| license: mit | |
| language: | |
| - en | |
| metrics: | |
| - accuracy | |
| pipeline_tag: text-classification | |
| library_name: transformers | |
| tags: | |
| - anti-spam | |
| - spam | |
| - spamclass | |
| - english | |
| - English | |
| - EN | |
| - en | |
| ## SpamClass 300k | |
| - acc=0.8347710683477106 | |
| - prec=0.7895869191049913 | |
| - rec=0.7835183603757472 | |
| - f1=0.7865409344192027 | |
| # Trained on a custom corpus. | |
| # 1x T4 was used during training. | |
| It primarily is meant for english spam, its accurate 83% of the time (acceptable) |