24 lines
1.9 KiB
Plaintext
24 lines
1.9 KiB
Plaintext
📚 Loading LIAR dataset...
|
|
🧮 Grouping into binary classes...
|
|
⬇️ Loading model from C:/Users/andre/OneDrive/Documents/code/fake_news_bert...
|
|
Some weights of the model checkpoint at C:/Users/andre/OneDrive/Documents/code/fake_news_bert were not used when initializing DistilBertForSequenceClassification: ['loss_fct.weight']
|
|
- This IS expected if you are initializing DistilBertForSequenceClassification from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
|
|
- This IS NOT expected if you are initializing DistilBertForSequenceClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
|
|
🪙 Tokenizing text...
|
|
Tokenizing: 100%|████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 16.52batch/s]
|
|
|
|
📝 Creating dataset...
|
|
|
|
🧪 Evaluating on LIAR test set...
|
|
Predicting: 100%|█████████████████████████████████████████████████████████████████████████████████████████████| 40/40 [00:09<00:00, 4.01it/s]
|
|
|
|
📊 DistilBERT Performance on LIAR Dataset:
|
|
precision recall f1-score support
|
|
|
|
Reliable 0.74 0.67 0.70 926
|
|
Fake 0.28 0.34 0.31 338
|
|
|
|
accuracy 0.59 1264
|
|
macro avg 0.51 0.51 0.51 1264
|
|
weighted avg 0.61 0.59 0.60 1264
|