A
Turkish educational content classifier that predicts educational quality scores (0-5) for Turkish web text.
This is the Turkish equivalent of
HuggingFaceFW/fineweb-edu-classifier.
1from transformers import AutoTokenizer, AutoModelForSequenceClassification
2import torch
3
4model_name = "YsK-dev/TurkWeb-Edu-Classifier-V3"
5tokenizer = AutoTokenizer.from_pretrained(model_name)
6model = AutoModelForSequenceClassification.from_pretrained(model_name)
7
8text = "Fotosentez, bitkilerin güneş ışığını kullanarak karbondioksit ve suyu glikoz ve oksijene dönüştürdüğü biyokimyasal bir süreçtir."
9
10inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=512)
11with torch.no_grad():
12 score = model(**inputs).logits.squeeze().item()
13
14print(f"Score: {score:.2f}")
15print(f"Int Score: {int(round(max(0, min(score, 5))))}")
16# Expected: High score (4-5) for this educational text about photosynthesis
