wev-ai 0.1.0__tar.gz

wev_ai-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 2026 Jared Palmer
+
+   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.

wev_ai-0.1.0/NOTICE

@@ -0,0 +1,44 @@
+wev
+Copyright 2026 the wev authors
+
+This product includes software adapted from kev
+(https://github.com/jaredpalmer/kev), licensed under the Apache License 2.0:
+  - wev/model.py: sequence packing with reserved Qwen delimiter tokens,
+    unforgeable option boundaries, block-causal branch mask with per-branch
+    position restart, pointer readout head.
+  - wev/api.py: System One request/response shapes and rendering.
+
+wev/prompts.py contains instruction text copied from jev-ultrafast
+(https://github.com/browser-use/jev-ultrafast), used under the MIT License:
+
+  MIT License
+  Copyright (c) 2026 Browser Use
+
+  Permission is hereby granted, free of charge, to any person obtaining a copy
+  of this software and associated documentation files (the "Software"), to deal
+  in the Software without restriction, including without limitation the rights
+  to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+  copies of the Software, and to permit persons to whom the Software is
+  furnished to do so, subject to the following conditions:
+
+  The above copyright notice and this permission notice shall be included in all
+  copies or substantial portions of the Software.
+
+  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+  IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+  FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+  AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+  LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+  OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+  SOFTWARE.
+
+Training data is derived from Mind2Web (https://huggingface.co/datasets/osunlp/Mind2Web),
+licensed CC BY 4.0: Deng et al., "Mind2Web: Towards a Generalist Agent for the Web", NeurIPS 2023
+(arXiv:2306.06070). Converted requests are a modified form of the original data.
+
+Training data is also derived from NNetNav (https://huggingface.co/datasets/stanfordnlp/nnetnav-live),
+licensed Apache-2.0: Murty et al., "NNetNav: Unsupervised Learning of Browser Agents Through Environment
+Interaction in the Wild" (arXiv:2410.02907).
+
+wev is not affiliated with or endorsed by TypeSafe AI. "Jev" and "System One" are referenced only to describe
+API compatibility.

wev_ai-0.1.0/PKG-INFO

@@ -0,0 +1,297 @@
+Metadata-Version: 2.4
+Name: wev-ai
+Version: 0.1.0
+Summary: Local System-One decision model: typed questions in, calibrated probabilities out, one forward pass. General decisions and browser-agent steps; drop-in POST /v1/systemone.
+License-Expression: Apache-2.0
+Project-URL: Homepage, https://github.com/alanhuangyoo/wev
+Project-URL: Models, https://huggingface.co/alanhuangya
+Project-URL: Issues, https://github.com/alanhuangyoo/wev/issues
+Keywords: decision-model,system-one,browser-agent,llm,lora,qwen,calibration
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Requires-Python: <3.14,>=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+License-File: NOTICE
+Requires-Dist: torch>=2.8
+Requires-Dist: transformers<5,>=4.56
+Requires-Dist: peft>=0.15
+Requires-Dist: pydantic>=2
+Requires-Dist: numpy
+Requires-Dist: safetensors
+Requires-Dist: huggingface_hub
+Provides-Extra: serve
+Requires-Dist: fastapi; extra == "serve"
+Requires-Dist: uvicorn; extra == "serve"
+Provides-Extra: train
+Requires-Dist: datasets>=3.0; extra == "train"
+Requires-Dist: lxml>=5; extra == "train"
+Provides-Extra: dev
+Requires-Dist: pytest; extra == "dev"
+Requires-Dist: datasets>=3.0; extra == "dev"
+Requires-Dist: lxml>=5; extra == "dev"
+Dynamic: license-file
+
+<div align="center">
+
+# wev
+
+**Local System-One decision models for agents.**<br>
+Typed questions in, calibrated probabilities out, in one forward pass. No text generation, no API key.
+
+[![Models](https://img.shields.io/badge/%F0%9F%A4%97%20Models-1.7B%20%7C%204B%20%7C%208B-ffcc4d)](https://huggingface.co/collections/alanhuangya/wev-6ab4eb5d872c9ae9fd68faa6)
+[![Dataset](https://img.shields.io/badge/%F0%9F%A4%97%20Dataset-wev--data-ffcc4d)](https://huggingface.co/datasets/alanhuangya/wev-data)
+[![pip](https://img.shields.io/badge/pip%20install-wev--ai-3775a9?logo=pypi&logoColor=white)](https://pypi.org/project/wev-ai/)
+[![Paper](https://img.shields.io/badge/Paper-PDF-b31b1b)](https://github.com/alanhuangyoo/wev/blob/main/paper/icml/main.pdf)
+[![License](https://img.shields.io/badge/License-Apache--2.0-2ea44f)](https://github.com/alanhuangyoo/wev/blob/main/LICENSE)
+
+[Quickstart](#quickstart) · [Results](#results) · [How it works](#how-it-works) · [Train your own](#train-your-own) · [Citation](#citation)
+
+</div>
+
+<p align="center">
+  <img src="https://raw.githubusercontent.com/alanhuangyoo/wev/main/assets/overview.png" width="88%" alt="Interface distillation: a browser agent's typed requests are answered by a teacher LLM on live websites, an LLM judge keeps the verified episodes, and the wev decision model trained on them serves the same interface locally.">
+</p>
+
+`wev` answers the `POST /v1/systemone` request shape: a free-form state plus any number of questions, each a
+**choice**, a **yes/no** or a **score**. It handles general decisions such as triage, routing, policy checks and
+agent monitoring, and it handles browser-agent steps: *which operation?* and *which element?* It runs on your own
+GPU, or on a laptop.
+
+## Highlights
+
+- 🌐 **Browser steps on unseen websites: 76% right**, where open decision models trained on general data reach at
+  most 21%.
+- 🧠 **General decisions stay strong.** wev beats Laya on every general benchmark, and wev-8b is level with Kev-8B out
+  of domain.
+- 🎓 **As good as its teacher, locally.** As the System One of an open browser agent, wev completes as many live
+  tasks as the LLM it was distilled from.
+- ⚡ **Fast.** 10–37 ms for a general decision and 91–322 ms for a full browser page on one RTX 5090; 77 ms on a
+  laptop.
+- 📏 **Calibrated.** Probabilities you can threshold: raise the bar for DONE and early stops drop from 8.5% to 3.8%.
+
+## Models
+
+| Model | Size | Best for | General decision | Browser step |
+|---|---|---|---|---|
+| [**wev-4b**](https://huggingface.co/alanhuangya/wev-4b) | 8.1 GB | the default: one consumer GPU, best on live websites | 15 ms | 217 ms |
+| [wev-8b](https://huggingface.co/alanhuangya/wev-8b) | 15.2 GB | the most accurate out of domain | 21 ms | 322 ms |
+| [wev-1.7b](https://huggingface.co/alanhuangya/wev-1.7b) | 3.5 GB | laptops and small GPUs; the fastest | 10 ms | 91 ms |
+
+*Median latency on one RTX 5090 (bf16), one request at a time.*
+
+## Quickstart
+
+```bash
+pip install "wev-ai[serve]"
+```
+
+```python
+import wev
+
+m = wev.load("alanhuangya/wev-4b")   # downloads once from the Hugging Face Hub
+out = m.predict(
+    state="Refund request: order #4411 arrived damaged, customer attached photos, first refund this year.",
+    questions={
+        "action": {"type": "choice", "instructions": "What should support do?",
+                   "criteria": {"refund": "Refund the order.", "replace": "Ship a replacement.",
+                                "escalate": "Send to a human agent."}},
+        "fraud_risk": {"type": "noul", "instructions": "This request looks fraudulent.",
+                       "criteria": {"true": "Likely fraud.", "false": "No sign of fraud."}},
+    },
+)
+out["answers"]
+# {'action': {'type': 'choice', 'choice': 'replace', 'probabilities': {'refund': 0.38, 'replace': 0.51, ...}},
+#  'fraud_risk': {'type': 'noul', 'noul': 0.02}}
+```
+
+**Serve it over HTTP** with the same request and response shapes as a System One API:
+
+```bash
+wev serve --model alanhuangya/wev-4b --port 8009
+curl -s localhost:8009/v1/systemone -H 'content-type: application/json' -d '{"state": "...", "questions": {...}}'
+```
+
+**Use it in a browser agent.** An agent written for a System One API, such as
+[jev-ultrafast](https://github.com/browser-use/jev-ultrafast), can call `http://127.0.0.1:8009/v1/systemone`
+instead of the hosted endpoint. jev-ultrafast hard-codes that URL in `jev_ultrafast/model.py` (`SYSTEM_ONE_URL`), so
+change that one line.
+
+## Results
+
+<p align="center"><img src="https://raw.githubusercontent.com/alanhuangyoo/wev/main/assets/teaser.png" width="52%" alt="Browser step success against out-of-domain general accuracy for wev, Kev and Laya."></p>
+
+Every model below receives the same requests and is scored the same way: per-question accuracy, with each model's
+most probable option taken as its answer. All numbers are on held-out test splits.
+
+**General typed decisions**
+
+| Model | Kev decision-v7 | Kev transfer-v4 | typed-decisions |
+|---|:---:|:---:|:---:|
+| **wev-8b** | 82.4 | 77.2 | 79.1 |
+| **wev-4b** | 80.6 | 73.8 | 79.4 |
+| **wev-1.7b** | 81.1 | 65.5 | **79.5** |
+| Kev-4B | **88.2** | **82.1** | 65.1 |
+| Kev-8B | 88.1 | 76.8 | 62.7 |
+| Laya (typed-decisions) | 65.7 | 62.8 | 76.8 |
+| Laya | 64.3 | 63.7 | 36.2 |
+
+**Browser steps.** Mind2Web test split: 873 requests on websites unseen in training. A step counts when both the
+operation and the target element are right.
+
+| Model | Step success | Operation |
+|---|:---:|:---:|
+| **wev-4b** | **75.9** | **91.2** |
+| **wev-8b** | 75.5 | 90.3 |
+| **wev-1.7b** | 68.2 | 88.1 |
+| Kev-4B | 21.2 | 35.7 |
+| Kev-8B | 19.0 | 73.3 |
+| Laya (typed-decisions) | 0.7 | 13.1 |
+| Laya | 0.0 | 2.5 |
+
+**End to end on live websites.** jev-ultrafast ran 153 held-out tasks with each model as its System One. A task
+succeeds when the agent says DONE and an LLM judge, reading the final page, agrees.
+
+| System One | Tasks completed |
+|---|:---:|
+| wev-4b | 30 / 153 |
+| wev-8b | 28 / 153 |
+| qwen3-max, prompted (the teacher) | 27 / 153 |
+
+<details>
+<summary><b>Notes on the comparison</b></summary>
+
+- *Kev decision-v7* is Kev's own training suite; wev also trains on its train split, and Kev leads there by 6–8
+  points. *Kev transfer-v4* is out of domain for every model in the table.
+- wev and Laya (typed-decisions) train on 80% of the typed-decisions train split; Kev and plain Laya do not, so on
+  that column they are generalists.
+- Kev and Laya ran as published, without web training data. wev is evaluated with a 4,096-token state; the 11
+  Mind2Web requests beyond it count as wrong for wev.
+- Live sites change from run to run, so treat end-to-end gaps of a few tasks as noise. wev-4b's gain from its second
+  teacher collection held on a paired comparison (10 tasks gained, 1 lost).
+- Test splits were held out from training and model selection, with one exception: wev-4b and wev-8b each had two
+  candidates, and both were read on test.
+- Raw results are in [`results/`](https://github.com/alanhuangyoo/wev/tree/main/results).
+
+</details>
+
+<details>
+<summary><b>Calibration: choosing when to stop</b></summary>
+
+<p align="center"><img src="https://raw.githubusercontent.com/alanhuangyoo/wev/main/assets/stopping.png" width="55%" alt="DONE recall against premature DONE rate as the threshold on the DONE probability is swept."></p>
+
+The hardest browser decision is when to stop. Accepting DONE only above a probability threshold trades missed stops
+for early ones: at 0.8, wev-4b stops early on 3.8% of unfinished steps (8.5% at 0.5), and wev-8b on 1.7%.
+
+</details>
+
+## How it works
+
+**Interface distillation.** An agent talks to its System One through typed requests. We serve a prompted LLM behind
+that interface while the agent works on live websites, so every logged request and answer is already a training
+example in the decision model's own format. An LLM judge reads each episode's final page and keeps only the episodes
+whose outcome it confirms. It rejected 36% of the teacher's own DONE claims.
+
+**Data.** We combine those episodes with Mind2Web and NNetNav steps converted to the same request format, and with
+general typed-decision corpora. An LLM-judge audit of NNetNav's stopping labels overturned 31% of its DONE labels.
+
+**Model.** A Qwen3 base model with its vocabulary head removed, adapted with LoRA and merged into the released
+weights. Each question sees the state and itself only, so one forward pass answers every question in a request, and
+question order never changes an answer. A pointer head scores each option against the question and normalises the
+scores into probabilities.
+
+```
+<state> page text · elements · recent actions
+<q> which operation? <opt> CLICK </opt> <opt> TYPE_TEXT </opt> … <opt> DONE </opt> <decide>
+<q> which element to CLICK? <opt> [1] Search </opt> <opt> [2] Sign in </opt> … <decide>
+```
+
+The [paper](https://github.com/alanhuangyoo/wev/blob/main/paper/icml/main.pdf) has the full method, ablations and failure analysis.
+
+## Train your own
+
+<details>
+<summary><b>Data, training, evaluation and export commands</b></summary>
+
+```bash
+pip install -e ".[train,serve,dev]"
+
+# data
+python -m wev.mind2web --out data/m2w-v2 --dev_frac 0.15 --test_frac 0.1
+python -m wev.nnetnav  --out data/nnetnav-v2
+python -m wev.general  --out data/general --kev_repo path/to/kev
+python -m wev.external --out data/ext
+# DONE relabelling of NNetNav by an LLM judge (OpenAI-compatible endpoint from WEV_JUDGE_BASE_URL / _API_KEY / _MODEL)
+python scripts/judge_done.py --data data/nnetnav-v2/train.jsonl --out judge-train.jsonl
+python scripts/apply_judge.py --data data/nnetnav-v2 --judge 'judge-{split}.jsonl' --out data/nnetnav-v3-clean
+
+# the wev-4b recipe; dir:K repeats a training file K times. One GPU works too: drop torchrun.
+# wev-8b and wev-1.7b use the same mix without data/teacher-v2.
+torchrun --nproc_per_node 8 -m wev.train --base Qwen/Qwen3-4B-Base --head pointer --lr 1e-4 --epochs 1 \
+  --batch_tokens 6000 --accum 1 --checkpointing 0 --out runs/wev-4b \
+  --data data/m2w-v2,data/nnetnav-v3-clean,data/teacher-v1:3,data/teacher-v2:3,data/general/kev-v7:2,data/general/typed-decisions-train:8,data/ext/tasksource-jev,data/ext/jev-distill,data/ext/td-synth
+
+wev evaluate --model runs/wev-4b --data data/general/kev-transfer-v4 --split dev
+wev export --run runs/wev-4b --out exports/wev-4b --check data/general/typed-decisions/dev.jsonl
+pytest -q tests
+```
+
+The converted browser data and both teacher collections are on the Hub as
+[`alanhuangya/wev-data`](https://huggingface.co/datasets/alanhuangya/wev-data); `validation.jsonl` is read as the
+development split. To collect new teacher episodes, use `scripts/teacher_server.py`, `make_tasks.py`, `collect.py`
+and `build_teacher_data.py`.
+
+</details>
+
+<details>
+<summary><b>Training data and licenses</b></summary>
+
+| Source | License | What it adds |
+|---|---|---|
+| [Mind2Web](https://huggingface.co/datasets/osunlp/Mind2Web) | CC BY 4.0 | human browser steps (click, type, select), split by website |
+| [NNetNav-live](https://huggingface.co/datasets/stanfordnlp/nnetnav-live) | Apache-2.0 | live-web steps; DONE relabelled by an LLM judge |
+| teacher episodes | outputs of qwen3-max | jev-ultrafast on live sites with qwen3-max as System One; judge-verified |
+| [Kev decision-v7](https://github.com/jaredpalmer/kev) | per source | ten public classification and QA sources plus rule-composition records |
+| [typed-decisions](https://huggingface.co/datasets/LocalLLaMA/typed-decisions) | Apache-2.0 | agent and ops workflows (80% of its train split) |
+| [tasksource-jev](https://huggingface.co/datasets/tasksource/tasksource-jev) | mixed, per source task | hundreds of classification tasks recast as decisions |
+| [jev-distill-corpus-v3](https://huggingface.co/datasets/SargeDev/jev-distill-corpus-v3) | Apache-2.0 | synthetic operational scenarios with soft labels |
+| [typed-decisions-synth](https://huggingface.co/datasets/n4ze3m/typed-decisions-synth) | MIT | multi-question cases over 149 domains |
+
+**Use terms.** The weights are released under Apache-2.0, but some training data carries its own terms: several
+tasksource-jev source tasks are licensed for research only, and the teacher episodes are outputs of qwen3-max,
+subject to its provider's terms. Treat the models as research artifacts and check those terms before commercial use.
+
+</details>
+
+## Limitations
+
+- English only. wev makes decisions and does not write text; in jev-ultrafast, typed values come from its separate
+  text model.
+- Browser targets are scored among the candidates the agent lists (8–40 per step), not every element on the page, so
+  the numbers are not comparable to the Mind2Web leaderboard.
+- Rare operations are rarely predicted: BLOCKED and scrolling have low recall. On NNetNav, wev-4b says DONE too early
+  on 8–10% of steps unless you threshold it.
+- On Kev's own suite, Kev is more accurate.
+- wev has not been compared with Jev itself, because we have no API access.
+
+## Citation
+
+Jun Huang and Xin Ren contributed equally (University of Electronic Science and Technology of China).
+
+```bibtex
+@misc{huang2026wev,
+  title  = {wev: Distilling LLM Browser Agents into Open, Local System-One Decision Models},
+  author = {Huang, Jun and Ren, Xin},
+  year   = {2026},
+  url    = {https://github.com/alanhuangyoo/wev}
+}
+```
+
+## Acknowledgements and license
+
+Apache-2.0. The model code builds on [kev](https://github.com/jaredpalmer/kev) (Apache-2.0) and uses instruction text
+from [jev-ultrafast](https://github.com/browser-use/jev-ultrafast) (MIT); the models build on Qwen3 base models
+(Apache-2.0). wev is an independent project, not affiliated with TypeSafe AI, and does not use Jev. See
+[NOTICE](https://github.com/alanhuangyoo/wev/blob/main/NOTICE).