sgl-eval 0.1.0__tar.gz

sgl_eval-0.1.0/LICENSE

@@ -0,0 +1,202 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

sgl_eval-0.1.0/NOTICE

@@ -0,0 +1,33 @@
+sgl-eval
+Copyright 2026 The SGLang Project
+
+Licensed under the Apache License, Version 2.0 (the "License"); see the
+top-level LICENSE file for the full text.
+
+================================================================================
+Third-party vendored sources
+================================================================================
+
+This product includes software developed at NVIDIA Corporation:
+    NeMo-Skills (https://github.com/NVIDIA/NeMo-Skills)
+    Released under the Apache License, Version 2.0.
+
+The vendored slice lives under `sgl_eval/_vendored/nemo_skills/`. The
+exact upstream commit, file list, and per-file provenance are recorded
+in `sgl_eval/_vendored/nemo_skills/SOURCES.yaml`. Every vendored .py
+file carries a banner header pointing back to its upstream path; .yaml
+prompt configs and bundled .txt data files carry the same provenance.
+
+Modifications applied to vendored Python sources at sync time:
+  - Intra-package imports under `nemo_skills.X` are textually rewritten
+    to `sgl_eval._vendored.nemo_skills.X` so the vendored modules
+    resolve under sgl-eval's package namespace. The mapping is recorded
+    in SOURCES.yaml under `import_rewrites`.
+  - `Lean4ProofEvaluator` (and the dependent imports under
+    `nemo_skills.code_execution.*`) is removed from
+    `evaluator/math.py`. We do not run Lean proofs.
+  - `_compute_reward_at_k` is removed from `metrics/math_metrics.py`.
+    We do not use reward-model-driven aggregation.
+
+No content edits beyond the above are made: every other line in a
+vendored file matches its upstream counterpart at the synced commit.

sgl_eval-0.1.0/PKG-INFO

@@ -0,0 +1,199 @@
+Metadata-Version: 2.4
+Name: sgl-eval
+Version: 0.1.0
+Summary: One-click accuracy evaluation harness for SGLang.
+Author: SGLang Team
+License: Apache-2.0
+Project-URL: Homepage, https://github.com/sgl-project/sgl-eval
+Project-URL: Repository, https://github.com/sgl-project/sgl-eval
+Project-URL: Issues, https://github.com/sgl-project/sgl-eval/issues
+Keywords: llm,evaluation,benchmark,sglang,openai-compatible,aime,gpqa,mmlu,gsm8k
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+License-File: NOTICE
+Requires-Dist: openai<3,>=1.0
+Requires-Dist: httpx<1,>=0.24
+Requires-Dist: tqdm<5,>=4.65
+Requires-Dist: pyyaml<7,>=6.0
+Requires-Dist: math_verify[antlr4_9_3]<1,>=0.5
+Requires-Dist: latex2sympy2_extended<2,>=1.0
+Requires-Dist: numpy<3,>=1.24
+Requires-Dist: pandas<3,>=2.0
+Requires-Dist: datasets<5,>=2.14
+Requires-Dist: pillow<12,>=10
+Requires-Dist: editdistance<1,>=0.6
+Provides-Extra: dev
+Requires-Dist: pytest<9,>=7.0; extra == "dev"
+Requires-Dist: pytest-asyncio<2,>=0.23; extra == "dev"
+Requires-Dist: ruff<1,>=0.5; extra == "dev"
+Requires-Dist: black<26,>=24.0; extra == "dev"
+Provides-Extra: longcontext
+Requires-Dist: transformers<6,>=4.40; extra == "longcontext"
+Requires-Dist: nltk<4,>=3.8; extra == "longcontext"
+Requires-Dist: wonderwords<3,>=2.2; extra == "longcontext"
+Requires-Dist: inflect<8,>=7.0; extra == "longcontext"
+Dynamic: license-file
+
+# sgl-eval
+
+[![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](LICENSE)
+[![Python](https://img.shields.io/badge/python-3.10%20%7C%203.11%20%7C%203.12%20%7C%203.13-blue.svg)](https://www.python.org/)
+
+One-click accuracy evaluation harness for [SGLang](https://github.com/sgl-project/sglang).
+
+Point at any OpenAI-compatible endpoint. Scoring logic (graders, evaluators,
+prompts, dataset configs) is vendored from [NeMo-Skills](https://github.com/NVIDIA/NeMo-Skills);
+sgl-eval contributes the transport, runner, and benchmark wiring.
+
+---
+
+## Quick start
+
+```bash
+pip install sgl-eval
+
+sgl-eval ping --base-url http://localhost:30000/v1
+sgl-eval run gsm8k --base-url http://localhost:30000/v1 --num-examples 50
+```
+
+Four subcommands: `run`, `list`, `ping`, `preset`. `sgl-eval run --help` is
+the full flag reference -- endpoint, sampling overrides (`--temperature`,
+`--seed`, `--thinking`, ...), and any flags the benchmark itself adds.
+
+---
+
+## Reading a run
+
+Each run prints the headline metric first -- single-shot accuracy, averaged
+across the `k` repeats when `k > 1` -- and writes the same payload plus
+provenance (model, endpoint, sampling config, vendored NS commit) as
+`metrics.json` under `--out-dir`.
+
+```
+== aime25 ==
+30 examples x 16 repeats  |  823.7s  |  4293 tok/s  |  3.5M tokens
+
+* pass@1[avg-of-16]  =  78.96% +/- 1.21% (SEM 0.30%)
+  pass@16            =  93.33%
+  majority@16        =  93.33%
+  no_answer          =  20.00%  [warn: consider --max-tokens]
+```
+
+While the run is going, the progress bar carries a live accuracy. For a
+sanity check that is usually the whole point: watch it, decide, stop.
+
+```
+gsm8k:  34%|###4      | 452/1319 [02:11<04:12, 3.4it/s, acc=81.42%]
+```
+
+Every scored sample is streamed to
+`<out-dir>/sgl_eval_<name>_<stamp>/output-rs*.jsonl` as it lands (disable
+with `--no-dump-predictions`), so the per-sample record survives however the
+run ends.
+
+---
+
+## Running less than the whole thing
+
+- **`--num-examples N`** -- only the first `N` examples.
+- **`Ctrl-C`** -- kills in-flight requests, keeps everything already scored,
+  and writes `metrics.json` flagged `partial: true` with how much ran, so a
+  half-run can't later be mistaken for a full one. Exits `130`; a second
+  `Ctrl-C` hard-exits if cleanup hangs. The preset `expected_vs_actual`
+  comparison is skipped -- a half-run isn't comparable to a baseline.
+- **`--from-dataset <path>`** -- swap in your own NS-shape JSONL
+  (`{id?, problem, expected_answer}`) for one run. Only the questions
+  change; scoring still goes through the vendored grader.
+
+---
+
+## Benchmarks
+
+`sgl-eval list` for the registered set, `sgl-eval list -v` for each one's
+defaults. See [`benchmarks.md`](benchmarks.md) for the ones that need more
+than an endpoint (today: `ruler2`), and for how to match a NeMo-Skills run.
+
+## Presets
+
+Save a `(benchmark, endpoint, sampling, n_repeats, expected)` bundle to
+`~/.sgl_eval/presets/<name>.yaml` and replay with `sgl-eval run --preset
+<name>`. See [`preset.md`](preset.md) for schema, example, usage, and
+override priority.
+
+For repository-maintained model defaults, select an exact supported model ID:
+
+```bash
+sgl-eval run BENCHMARK \
+  --base-url BASE_URL \
+  --load-preset-from-model-id MODEL_ID
+```
+
+This sets the served model and its recommended generation parameters, but not
+the deployment-specific `--base-url`. See [`preset.md`](preset.md#built-in-model-presets)
+for the supported model list, resolved values, and override priority.
+
+---
+
+## Architecture
+
+**Anything that decides a score is vendored verbatim from NeMo-Skills.**
+sgl-eval contributes only transport: an OpenAI client, a threadpool runner,
+a CLI, and the thin glue that wires upstream pieces into one command.
+
+```
++----------------------------------------------------+
+|  sgl-eval                                          |
+|    cli, sampler, runner, registry, metrics         |
+|    evals/                                          |
++----------------------------------------------------+
+|  vendored from NeMo-Skills                         |
+|    math_grader, evaluator/, metrics/,              |
+|    dataset/<bench>/, prompts/*.yaml                |
++----------------------------------------------------+
+```
+
+The slice is pinned at a specific commit in
+`sgl_eval/_vendored/nemo_skills/SOURCES.yaml`. To upgrade, bump
+`synced_from_sha` there and run:
+
+```bash
+python scripts/sync_vendored.py    # re-fetch all vendored files
+pytest                             # upstream's own tests run against the
+                                   # new slice -- catches behavior drift
+```
+
+Adding a benchmark inside an existing category (`math`, `multichoice`) is one
+row in `_registry.py:_TABLE`. A new category needs a runner alongside it --
+graders are usually already in NeMo-Skills.
+
+---
+
+## Scope
+
+The goal is to be the single accuracy-eval client SGLang's CI calls, in place
+of `sglang.test.run_eval` and the assorted per-test harnesses.
+
+Not in scope: performance benchmarking (latency / throughput / scheduling --
+that is SGLang's `bench_serving.py`; sgl-eval records them only as side
+metrics, never as the headline), training or fine-tuning, multi-server
+orchestration (one endpoint per invocation), and OS-level agent loops.
+
+---
+
+## License
+
+Apache-2.0. See `LICENSE`. Vendored NeMo-Skills sources are also Apache-2.0;
+see `NOTICE` for attribution and the list of vendored files.

sgl_eval-0.1.0/README.md

@@ -0,0 +1,150 @@
+# sgl-eval
+
+[![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](LICENSE)
+[![Python](https://img.shields.io/badge/python-3.10%20%7C%203.11%20%7C%203.12%20%7C%203.13-blue.svg)](https://www.python.org/)
+
+One-click accuracy evaluation harness for [SGLang](https://github.com/sgl-project/sglang).
+
+Point at any OpenAI-compatible endpoint. Scoring logic (graders, evaluators,
+prompts, dataset configs) is vendored from [NeMo-Skills](https://github.com/NVIDIA/NeMo-Skills);
+sgl-eval contributes the transport, runner, and benchmark wiring.
+
+---
+
+## Quick start
+
+```bash
+pip install sgl-eval
+
+sgl-eval ping --base-url http://localhost:30000/v1
+sgl-eval run gsm8k --base-url http://localhost:30000/v1 --num-examples 50
+```
+
+Four subcommands: `run`, `list`, `ping`, `preset`. `sgl-eval run --help` is
+the full flag reference -- endpoint, sampling overrides (`--temperature`,
+`--seed`, `--thinking`, ...), and any flags the benchmark itself adds.
+
+---
+
+## Reading a run
+
+Each run prints the headline metric first -- single-shot accuracy, averaged
+across the `k` repeats when `k > 1` -- and writes the same payload plus
+provenance (model, endpoint, sampling config, vendored NS commit) as
+`metrics.json` under `--out-dir`.
+
+```
+== aime25 ==
+30 examples x 16 repeats  |  823.7s  |  4293 tok/s  |  3.5M tokens
+
+* pass@1[avg-of-16]  =  78.96% +/- 1.21% (SEM 0.30%)
+  pass@16            =  93.33%
+  majority@16        =  93.33%
+  no_answer          =  20.00%  [warn: consider --max-tokens]
+```
+
+While the run is going, the progress bar carries a live accuracy. For a
+sanity check that is usually the whole point: watch it, decide, stop.
+
+```
+gsm8k:  34%|###4      | 452/1319 [02:11<04:12, 3.4it/s, acc=81.42%]
+```
+
+Every scored sample is streamed to
+`<out-dir>/sgl_eval_<name>_<stamp>/output-rs*.jsonl` as it lands (disable
+with `--no-dump-predictions`), so the per-sample record survives however the
+run ends.
+
+---
+
+## Running less than the whole thing
+
+- **`--num-examples N`** -- only the first `N` examples.
+- **`Ctrl-C`** -- kills in-flight requests, keeps everything already scored,
+  and writes `metrics.json` flagged `partial: true` with how much ran, so a
+  half-run can't later be mistaken for a full one. Exits `130`; a second
+  `Ctrl-C` hard-exits if cleanup hangs. The preset `expected_vs_actual`
+  comparison is skipped -- a half-run isn't comparable to a baseline.
+- **`--from-dataset <path>`** -- swap in your own NS-shape JSONL
+  (`{id?, problem, expected_answer}`) for one run. Only the questions
+  change; scoring still goes through the vendored grader.
+
+---
+
+## Benchmarks
+
+`sgl-eval list` for the registered set, `sgl-eval list -v` for each one's
+defaults. See [`benchmarks.md`](benchmarks.md) for the ones that need more
+than an endpoint (today: `ruler2`), and for how to match a NeMo-Skills run.
+
+## Presets
+
+Save a `(benchmark, endpoint, sampling, n_repeats, expected)` bundle to
+`~/.sgl_eval/presets/<name>.yaml` and replay with `sgl-eval run --preset
+<name>`. See [`preset.md`](preset.md) for schema, example, usage, and
+override priority.
+
+For repository-maintained model defaults, select an exact supported model ID:
+
+```bash
+sgl-eval run BENCHMARK \
+  --base-url BASE_URL \
+  --load-preset-from-model-id MODEL_ID
+```
+
+This sets the served model and its recommended generation parameters, but not
+the deployment-specific `--base-url`. See [`preset.md`](preset.md#built-in-model-presets)
+for the supported model list, resolved values, and override priority.
+
+---
+
+## Architecture
+
+**Anything that decides a score is vendored verbatim from NeMo-Skills.**
+sgl-eval contributes only transport: an OpenAI client, a threadpool runner,
+a CLI, and the thin glue that wires upstream pieces into one command.
+
+```
++----------------------------------------------------+
+|  sgl-eval                                          |
+|    cli, sampler, runner, registry, metrics         |
+|    evals/                                          |
++----------------------------------------------------+
+|  vendored from NeMo-Skills                         |
+|    math_grader, evaluator/, metrics/,              |
+|    dataset/<bench>/, prompts/*.yaml                |
++----------------------------------------------------+
+```
+
+The slice is pinned at a specific commit in
+`sgl_eval/_vendored/nemo_skills/SOURCES.yaml`. To upgrade, bump
+`synced_from_sha` there and run:
+
+```bash
+python scripts/sync_vendored.py    # re-fetch all vendored files
+pytest                             # upstream's own tests run against the
+                                   # new slice -- catches behavior drift
+```
+
+Adding a benchmark inside an existing category (`math`, `multichoice`) is one
+row in `_registry.py:_TABLE`. A new category needs a runner alongside it --
+graders are usually already in NeMo-Skills.
+
+---
+
+## Scope
+
+The goal is to be the single accuracy-eval client SGLang's CI calls, in place
+of `sglang.test.run_eval` and the assorted per-test harnesses.
+
+Not in scope: performance benchmarking (latency / throughput / scheduling --
+that is SGLang's `bench_serving.py`; sgl-eval records them only as side
+metrics, never as the headline), training or fine-tuning, multi-server
+orchestration (one endpoint per invocation), and OS-level agent loops.
+
+---
+
+## License
+
+Apache-2.0. See `LICENSE`. Vendored NeMo-Skills sources are also Apache-2.0;
+see `NOTICE` for attribution and the list of vendored files.