Text Classification
Transformers
PyTorch
Safetensors
Spanish
bert
biomedical
clinical
eHR
spanish
bert-base-spanish-wwm-cased
Eval Results (legacy)
text-embeddings-inference
Instructions to use IIC/bert-base-spanish-wwm-cased-cantemist with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use IIC/bert-base-spanish-wwm-cased-cantemist with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="IIC/bert-base-spanish-wwm-cased-cantemist")# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("IIC/bert-base-spanish-wwm-cased-cantemist") model = AutoModelForSequenceClassification.from_pretrained("IIC/bert-base-spanish-wwm-cased-cantemist", device_map="auto") - Notebooks
- Google Colab
- Kaggle
metadata
language: es
tags:
- biomedical
- clinical
- eHR
- spanish
- bert-base-spanish-wwm-cased
license: cc-by-4.0
datasets:
- PlanTL-GOB-ES/cantemist-ner
metrics:
- f1
model-index:
- name: IIC/bert-base-spanish-wwm-cased-cantemist
results:
- task:
type: token-classification
dataset:
name: cantemist-ner
type: PlanTL-GOB-ES/cantemist-ner
metrics:
- name: f1
type: f1
value: 0.898
widget:
- text: >-
El diagnóstico definitivo de nuestro paciente fue de un Adenocarcinoma de
pulmón cT2a cN3 cM1a Estadio IV (por una única lesión pulmonar
contralateral) PD-L1 90%, EGFR negativo, ALK negativo y ROS-1 negativo.
- text: >-
Durante el ingreso se realiza una TC, observándose un nódulo pulmonar en
el LII y una masa renal derecha indeterminada. Se realiza punción biopsia
del nódulo pulmonar, con hallazgos altamente sospechosos de carcinoma.
- text: >-
Trombosis paraneoplásica con sospecha de hepatocarcinoma por imagen, sobre
hígado cirrótico, en paciente con índice Child-Pugh B.
bert-base-spanish-wwm-cased-cantemist
This model is a finetuned version of bert-base-spanish-wwm-cased for the cantemist dataset used in a benchmark in the paper A comparative analysis of Spanish Clinical encoder-based models on NER and classification tasks. The model has a F1 of 0.898
Please refer to the original publication for more information.
Parameters used
| parameter | Value |
|---|---|
| batch size | 64 |
| learning rate | 4e05 |
| classifier dropout | 0.1 |
| warmup ratio | 0 |
| warmup steps | 0 |
| weight decay | 0 |
| optimizer | AdamW |
| epochs | 10 |
| early stopping patience | 3 |
BibTeX entry and citation info
@article{10.1093/jamia/ocae054,
author = {García Subies, Guillem and Barbero Jiménez, Álvaro and Martínez Fernández, Paloma},
title = {A comparative analysis of Spanish Clinical encoder-based models on NER and classification tasks},
journal = {Journal of the American Medical Informatics Association},
volume = {31},
number = {9},
pages = {2137-2146},
year = {2024},
month = {03},
issn = {1527-974X},
doi = {10.1093/jamia/ocae054},
url = {https://doi.org/10.1093/jamia/ocae054},
}