pytorch skills

huggingface/transformers

Skill Claude CodeCodex

Fixes broken typing checks detected by ty, make typing, or make check-repo. Use when typing errors appear in local runs, CI, or PR logs.

huggingface/diffusers

Skill Claude CodeCodex

Use when the user has written (or wants to write) a ModularPipelineBlocks subclass in a local Python file and needs to package it into a Hub-uploadable directory. Covers the workflow from a single block.py file to a published custom-block repo that consumers can load via ModularPipeline.frompretrained( …

huggingface/diffusers

Skill Claude CodeCodex

Use when the user wants to run a diffusers pipeline from a terminal (one-off generation, batch jobs, smoke-testing a new model), run on HF Sandbox hardware via --remote, introspect a pipeline's input schema before calling it, or attach a LoRA at inference time. Prefer this over writing ad-hoc Python scripts for…

huggingface/diffusers

Skill Claude CodeCodex

Use when adding a new model or pipeline to diffusers, setting up file structure for a new model, converting a pipeline to modular format, or converting weights for a new version of an already-supported model.

stas00/ml-engineering

Skill Claude CodeCodex

Field-tested methodology and concrete recipes for training and operating large-scale LLM/VLM/multi-modal models end to end - choosing and benchmarking accelerators, storage and network; SLURM/Kubernetes orchestration; maximizing training throughput and fitting models in memory; diagnosing and surviving training…

NVIDIA/TensorRT-LLM

Skill Claude CodeCodex

Translates a HuggingFace model into a prefill-only AutoDeploy custom model using reference custom ops, validates with hierarchical equivalence tests.

NVIDIA/TensorRT-LLM

Skill Claude CodeCodex

Run TensorRT-LLM test cases, benchmarks, evaluations, or custom scripts by checking the environment (local GPU or Slurm), selecting the appropriate Docker image, and executing either locally or via Slurm job submission. Accepts pre-built command strings — command construction for trtllm-bench, trtllm-eval, and…

NVIDIA/TensorRT-LLM

Skill Claude CodeCodex

Onboard a HuggingFace multimodal model (vision/audio/video + text) to the TensorRT-LLM PyTorch backend. Use when writing a new tensorrtllm/torch/models/modeling .py plus its input processor and weight mapper, or extending an existing VLM. Not for AutoDeploy — use ad-model-onboard for that path.

flashinfer-ai/flashinfer

Skill Claude CodeCodex

Step-by-step tutorial for adding new CUDA kernels to FlashInfer.

flashinfer-ai/flashinfer

Skill Claude CodeCodex

Guide for benchmarking FlashInfer kernels with CUPTI timing.

flashinfer-ai/flashinfer

Skill Claude CodeCodex

Tutorial for debugging CUDA crashes using API logging.

remsky/Kokoro-FastAPI

Skill Claude CodeCodex

Contributing to the Kokoro-FastAPI Python API: module layout, endpoint gating pattern, test expectations. Use when adding or changing endpoints, services, or inference code.

remsky/Kokoro-FastAPI

Skill Claude CodeCodex

Running the Kokoro-FastAPI benchmark + transcription-roundtrip suites and regenerating the README plots. Use when asked to run/refresh benchmarks, RTF/first-token plots, transcription sanity checks, or the long-form baseline.

remsky/Kokoro-FastAPI

Skill Claude CodeCodex

Contributing to the Kokoro-FastAPI web player: vanilla JS constraints, MSE/audio gotchas, unit and e2e test setup. Use when changing anything under web/.

huggingface/huggingface_hub

Skill Claude CodeCodex

Generate Hugging Face Hub (huggingfacehub) release notes from cached PR JSON files. Use when asked to draft release notes from PR files.

NVIDIA/physicsnemo

Skill Claude CodeCodex

Official NVIDIA-authored guidance for navigating PhysicsNeMo — pick the model, datapipe, or example for a SciML/AI4Science task (surrogates, forecasting, downscaling, physics-informed, inverse, generative). Points at existing files via live repo search; never writes code. Do NOT use for installation or environment…

NVIDIA/physicsnemo

Skill Claude CodeCodex

Official NVIDIA-authored guidance for PhysicsNeMo ShardTensor domain parallelism — integrate domain parallelism into training/inference scripts (new or existing) with DDP or FSDP2, write and register shard patches to enable new layers/ops, and bootstrap multi-GPU correctness tests. Use when working with ShardTensor…

stas00/the-art-of-debugging

Skill Claude CodeCodex

Systematic methodology and concrete tool recipes for debugging Unix, Python, and PyTorch programs - crashes, hangs, segfaults, wrong output, CUDA OOM, NaN/Inf, slowness, and multi-node/multi-GPU issues. Use when a program crashes, hangs, deadlocks, segfaults, runs out of memory (OOM), produces NaN/Inf or wrong…

PharMolix/OpenBioMed

Skill Claude CodeCodex

Query CZ CELLxGENE Census (61M+ cells). Filter by cell type/tissue/disease, retrieve expression data, and integrate with scanpy/PyTorch for population-scale single-cell analysis. Use this skill when: (1) Querying single-cell expression data by cell type, tissue, or disease, (2) Exploring available single-cell datasets…

PINA-org/PINA

Skill Claude CodeCodex

../../../.opencode/skills/condition-setup/SKILL.md.

PINA-org/PINA

Skill Claude CodeCodex

../../../.opencode/skills/create-problem/SKILL.md.

PINA-org/PINA

Skill Claude CodeCodex

Audits SKILL.md files in this repo's skills directory for references to functions, classes, or modules (mentioned by name in prose, e.g. parseinvoice) that no longer exist in the codebase because they were renamed, moved, or deleted. Proposes a targeted fix to the one skill being checked, never a blanket rewrite. ONLY…

graphsignal/graphsignal-profiler

Skill Claude CodeCodex

Set up the Graphsignal Profiler for inference workloads — vLLM, SGLang, PyTorch, and dstack services. Use when the user wants GPU profiling, tracing, or monitoring for inference, asks about graphsignal-run or graphsignal.watch(), or asks about CUPTI / Prometheus / OTLP setup.

frgfm/torch-scan

Skill Claude CodeCodex

Inspect and compare PyTorch models with TorchScan reports, operator FLOPs, and peak-memory workloads. Use when an agent must analyze model structure, parameters, compute, memory, regressions, unsupported operations, or an owner-provided model budget without inventing completeness.