google/mobilebert-uncased

1---2language: en3thumbnail: https://huggingface.co/front/thumbnails/google.png4 5license: apache-2.06---7 8## MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices9 10MobileBERT is a thin version of BERT_LARGE, while equipped with bottleneck structures and a carefully designed balance11between self-attentions and feed-forward networks.12 13This checkpoint is the original MobileBert Optimized Uncased English: 14[uncased_L-24_H-128_B-512_A-4_F-4_OPT](https://storage.googleapis.com/cloud-tpu-checkpoints/mobilebert/uncased_L-24_H-128_B-512_A-4_F-4_OPT.tar.gz) 15checkpoint.16 17## How to use MobileBERT in `transformers`18 19```python20from transformers import pipeline21 22fill_mask = pipeline(23	"fill-mask",24	model="google/mobilebert-uncased",25	tokenizer="google/mobilebert-uncased"26)27 28print(29	fill_mask(f"HuggingFace is creating a {fill_mask.tokenizer.mask_token} that the community uses to solve NLP tasks.")30)31 32```33