Here are the success numbers

================ VALIDATION (ALL LABELS) ================
              precision    recall  f1-score   support

    CARDINAL     0.7386    0.8614    0.7953       938
        DATE     0.7042    0.8388    0.7656      1507
       EVENT     0.3429    0.3357    0.3392       143
    LANGUAGE     0.6250    0.6061    0.6154        33
         LAW     0.2388    0.4000    0.2991        40
       MONEY     0.7675    0.8796    0.8197       274
        NORP     0.8095    0.8831    0.8447       847
     ORDINAL     0.8156    0.8578    0.8361       232
         ORG     0.6110    0.6833    0.6451      1740
     PERCENT     0.8042    0.8588    0.8306       177
      PERSON     0.8274    0.9089    0.8662      2020
       PLACE     0.8259    0.8964    0.8597      2587
     PRODUCT     0.3017    0.4861    0.3723        72
    QUANTITY     0.5625    0.7200    0.6316       100
        TIME     0.5567    0.7336    0.6331       214
 WORK_OF_ART     0.2260    0.2817    0.2508       142

   micro avg     0.7330    0.8263    0.7769     11066
   macro avg     0.6098    0.7019    0.6503     11066
weighted avg     0.7375    0.8263    0.7789     11066

F1 overall: 0.7768573977316172

================ VALIDATION (TARGET ONLY) ================
              precision    recall  f1-score   support

      PERSON     0.8274    0.9089    0.8662      2020
       PLACE     0.8259    0.8964    0.8597      2587
     PRODUCT     0.3017    0.4861    0.3723        72
 WORK_OF_ART     0.2260    0.2817    0.2508       142

   micro avg     0.7951    0.8774    0.8342      4821
   macro avg     0.5452    0.6433    0.5873      4821
weighted avg     0.8010    0.8774    0.8372      4821

F1 target: 0.8342372547086085
Downloads last month
22
Safetensors
Model size
14.3M params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support