Link the in-browser demo Space and describe the ONNX files
Browse files
README.md
CHANGED
|
@@ -16,6 +16,8 @@ Fine-tuned on **6208** reviews from the CodedLang dataset
|
|
| 16 |
([Wan, Li & Huang, ACL 2026](https://doi.org/10.18653/v1/2026.acl-long.426),
|
| 17 |
[RuyuanWan/CodedLang](https://huggingface.co/datasets/RuyuanWan/CodedLang), CC BY 4.0).
|
| 18 |
|
|
|
|
|
|
|
| 19 |
## Held-out test metrics (this checkpoint, 763 reviews never seen during training)
|
| 20 |
|
| 21 |
- precision: 0.92
|
|
|
|
| 16 |
([Wan, Li & Huang, ACL 2026](https://doi.org/10.18653/v1/2026.acl-long.426),
|
| 17 |
[RuyuanWan/CodedLang](https://huggingface.co/datasets/RuyuanWan/CodedLang), CC BY 4.0).
|
| 18 |
|
| 19 |
+
**Try it in your browser:** [nopp23/codedlang-detector-demo](https://huggingface.co/spaces/nopp23/codedlang-detector-demo) (runs locally via Transformers.js, nothing is sent to a server). The demo loads `onnx/model_quantized.onnx`, a per-channel 8-bit copy; on the held-out test it scored F1 0.891 vs 0.885 for the full model, with 1 of 763 reviews decided differently. `onnx/model.onnx` is the full-precision export and matches PyTorch exactly on the same split.
|
| 20 |
+
|
| 21 |
## Held-out test metrics (this checkpoint, 763 reviews never seen during training)
|
| 22 |
|
| 23 |
- precision: 0.92
|