo starsObserved in 1 repos
Development
add-export-format
Add a new model export format to AutoRound (e.g., auto_round, auto_gptq, auto_awq, gguf, llm_compressor). Use when implementing a new quantized model serialization format, adding a new packing method, or extending export compatibility for deployment frameworks like vLLM, SGLang, or llama.cpp.
1.52k repo starsObserved in 1 repos
Development
add-inference-backend
Add a new hardware inference backend to AutoRound for deploying quantized models (e.g., CUDA/Marlin, Triton, CPU, HPU, ARK). Use when implementing QuantLinear kernels, registering backend capabilities, or enabling quantized model inference on a new hardware platform.
1.52k repo starsObserved in 1 repos
Development
add-quantization-datatype
Add a new quantization data type to AutoRound (e.g., INT, FP8, MXFP, NVFP, GGUF variants). Use when implementing a new weight/activation quantization scheme, registering a new quant function, or extending the data_type registry.
1.52k repo starsObserved in 1 repos
Development
add-vlm-model
Add support for a new Vision-Language Model (VLM) to AutoRound, including multimodal block handler, calibration dataset template, and special model handling. Use when integrating a new VLM like LLaVA, Qwen2-VL, GLM-Image, Phi-Vision, or similar multi-modal models for quantization.
1.52k repo starsObserved in 1 repos
Development
autocorrelation-and-lag-selection
Analyzes time series dynamics with the fast skforecast.stats functions acf, pacf and calculate_lag_autocorrelation. Covers reading ACF/PACF patterns to identify AR/MA orders and seasonality, ranking lags by partial autocorrelation, and feeding the result to the lags argument of any skforecast forecaster. Use when the user wants to understand the dynamics of a series, choose a candidate set of lags before hyperparameter tuning, or replace a slow statsmodels acf/pacf call.
1.51k repo starsObserved in 2 repos
Development