|
Welcome to **RoBERTArg**! |
|
|
|
π€ **Training**: Model (RoBERTA_base_) fine-tuned on a heterogeneous argument corpus (~40K sentences, Stab et al. (2018) π) corpus of several controversial topics (abortion etc.). |
|
|
|
π· Using **NON-ARGUMENT** (0) and **ARGUMENT** (1) as labels. |
|
|
|
**Performance** |
|
|
|
| Model | Acc | F1 | R arg | R non | P arg | P non | |
|
|----|----|----|----|----|----|----| |
|
| RoBERTArg | 0.8193 | 0.8021 | 0.8463 | 0.7986 | 0.7623 | 0.8719 | |
|
|
|
**Confusion Matrix** |
|
|
|
| | 1 | 0 | |
|
|----|----|----| |
|
| 1 | 2213 | 558 | |
|
| 0 | 325 | 1790 | |
|
|
|
ππΎ Check out _chkla/argument-analyzer/_ for more details. |
|
|
|
Enjoy and stay tuned! π |