Publish experimental guarded KEEP/DELETE cleanup model

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.

NOTICE

@@ -0,0 +1,8 @@
+Transcript cleanup tiny, experimental fine-tune by Arach, 2026.
+Derived from Google BERT-Tiny (google/bert_uncased_L-2_H-128_A-2),
+revision 30b0a37ccaaa32f332884b96992754e246e48c5f, Apache License 2.0.
+Modifications: a new KEEP/DELETE token-classification head and full-model
+fine-tuning on assistant-authored synthetic cleanup examples.
+The tokenizer is redistributed unchanged.
+Base model: https://huggingface.co/google/bert_uncased_L-2_H-128_A-2
+Recipe code has a separate MIT license; Parakeet weights use CC-BY-4.0.

README.md

@@ -0,0 +1,119 @@
+---
+license: apache-2.0
+language:
+  - en
+library_name: transformers
+pipeline_tag: token-classification
+base_model: google/bert_uncased_L-2_H-128_A-2
+base_model_relation: finetune
+tags:
+  - transcript-cleanup
+  - experimental
+  - bert
+  - keep-delete
+---
+
+# Transcript cleanup tiny
+
+An experimental English KEEP/DELETE tagger for an optional cleanup step after transcription. The model proposes word deletions. The [shared recipe runtime](https://huggingface.co/arach/parakeet-coreml-recipe/tree/main/cleanup) checks those proposals and renders edits from the original text.
+
+The frozen pilot matches 39 of 48 synthetic held-out references. It preserves all 24 cases that should remain unchanged and correctly cleans 15 of the 24 cases that need edits. The nine misses remain unchanged. These results do not establish reliability on real dictation.
+
+This repository contains the weights, tokenizer, and training manifest. Runtime code and experiment evidence live in the recipe's `cleanup/` folder. No ASR application or streaming integration is included.
+
+## Model
+
+| Property | Value |
+| --- | --- |
+| Base | [Google BERT-Tiny](https://huggingface.co/google/bert_uncased_L-2_H-128_A-2/tree/30b0a37ccaaa32f332884b96992754e246e48c5f) |
+| Base revision | `30b0a37ccaaa32f332884b96992754e246e48c5f` |
+| Architecture | Two-layer encoder, 128 hidden dimensions, two attention heads |
+| Task head | KEEP / DELETE token classification |
+| Fine-tuning | Full encoder and new classification head |
+| Parameters | 4,369,666, approximately 4.37 million |
+| Weight file | `model.safetensors`, 17,483,080 bytes, approximately 17.48 MB |
+| Deletion threshold | Probability greater than 0.99 |
+| Context limit | 512 tokenizer tokens, including special tokens |
+
+Weight SHA-256: `1d7ec6c86525c1d1072a1889b23493df20fc9f59d6c9e0f412e3b5a3fa086208`.
+
+## Use the guarded runtime
+
+Install the Python dependencies:
+
+```sh
+python -m pip install torch==2.15.0 transformers==5.22.2 huggingface-hub==2.4.0 safetensors==0.8.0
+```
+
+Download the recipe code and model, then keep the `Tagger` instance resident between utterances:
+
+```python
+from pathlib import Path
+import sys
+
+from huggingface_hub import snapshot_download
+
+recipe_path = snapshot_download(
+    repo_id="arach/parakeet-coreml-recipe",
+    revision="e195703e5060c9945c2478c3749dc3508b8eb0b6",
+    allow_patterns=["cleanup/cleanup.py", "cleanup/edit_policy.py"],
+)
+sys.path.insert(0, str(Path(recipe_path) / "cleanup"))
+from cleanup import Tagger
+
+model_path = snapshot_download(
+    repo_id="arach/transcript-cleanup-tiny",
+    # Set revision to a published model commit for reproducible use.
+)
+tagger = Tagger(Path(model_path))
+result = tagger.run("Um, the wicker basket is beside the radiator.")
+print(result["output"])
+# The wicker basket is beside the radiator.
+```
+
+The reported results require the complete `Tagger` pipeline: source alignment, the frozen threshold, eligibility guards, and the original-text renderer. A generic Transformers token-classification pipeline omits these steps. In the held-out evaluation, the model proposed deleting one quoted `uh`; the guard blocked that proposal.
+
+The intended experiment is an optional pass on a completed utterance, with the original transcript retained. Speculative streaming is not implemented. See the [recipe instructions](https://huggingface.co/arach/parakeet-coreml-recipe/blob/main/cleanup/README.md) for the CLI and result fields.
+
+## Edit policy and limits
+
+Eligible edits remove unquoted standalone `um` or `uh`, or one copy of an adjacent repeated word. Repeated words must be separated by horizontal whitespace; comma-separated repetitions are ineligible in this version. The renderer retains at least one copy and repairs neighboring separators. It can capitalize the first surviving letter.
+
+The policy leaves grammar, hedges, emphasis, explicit self-corrections, and incomplete speech intact. Guards protect recognized quoted text, code, URLs, digits, negations, underscore identifiers, and camelCase. They do not recognize every name or literal. The model can still make incorrect eligible deletions on unseen inputs.
+
+Inputs over the context limit or with unsafe source alignment return the original text with a fallback reason. They are not truncated. Edit offsets are Python character indices. The edit list records lexical deletions; separator and capitalization repairs are implicit, so the list is not a complete patch format for an application.
+
+A known prefix-capitalization discrepancy remains: deleting the first `go` in `go go home.` can render `Go home.`, while training-label derivation permits that capitalization change only after a prefix filler. The frozen holdout does not exercise this case. The policy and label derivation need alignment before a subsequent experiment.
+
+## Training and evaluation
+
+All examples are assistant-authored synthetic text. There are no human-approved labels or real ASR captures. The development set contains 160 pairs from 80 families, including 80 unchanged pairs. A family-level split assigns 128 pairs to training and 32 to validation. The model trains for 12 fixed epochs with seed 23, batch size 8, learning rate 0.0003, and DELETE loss weight 4.
+
+Validation selects the threshold by minimizing cases that lose required words, then changed unchanged cases, then reference mismatches. Ties prefer the higher threshold. The selected threshold is 0.99. The weights, threshold, and renderer were frozen before evaluation on the separate 48-case pack.
+
+| System | Reference matches | Correct cleanup | Unchanged cases preserved | Cases losing required words | Warm p50 / p95 |
+| --- | ---: | ---: | ---: | ---: | ---: |
+| Return input | 24/48 | 0/24 | 24/24 | 0 | Approximately 0 ms |
+| Rules with the same renderer | 45/48 | 22/24 | 23/24 | 1 | 0.08 / 0.91 ms |
+| This tagger with guards | 39/48 | 15/24 | 24/24 | 0 | 2.87 / 35.62 ms |
+| Earlier Qwen3-0.6B adapter, restricted prompt | 40/48 | 22/24 | 18/24 | 6 | 383.33 / 725.10 ms |
+
+The tagger's nine misses comprise two unpunctuated fillers, five eligible repetitions below the threshold, and two comma-separated repetitions outside renderer scope. Rules clean more cases but incorrectly shorten grammatical `had had`. The Qwen adapter was trained earlier under a broader cleanup policy and runs in MLX; this is a comparison of practical prototypes, not an isolated architecture comparison.
+
+Reference matching normalizes outer whitespace and quote typography. Unchanged-case preservation uses exact string equality. Required-word loss is a lexical subsequence diagnostic, not proof of semantic fidelity. Four frozen `preserve` annotations contain cleaned phrases absent as contiguous input spans. The comparison report accounts for this diagnostic artifact; raw protected-span failure counts must not be interpreted as lost source text.
+
+An audit found no exact or normalized duplicate sentences and no shared six-word sequences between development and holdout data. Four similar templates remain. Excluding those templates gives 35/44 matches for the tagger, with all 21 unchanged cases preserved. Synthetic independence remains limited.
+
+The [experiment folder](https://huggingface.co/arach/parakeet-coreml-recipe/tree/main/cleanup/experiment) contains the training source, data, and evaluation harness. [Recorded results](https://huggingface.co/arach/parakeet-coreml-recipe/tree/main/cleanup/experiment/results) include the comparison, raw evaluations, and length probes. Training details and hashes are also recorded in `tagger_manifest.json`.
+
+## Timing scope
+
+Measurements use an Apple M1 with 16 GiB RAM, PyTorch CPU inference, four threads, and batch size one. Five serial repeats per held-out case produce 240 timing observations, not 240 independent quality cases. Five warmups are excluded. Times include tokenization, inference, guards, and rendering; they exclude model loading, ASR, and delivery.
+
+Fresh-process initialization, including Python library imports, took 13.50 seconds in the holdout run and 6.07 seconds in the length run. Warm latency assumes a resident model. Overlapping length probes with 20, 71, and 191 body words took 4.81–86.57 ms, 4.41–63.21 ms, and 9.31–48.11 ms respectively. Workstation scheduling varied substantially; these observations are not a latency guarantee or controlled scaling curve.
+
+Native startup, Core ML conversion, process memory, energy, thermal stability, and concurrent ASR latency or memory have not been measured. Real Parakeet transcripts with human-approved minimal edits and a fresh recording-session holdout are needed before application integration.
+
+## Licenses
+
+These weights derive from the Apache-2.0 BERT-Tiny base and are released under Apache-2.0. The shared recipe code is MIT licensed. Parakeet artifacts have a separate CC-BY-4.0 license; that license does not apply to these cleanup weights.

config.json

@@ -0,0 +1,36 @@
+{
+  "add_cross_attention": false,
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": null,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "eos_token_id": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 128,
+  "id2label": {
+    "0": "KEEP",
+    "1": "DELETE"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 512,
+  "is_decoder": false,
+  "label2id": {
+    "DELETE": 1,
+    "KEEP": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 2,
+  "num_hidden_layers": 2,
+  "pad_token_id": 0,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.22.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d7ec6c86525c1d1072a1889b23493df20fc9f59d6c9e0f412e3b5a3fa086208
+size 17483080

publication_manifest.json

@@ -0,0 +1,19 @@
+{
+  "status": "experimental synthetic pilot",
+  "source_commit": "c3b9594",
+  "model_repo": "arach/transcript-cleanup-tiny",
+  "recipe_repo": "arach/parakeet-coreml-recipe",
+  "recipe_base_commit": "3c3658de9aaeaf5700adb6e7b24efcffc7c802af",
+  "recipe_commit": "e195703e5060c9945c2478c3749dc3508b8eb0b6",
+  "model_artifact_sha256": {
+    "model.safetensors": "1d7ec6c86525c1d1072a1889b23493df20fc9f59d6c9e0f412e3b5a3fa086208",
+    "config.json": "c00aee6a050ada6caf07d28cd7ec1a7e89b05df66563b25597a0d44b515572ff",
+    "tokenizer.json": "d241a60d5e8f04cc1b2b3e9ef7a4921b27bf526d9f6050ab90f9267a1f9e5c66",
+    "tokenizer_config.json": "9f0c2c65a70ea18113ffa2e717103e7210ea826e7acc3442b69b346686b55a48",
+    "tagger_manifest.json": "4e414f2dee872a5a76eebb149be4dd6bafe8692f9bbff11da2c5abf295e251a9"
+  },
+  "runtime_sha256": "b7bae82121b2bb7066e167a4c203c3f15612a81e82e0059f04eaa87449b0dba5",
+  "renderer_sha256": "7977564c691039e69c492324cfdb1f1f5cb0f4ac07d5962e7860094f06ab3104",
+  "training_pairs_sha256": "582a5137099bb81351a83eea9c15caf6cd05653fa982b26049166d50ba0f898c",
+  "heldout_sha256": "6a4b7c565e6322c6a5008e377c0ccdc87445cd6d70f49e8a1dfee7d2973cd593"
+}

tagger_manifest.json

@@ -0,0 +1,413 @@
+{
+  "status": "complete",
+  "timestamp": "2026-11-07T02:14:14.279533+00:00",
+  "model": "google/bert_uncased_L-2_H-128_A-2",
+  "revision": "30b0a37ccaaa32f332884b96992754e246e48c5f",
+  "config": {
+    "seed": 23,
+    "epochs": 12,
+    "batch_size": 8,
+    "learning_rate": 0.0003,
+    "weight_decay": 0.01,
+    "delete_loss_weight": 4.0,
+    "max_tokens": 512,
+    "threads": 4,
+    "device": "cpu",
+    "thresholds": [
+      0.5,
+      0.7,
+      0.8,
+      0.9,
+      0.95,
+      0.99,
+      1.0
+    ]
+  },
+  "provenance": "Synthetic assistant-authored experiment; no real captures or user-approved targets.",
+  "parameter_count": 4369666,
+  "pairs_sha256": "582a5137099bb81351a83eea9c15caf6cd05653fa982b26049166d50ba0f898c",
+  "heldout_sha256": "6a4b7c565e6322c6a5008e377c0ccdc87445cd6d70f49e8a1dfee7d2973cd593",
+  "script_sha256": "2ff06711560dae1d79673a6b7e5603d2a758ac636d9c980790af7916ad0f32ed",
+  "renderer_sha256": "7977564c691039e69c492324cfdb1f1f5cb0f4ac07d5962e7860094f06ab3104",
+  "packages": {
+    "torch": "2.15.0",
+    "transformers": "5.22.2",
+    "huggingface-hub": "2.4.0",
+    "safetensors": "0.8.0"
+  },
+  "train_ids": [
+    "seedling-trays-01",
+    "seedling-trays-02",
+    "ferry-ramp-01",
+    "ferry-ramp-02",
+    "spare-reed-01",
+    "spare-reed-02",
+    "wool-pattern-01",
+    "wool-pattern-02",
+    "courtyard-bench-01",
+    "courtyard-bench-02",
+    "storm-shutters-01",
+    "storm-shutters-02",
+    "paper-lantern-01",
+    "paper-lantern-02",
+    "marble-stairs-01",
+    "marble-stairs-02",
+    "lemon-peel-01",
+    "lemon-peel-02",
+    "trail-junction-01",
+    "trail-junction-02",
+    "copper-pan-01",
+    "copper-pan-02",
+    "plum-jam-01",
+    "plum-jam-02",
+    "woven-rug-01",
+    "woven-rug-02",
+    "ginger-tea-01",
+    "ginger-tea-02",
+    "valve-handle-01",
+    "valve-handle-02",
+    "small-funnel-01",
+    "small-funnel-02",
+    "orange-crate-01",
+    "orange-crate-02",
+    "linen-drawer-01",
+    "linen-drawer-02",
+    "box-chalk-01",
+    "box-chalk-02",
+    "silk-thread-01",
+    "silk-thread-02",
+    "winding-path-01",
+    "winding-path-02",
+    "island-postcard-01",
+    "island-postcard-02",
+    "oak-stool-01",
+    "oak-stool-02",
+    "blue-scarf-01",
+    "blue-scarf-02",
+    "skating-gloves-01",
+    "skating-gloves-02",
+    "weekend-route-01",
+    "weekend-route-02",
+    "garden-ladder-01",
+    "garden-ladder-02",
+    "canvas-apron-01",
+    "canvas-apron-02",
+    "concert-seat-01",
+    "concert-seat-02",
+    "faded-map-01",
+    "faded-map-02",
+    "soup-ladle-01",
+    "soup-ladle-02",
+    "quoted-brochure-01",
+    "quoted-brochure-02",
+    "quoted-rehearsal-01",
+    "quoted-rehearsal-02",
+    "quoted-audio-note-01",
+    "quoted-audio-note-02",
+    "very-steep-01",
+    "very-steep-02",
+    "long-winter-01",
+    "long-winter-02",
+    "no-extra-salt-01",
+    "no-extra-salt-02",
+    "had-breakfast-01",
+    "had-breakfast-02",
+    "literal-python-01",
+    "literal-python-02",
+    "literal-path-01",
+    "literal-path-02",
+    "literal-camel-01",
+    "literal-camel-02",
+    "literal-iphone-01",
+    "literal-iphone-02",
+    "correction-reservation-01",
+    "correction-reservation-02",
+    "unfinished-paint-01",
+    "unfinished-paint-02",
+    "unfinished-melody-01",
+    "unfinished-melody-02",
+    "instruction-marigold-01",
+    "instruction-marigold-02",
+    "numeric-repeat-01",
+    "numeric-repeat-02",
+    "hedge-feeling-01",
+    "hedge-feeling-02",
+    "discourse-literal-01",
+    "discourse-literal-02",
+    "grammar-that-01",
+    "grammar-that-02",
+    "grammar-is-01",
+    "grammar-is-02",
+    "emphatic-more-01",
+    "emphatic-more-02",
+    "metalinguistic-duplicates-01",
+    "metalinguistic-duplicates-02",
+    "two-hesitations-01",
+    "two-hesitations-02",
+    "filler-no-commas-01",
+    "filler-no-commas-02",
+    "filler-question-01",
+    "filler-question-02",
+    "filler-fragment-01",
+    "filler-fragment-02",
+    "filler-longer-01",
+    "filler-longer-02",
+    "only-hesitation-01",
+    "only-hesitation-02",
+    "repeat-tentative-01",
+    "repeat-tentative-02",
+    "repeat-negative-01",
+    "repeat-negative-02",
+    "repeat-literal-context-01",
+    "repeat-literal-context-02",
+    "repeat-multiple-01",
+    "repeat-multiple-02",
+    "repeat-with-filler-01",
+    "repeat-with-filler-02"
+  ],
+  "valid_ids": [
+    "ceramic-shelf-01",
+    "ceramic-shelf-02",
+    "bicycle-chain-01",
+    "bicycle-chain-02",
+    "dried-rosemary-01",
+    "dried-rosemary-02",
+    "library-atlas-01",
+    "library-atlas-02",
+    "embroidered-bag-01",
+    "embroidered-bag-02",
+    "daylight-window-01",
+    "daylight-window-02",
+    "ticket-desk-01",
+    "ticket-desk-02",
+    "roof-rain-01",
+    "roof-rain-02",
+    "sketch-frame-01",
+    "sketch-frame-02",
+    "quoted-puppet-01",
+    "quoted-puppet-02",
+    "correction-colour-01",
+    "correction-colour-02",
+    "correction-approval-01",
+    "correction-approval-02",
+    "correction-direction-01",
+    "correction-direction-02",
+    "instruction-dialogue-01",
+    "instruction-dialogue-02",
+    "numeric-contrast-01",
+    "numeric-contrast-02",
+    "repeat-soft-request-01",
+    "repeat-soft-request-02"
+  ],
+  "train_families": [
+    "blue-scarf",
+    "box-chalk",
+    "canvas-apron",
+    "concert-seat",
+    "copper-pan",
+    "correction-reservation",
+    "courtyard-bench",
+    "discourse-literal",
+    "emphatic-more",
+    "faded-map",
+    "ferry-ramp",
+    "filler-fragment",
+    "filler-longer",
+    "filler-no-commas",
+    "filler-question",
+    "garden-ladder",
+    "ginger-tea",
+    "grammar-is",
+    "grammar-that",
+    "had-breakfast",
+    "hedge-feeling",
+    "instruction-marigold",
+    "island-postcard",
+    "lemon-peel",
+    "linen-drawer",
+    "literal-camel",
+    "literal-iphone",
+    "literal-path",
+    "literal-python",
+    "long-winter",
+    "marble-stairs",
+    "metalinguistic-duplicates",
+    "no-extra-salt",
+    "numeric-repeat",
+    "oak-stool",
+    "only-hesitation",
+    "orange-crate",
+    "paper-lantern",
+    "plum-jam",
+    "quoted-audio-note",
+    "quoted-brochure",
+    "quoted-rehearsal",
+    "repeat-literal-context",
+    "repeat-multiple",
+    "repeat-negative",
+    "repeat-tentative",
+    "repeat-with-filler",
+    "seedling-trays",
+    "silk-thread",
+    "skating-gloves",
+    "small-funnel",
+    "soup-ladle",
+    "spare-reed",
+    "storm-shutters",
+    "trail-junction",
+    "two-hesitations",
+    "unfinished-melody",
+    "unfinished-paint",
+    "valve-handle",
+    "very-steep",
+    "weekend-route",
+    "winding-path",
+    "wool-pattern",
+    "woven-rug"
+  ],
+  "valid_families": [
+    "bicycle-chain",
+    "ceramic-shelf",
+    "correction-approval",
+    "correction-colour",
+    "correction-direction",
+    "daylight-window",
+    "dried-rosemary",
+    "embroidered-bag",
+    "instruction-dialogue",
+    "library-atlas",
+    "numeric-contrast",
+    "quoted-puppet",
+    "repeat-soft-request",
+    "roof-rain",
+    "sketch-frame",
+    "ticket-desk"
+  ],
+  "selection_policy": "Final fixed-epoch model. Threshold chosen on validation only: minimize cases losing required words, then changed no-ops, then maximize strict accepted references; prefer higher threshold on ties.",
+  "gold_renderer_mismatches": [],
+  "epochs": [
+    {
+      "epoch": 1,
+      "mean_batch_loss": 0.46413037180900574
+    },
+    {
+      "epoch": 2,
+      "mean_batch_loss": 0.3076215647161007
+    },
+    {
+      "epoch": 3,
+      "mean_batch_loss": 0.21633667964488268
+    },
+    {
+      "epoch": 4,
+      "mean_batch_loss": 0.13570373295806348
+    },
+    {
+      "epoch": 5,
+      "mean_batch_loss": 0.10943305736873299
+    },
+    {
+      "epoch": 6,
+      "mean_batch_loss": 0.08478387643117458
+    },
+    {
+      "epoch": 7,
+      "mean_batch_loss": 0.06221661943709478
+    },
+    {
+      "epoch": 8,
+      "mean_batch_loss": 0.03604193188948557
+    },
+    {
+      "epoch": 9,
+      "mean_batch_loss": 0.06867107070866041
+    },
+    {
+      "epoch": 10,
+      "mean_batch_loss": 0.05189189137308858
+    },
+    {
+      "epoch": 11,
+      "mean_batch_loss": 0.04755593434674665
+    },
+    {
+      "epoch": 12,
+      "mean_batch_loss": 0.045788399802404456
+    }
+  ],
+  "threshold": 0.99,
+  "validation_thresholds": [
+    {
+      "threshold": 0.5,
+      "unique_cases": 32,
+      "accepted_unique": 27,
+      "cases_missing_required_words": 0,
+      "cases_with_new_words": 0,
+      "no_op_cases": 17,
+      "changed_no_op_cases": 0,
+      "scope": "Lexical subsequence diagnostics and exact no-op preservation; not semantic safety. Repeats are timing observations, not independent quality cases."
+    },
+    {
+      "threshold": 0.7,
+      "unique_cases": 32,
+      "accepted_unique": 27,
+      "cases_missing_required_words": 0,
+      "cases_with_new_words": 0,
+      "no_op_cases": 17,
+      "changed_no_op_cases": 0,
+      "scope": "Lexical subsequence diagnostics and exact no-op preservation; not semantic safety. Repeats are timing observations, not independent quality cases."
+    },
+    {
+      "threshold": 0.8,
+      "unique_cases": 32,
+      "accepted_unique": 27,
+      "cases_missing_required_words": 0,
+      "cases_with_new_words": 0,
+      "no_op_cases": 17,
+      "changed_no_op_cases": 0,
+      "scope": "Lexical subsequence diagnostics and exact no-op preservation; not semantic safety. Repeats are timing observations, not independent quality cases."
+    },
+    {
+      "threshold": 0.9,
+      "unique_cases": 32,
+      "accepted_unique": 27,
+      "cases_missing_required_words": 0,
+      "cases_with_new_words": 0,
+      "no_op_cases": 17,
+      "changed_no_op_cases": 0,
+      "scope": "Lexical subsequence diagnostics and exact no-op preservation; not semantic safety. Repeats are timing observations, not independent quality cases."
+    },
+    {
+      "threshold": 0.95,
+      "unique_cases": 32,
+      "accepted_unique": 27,
+      "cases_missing_required_words": 0,
+      "cases_with_new_words": 0,
+      "no_op_cases": 17,
+      "changed_no_op_cases": 0,
+      "scope": "Lexical subsequence diagnostics and exact no-op preservation; not semantic safety. Repeats are timing observations, not independent quality cases."
+    },
+    {
+      "threshold": 0.99,
+      "unique_cases": 32,
+      "accepted_unique": 27,
+      "cases_missing_required_words": 0,
+      "cases_with_new_words": 0,
+      "no_op_cases": 17,
+      "changed_no_op_cases": 0,
+      "scope": "Lexical subsequence diagnostics and exact no-op preservation; not semantic safety. Repeats are timing observations, not independent quality cases."
+    },
+    {
+      "threshold": 1.0,
+      "unique_cases": 32,
+      "accepted_unique": 17,
+      "cases_missing_required_words": 0,
+      "cases_with_new_words": 0,
+      "no_op_cases": 17,
+      "changed_no_op_cases": 0,
+      "scope": "Lexical subsequence diagnostics and exact no-op preservation; not semantic safety. Repeats are timing observations, not independent quality cases."
+    }
+  ],
+  "model_sha256": "1d7ec6c86525c1d1072a1889b23493df20fc9f59d6c9e0f412e3b5a3fa086208",
+  "training_seconds": 4.683
+}

tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "backend": "tokenizers",
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}