Thoth Ai, Türkçe hakaret ve küfürleri tespit etmek için oluşturulmuştur. Akademik projelerde kaynak gösterilmesi halinde kullanılabilir.

Validation Metrics

  • Loss: 0.230
  • Accuracy: 0.936
  • Macro F1: 0.927
  • Micro F1: 0.936
  • Weighted F1: 0.936
  • Macro Precision: 0.929
  • Micro Precision: 0.936
  • Weighted Precision: 0.936
  • Macro Recall: 0.925
  • Micro Recall: 0.936
  • Weighted Recall: 0.936

from transformers import AutoModelForSequenceClassification, AutoTokenizer

model = AutoModelForSequenceClassification.from_pretrained("thothai/turkce-kufur-tespiti", use_auth_token=True)

tokenizer = AutoTokenizer.from_pretrained("thothai/turkce-kufur-tespiti", use_auth_token=True)

inputs = tokenizer("Merhaba", return_tensors="pt")

outputs = model(**inputs) ```

Downloads last month
37
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API: The model has no pipeline_tag.