Downloads · 30 days
9
16% of all-time downloads
wiflore/SciBETO-IMRaD
SciBETO-IMRaD is a text classification model from wiflore. Use it when you need a label for a piece of text. The card lists the license as mit.
Fine-tuned SciBETO-large para clasificación de segmentos de papers científicos en español en estructura IMRaD de 8 clases: INTRO, BACK, METH, RES, DISC, CONC, CONTR, LIM.
Downloads · 30 days
9
16% of all-time downloads
All-time downloads
58
Public
Parameters
355M
1.4 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors1.4 GB · 100%
From the Hugging Face model README
Fine-tuned SciBETO-large para clasificación de segmentos de papers científicos en español
en estructura IMRaD de 8 clases:
INTRO, BACK, METH, RES, DISC, CONC, CONTR, LIM.
Flaglab/SciBETO-largedocumento_id, seed=42| Métrica | 4 clases | 8 clases |
|---|---|---|
| F1-macro | 0.5781 | 0.7083 |
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
tokenizer = AutoTokenizer.from_pretrained("wiflore/SciBETO-IMRaD")
model = AutoModelForSequenceClassification.from_pretrained("wiflore/SciBETO-IMRaD")
text = "En este estudio proponemos un método para..."
inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=512)
with torch.no_grad():
logits = model(**inputs).logits
pred = logits.argmax(-1).item()
labels = ['INTRO','BACK','METH','RES','DISC','CONC','CONTR','LIM']
print(labels[pred])