prompt_iteration_viewer
Apply PromptSource-style templates across multiple examples and summarize browse, sourcing, and review diagnostics for prompt iteration.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- prompt_iteration_viewer
- description
- Apply PromptSource-style templates across multiple examples and summarize browse, sourcing, and review diagnostics for prompt iteration.
02 资源文件 2
- 脚本scripts/iterate_prompts.py3.2 KB
- 脚本tests/test_prompt_iteration_viewer.py2.2 KB
03 SKILL.md 正文
Prompt Iteration Viewer
Use this skill when you need a UI-neutral version of PromptSource's browse/sourcing workflow: render prompt variations over examples, preserve diagnostics, and summarize coverage.
Inputs
examples: JSON list of dataset example dictionaries.templates: JSON list withid,name,template, optionalanswer_choices, and metadata.choice_index: deterministic choice helper index.include_skipped: whether skipped rows should appear in output rows.
Outputs
A JSON report with rows, coverage, variation_summary, and errors.
Workflow
- For every template/example pair, call the prompt template rendering helper.
- Keep raw rendering diagnostics separate from rendered input and target.
- Count produced, skipped, and error rows per template.
- Summarize prompt variation names and target variants.
- Use the report for prompt review or recovery evidence without changing targets.
Validation
Run:
python tests/test_prompt_iteration_viewer.py
python scripts/iterate_prompts.py --examples-json '[{"premise":"A dog runs","hypothesis":"An animal moves","label":0}]' --templates-json '[{"id":"t1","name":"nli","template":"{{premise}} Question: {{hypothesis}}? ||| {{answer_choices[label]}}","answer_choices":["Yes","No","Maybe"]}]'
Limitations
This skill does not launch Streamlit or load remote datasets. It provides the mechanism behind browse/iteration diagnostics on supplied examples.
04 同仓库更多 Skills
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.
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.
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.
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.
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.
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).