Spaces:
Running
Running
File size: 1,246 Bytes
f7a0325 e5b0f7e f7a0325 f544e76 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 | ---
title: LBW Guard Colab Tests
emoji: 🧪
colorFrom: blue
colorTo: green
sdk: gradio
python_version: "3.10"
app_file: app.py
suggested_hardware: cpu-basic
models:
- Qwen/Qwen2.5-0.5B
datasets:
- Salesforce/wikitext
tags:
- optimizer
- training
- colab
- gradio
---
Copyright (c) Qluon Inc. All rights reserved.
Provided for Learn-By-Wire Guard evaluation and customer testing under the applicable Qluon license terms.
# LBW Guard Colab Tests
This Space packages the LBW Guard customer Colab notebooks:
- `LBW_Guard_Easy_Test_COLAB.ipynb`: fast AdamW vs `lbw_guard` smoke test.
- `LBW_Guard_Ablation_Test_COLAB.ipynb`: ablation-style AdamW baseline and `lbw_guard` comparison.
Use the app to download either notebook, then open it in Google Colab with a GPU runtime.
The notebooks install `LBW-Guard` from Python package distribution and do not import local source folders.
## Related Paper
**Learn-by-Wire Training Control Governance: Bounded Autonomous Training Under Stress for Stability and Efficiency**
- Author: Anis Radianis
- arXiv: https://arxiv.org/abs/2605.19008
- DOI: https://doi.org/10.48550/arXiv.2605.19008
If you use these notebooks or the `lbw-guard` package in evaluation, please cite the paper above.
|