| # Statement of changes |
|
|
| Apache License 2.0, section 4(b): prominent notice that files carry modifications. |
|
|
| ## What is modified |
|
|
| **No upstream source file is modified.** No file from `Qwen/Qwen3.8-27B` is edited, |
| renamed, patched or redistributed in this repository, with one exception: the |
| upstream licence text is reproduced byte-for-byte at `licenses/Qwen-Apache-2.0.txt` |
| (sha256 `bbedc3fda3305820b977265f01b8619d87570a6739de3a5582c3464840f1e57a`) because section 4(a) requires it. That file is |
| unmodified and carries the upstream copyright. |
|
|
| The modification this work carries is not an edit to a source file. It is a set |
| of **trained parameters applied to the base model at inference time**, plus an |
| original serving layer that reads the model's logits. Concretely: |
|
|
| | Change | Artifact | sha256 | |
| |---|---|---| |
| | LoRA adapter, rank 64, question-side placement, float32 | `adapter/adapter.safetensors` | `d122466d430a058bb6457d919f811160e97fbd20149f4f24ca455c5d83e360a0` | |
| | Trained linear answer heads | `adapter/heads.npz` | `f766d752d7768a419a9657155cf27f042834d9de29392cf7470d8725130e67ab` | |
| | Readout calibration, one positive scalar per answer type | `serving/readout-temperature-v3.json` | file `1a2285d8fd56d17ee1d06a1e9fce866cc0d3b0263730754babb11deea5f1f7c9` / payload `945bad449b7f5ffc88e597277d632fbab81c3c8729e22c8babd3f4a45fe1378b` | |
| | Runtime identity binding (bf16, default) | `serving/serving-binding.json` | payload `0add0efda28902180db757a12160953e0f5d8dc303decc27b2cfda84db900da8` | |
| | Runtime identity binding (fp32) | `serving/serving-binding-fp32.json` | payload `517f263000cf65457751c4fba519221d48ac550e060b241f198b007ae88c59db` | |
| | Runtime identity binding (int8) | `serving/serving-binding-int8.json` | payload `b550254777ceb3f53e7e10e15f7dc9f80f620ddd9c1ed69592190da5f2107f16` | |
| | Experimental evidence head weights | `serving/evidence-head.safetensors` | `5088019adb67e523ef7411cf75e354f1070e5752c132e50f5f3acc4feb1c0f6c` | |
| | Experimental evidence head config | `serving/evidence-head.json` | `b8bf1d642af9f3a830f1ee47c49bda7336d65b56b97fe3625ea1741f133fa244` | |
| | Experimental evidence policy | `serving/evidence-policy.json` | `866888d26a5da9c91ffb1b418334e174a80320ea5f8d2216c2d459f2e25f3c84` | |
| | Serving layer (original work, not derived from upstream code) | `src/` | see `MANIFEST.json` | |
|
|
| Base model, unmodified and not redistributed: `Qwen/Qwen3.8-27B` at revision |
| `1d4bf0f2ff6012fd82039f2fa52739d0dd7c60c0`, Copyright 2026 Alibaba Cloud, Apache-2.0. |
|
|
| The v1.1 adapter and heads were trained on 21 September 2026 (v1.0: 16 to 20 September 2026). The shipped artifacts are |
| identified by the checksums in the table above and in `MANIFEST.json`. |
|
|
| ## Which files carry a change notice |
|
|
| | File | Why | |
| |---|---| |
| | `NOTICE` | Section 4(b) and 4(c) statement, carried with every copy | |
| | `MODIFICATIONS.md` | this file | |
| | `README.md` | names the base model, the pinned revision and the derivative relationship in the front matter and in the body | |
| | `adapter/config.json` | machine-readable record of the base repo, the pinned revision and both weight checksums | |
| | `serving/serving-binding.json` | pins the exact runtime the weights were measured on and refuses to load against a different one | |
|
|
| No file under `src/` carries an upstream change notice, because no file under |
| `src/` contains upstream code. Every file there is original work, written for |
| this project, and is covered by the repository's own Apache-2.0 `LICENSE` and by |
| the copyright line in `NOTICE`. |
|
|
| All of the changes described above — the adapter, the heads, the calibration, the |
| serving binding and the serving layer — are Copyright 2026 |
| Doccy Pty Ltd and licensed under Apache-2.0. |
|
|
| ## Third-party text scan |
|
|
| Before release, **every file staged into this repository was scanned for text |
| originating in third-party source documents.** The scan compared normalised |
| 6-gram and 8-gram shingles of every staged text file against: |
|
|
| 1. the 42 third-party source records the training and evaluation panels were |
| built from (Apache-2.0, MIT, BSD-3-Clause, CC BY 4.0, CC BY-SA 4.0, CC0, |
| OGL v3.0 and US-government public-domain assertions); and |
| 2. every generated panel and document corpus on disk. |
|
|
| **Result: no third-party document text is present in any shipped file.** The |
| only matches were: |
|
|
| * the reproduced Apache-2.0 licence text at `licenses/Qwen-Apache-2.0.txt`, which |
| matches an Apache-2.0 licence document held in the evaluation corpus and is |
| required to be here verbatim; and |
| * the phrase *"A missing fact is not a negative fact"*, which is **our own |
| prompt-template wording** appearing in our own evaluation panels, not |
| third-party text entering our prompts. |
|
|
| Acceptance fixtures are excluded from this repository entirely. The fixture |
| documents used in live acceptance are original synthetic text authored for this |
| project and held in tooling that is not distributed. |
|
|
| Re-run the scan with `ops/solomon_package.py --scan` in the source project. |
|
|
| ## Maintainer notes |
|
|
| These warnings used to live in a `PUSH.md` that carried its own instruction to be |
| deleted before the repository was made public. The repository is public now, so |
| that file is gone — from the manifest and from the tree — and the parts of it that |
| are still true are kept here. |
|
|
| **Never mutate a published revision.** The serving binding pins by hash and |
| consumers pin by revision. If something is wrong with a published revision, push a |
| **new** revision and a **new** tag. Do not force-push over a revision that has |
| already been fetched: a consumer who pinned it would silently get different weights |
| under a hash they already trusted. |
|
|
| **Do not trust a fetch tool's exit code.** The remote-volume fetch used to retrieve the adapter PRINTS "No such file or directory" AND EXITS 0 on a wrong path: a failed download is indistinguishable from a successful one by return code. Never trust the exit status; check that the file exists and that its sha256 matches before treating a fetch as done. |
| The 870 MB adapter in this repository was fetched that way, and the only thing that |
| made the fetch trustworthy was re-hashing the file afterwards. Verify what you |
| downloaded from here the same way: |
|
|
| ```sh |
| shasum -a 256 adapter/adapter.safetensors # d122466d430a058bb6457d919f811160e97fbd20149f4f24ca455c5d83e360a0 |
| shasum -a 256 adapter/heads.npz # f766d752d7768a419a9657155cf27f042834d9de29392cf7470d8725130e67ab |
| ``` |
|
|
| `MANIFEST.json` carries the size and sha256 of every file here. |
|
|
| **Both weight files must go through git-lfs.** `.gitattributes` tracks |
| `*.safetensors` and `*.npz`; confirm with `git lfs ls-files` before committing. An |
| 870 MB blob committed outside LFS has to be undone by rewriting history. |
|
|
| **Every generated file here comes from `release/solomon-release.json`.** Editing a |
| generated file by hand breaks the manifest hash and is caught by |
| `ops/solomon_package.py --enforce` as `DRIFTED_SINCE_STAGING`. Change the descriptor |
| or the generator and re-stage. |
|
|
| **A calibration may not be carried onto a different adapter.** The temperatures were |
| fitted on this model's logits. Refit and register them for a new adapter, or ship no |
| calibration and serve at T = 1.0, which is always permitted. |
|
|
