AGENT SKILL

serving-export-and-deployment

Guides agents serving Ludwig models, building prediction payloads, exporting artifacts, using Ray/KServe/vLLM shims, and uploading model outputs.

License 许可
Apache-2.0
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
serving-export-and-deployment
license
Apache 2.0
metadata
{ "disco-role": "operating" }
description
Guides agents serving Ludwig models, building prediction payloads, exporting artifacts, using Ray/KServe/vLLM shims, and uploading model outputs.
disable-model-invocation
true

02 资源文件 5

03 SKILL.md 正文

Serving, Export, and Deployment

Use this sub-skill when the task is about ludwig serve, FastAPI endpoints, Ray Serve, KServe, vLLM serving, request payloads, export formats, MLflow export, or Hub uploads.

Start here

  1. Confirm a complete trained model artifact exists. Use the inspection helper from prediction-evaluation-and-inspection if needed.
  2. Check serving/export optional dependencies:
python scripts/check_serving_prereqs.py
  1. Build a payload before starting a server:
python scripts/build_serving_payload.py --config config.yaml --mode predict --output payload.json
  1. Read workflows.md for local server, Ray Serve, KServe, vLLM, export, and upload patterns.

Core commands

ludwig serve --model_path results/experiment_run/model --host 0.0.0.0 --port 8000
ludwig export_model --model_path results/experiment_run/model --output_path exported_model --format safetensors
ludwig export_mlflow --model_path results/experiment_run/model --output_path mlflow_model
ludwig upload hf_hub --repo_id owner/name --model_path results/experiment_run/model

Do not run servers or uploads unless the user approves long-running processes or remote side effects.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351