Downloads · 30 days
9
10% of all-time downloads
MMHusnain/semeval26-psysemev-extraction-Evidence-2026-05-17-1834
semeval26-psysemev-extraction-Evidence-2026-05-17-1834 is a token classification model from MMHusnain. Use it when you need labels on individual words, such as names. It is set up for peft. The card lists the license as mit.
RoBERTa-large with LoRA for Evidence span extraction (SemEval 2026 Task 10, marker extraction subtask).
Downloads · 30 days
9
10% of all-time downloads
All-time downloads
91
Public
Repo size
28.4 MB
Likes
0
Public
Click a slice to open those files.
.safetensors28.4 MB · 85%
From the Hugging Face model README
RoBERTa-large with LoRA for Evidence span extraction (SemEval 2026 Task 10, marker extraction subtask).
Spans denoting supporting information or proof (e.g. articles, data, documents, testimony) cited in the text.
roberta-large + PEFT LoRA (RobertaForTokenClassification).Research and evaluation on SemEval 2026 marker extraction.
from transformers import RobertaForTokenClassification, RobertaTokenizerFast
from peft import PeftModel
repo_id = "MMHusnain/semeval26-psysemev-extraction-Evidence-2026-05-17-1834"
base_model = "roberta-large"
tokenizer = RobertaTokenizerFast.from_pretrained(repo_id, add_prefix_space=True)
model = RobertaForTokenClassification.from_pretrained(base_model, num_labels=2)
model = PeftModel.from_pretrained(model, repo_id)
Trained under Markers-Extraction/ with LoRA.
SemEval-2026 Task 10: Psycholinguistic Markers of Conspiracy Theories in Social Media Conversations.
| Field | Value |
|---|---|
| Hub repo | MMHusnain/semeval26-psysemev-extraction-Evidence-2026-05-17-1834 |
| Marker type | Evidence |
| Upload batch | 2026-05-17-1834 |
| Base model | roberta-large |