mlm-spanish-roberta-base

This model has a RoBERTa base architecture and was trained from scratch with 3.6 GB of raw text over 10 epochs. 4 Tesla V-100 GPUs were used for the training.

To test the quality of the resulting model we evaluate it over the GLUES benchmark for Spanish NLU. The results are the following:

Task Score (metric)
XNLI 71.99 (accuracy)
Paraphrasing 74.85 (accuracy)
NER 85.34 (F1)
POS 97.49 (accuracy)
Dependency Parsing 85.14/81.08 (UAS/LAS)
Document Classification 93.00 (accuracy)
Downloads last month
54
Safetensors
Model size
0.1B params
Tensor type
I64
ยท
F32
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for MMG/mlm-spanish-roberta-base

Finetunes
2 models