Views
No views yet
google/gemma-4-31b-it, fine-tuned for Relation Classification on a Romanian translation of SemEval-2010 Task 8 plus the original English data. Given a sentence with two marked entities (<e1>, <e2>), the model assigns one of ten directional relations.| Language | F1-Score | Accuracy |
|---|---|---|
| English | 0.880 | 0.868 |
| Romanian | 0.865 | 0.850 |
google/gemma-4-31b-it, loaded in 4-bit1from transformers import AutoModelForCausalLM, AutoTokenizer
2from peft import PeftModel
3
4base = AutoModelForCausalLM.from_pretrained("google/gemma-4-31b-it", load_in_4bit=True, device_map="auto")
5model = PeftModel.from_pretrained(base, "<HF_REPO_RE>")
6tok = AutoTokenizer.from_pretrained("<HF_REPO_RE>")run_inference.py script in the code repository.1@misc{vasile2026crosslingual,
2 title = {Cross-lingual Relation Extraction with Large Language Models: Zero-Shot, Few-Shot, and Fine-Tuned Evaluation on Romanian},
3 author = {Vasile, Drago\c{s}-Mitru\c{t} and Apostol, Elena-Simona and Toma, \c{S}tefan-Adrian and Paschke, Adrian and Truic\u{a}, Ciprian-Octavian},
4 year = {2026},
5 note = {Preprint}
6}