Train-Test Set: "teknofest_train_final.csv"

Model: "dbmdz/bert-base-turkish-128k-uncased"

Önişleme

• Karakterler küçültülmüştür
• Noktalama işaretleri silinmiştir

Tokenizer Parametreleri

max_length=64
padding=True
truncation=True

Eğitim Parametreleri

• Epoch: 3
• Learning Rate: 7e-5
• Batch-Size: 64
• Tokenizer Length: 64
• Loss: BCE
• Online Hard Example Mining: Açık
• Class-Weighting: Açık (^0.3)
• Early Stopping: Kapalı
• Stratified Batch Sampling: Açık
• Gradient Accumulation: Kapalı
• LR Scheduler: Cosine-with-Warmup
• Warmup Ratio: 0.1
• Weight Decay: 0.01
• LLRD: 0.95
• Label Smoothing: 0.05
• Gradient Clipping: 1.0
• MLM Pre-Training: Kapalı

CV10 Sonuçları

              precision    recall  f1-score   support

      INSULT     0.9098    0.9143    0.9120      2393
       OTHER     0.9596    0.9481    0.9538      3528
   PROFANITY     0.9599    0.9575    0.9587      2376
      RACIST     0.9551    0.9636    0.9594      2033
      SEXIST     0.9552    0.9635    0.9593      2081

    accuracy                         0.9485     12411
   macro avg     0.9479    0.9494    0.9486     12411
weighted avg     0.9486    0.9485    0.9485     12411