model-zoo-and-converters
Route detrex model zoo selection, pretrained backbones, and safe checkpoint conversion.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- model-zoo-and-converters
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Route detrex model zoo selection, pretrained backbones, and safe checkpoint conversion.
- disable-model-invocation
- true
02 资源文件 5
- 参考文档references/backbones.md2.2 KB
- 参考文档references/converters.md3.1 KB
- 参考文档references/project-guide.md3.8 KB
- 参考文档references/troubleshooting.md2.5 KB
- 脚本scripts/checkpoint_tools.py22.1 KB
03 SKILL.md 正文
model-zoo-and-converters
Use this sub-skill when you need to:
- choose the right detrex project family or pretrained variant;
- decide whether a model zoo row is a converted checkpoint, a hacked-trainer run, or a standard detrex config;
- inspect a checkpoint before converting it;
- convert an official DETR-family checkpoint into detrex format;
- route DINO, MaskDINO, and CO-MOT work to the right project notes.
Read in this order
- Project guide to identify the family and trainer/config route.
- Converters if the checkpoint needs key remapping or class-head reshaping.
- Backbones if the problem is pretrained backbone selection or shape compatibility.
- Troubleshooting for mismatch, head-shape, or missing-key failures.
- Checkpoint tools to inspect first, then run a bounded local conversion.
Guardrails
- Prefer inspect mode before conversion.
- Keep conversions local; do not fetch checkpoints in this skill.
- If a checkpoint already looks detrex-shaped, do not reconvert it.
- If the family is not one of the bundled converter modes, stop and route to the closest project reference.
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).