SAVRN
Search Contact SAVRN

Open-weight model · Fill mask

bert-base-chinese

by BERT community google-bert/bert-base-chinese

This model has been pre-trained for Chinese, training and random input masking has been applied independently to word pieces (as in the original BERT paper).

Parameters103M
Context512
Weights1.7 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads970.9k

Runs On

What it takes to serve bert-base-chinese (103M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 0.2 GB 0.2 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 0.1 GB 0.1 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 0.1 GB 0.1 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

SAVRN's Notes on bert-base-chinese

We still see this encoder in production more than most people would guess: 970,946 downloads a month for a 103 million parameter masked language model published March 2, 2022. Fill mask is the task, the piece under a Chinese text classifier or embedding job, not a chat model. Memory is trivial, 0.2 GB at 16-bit and 0.1 GB at 8-bit or 4-bit. The cheapest Index host is an MI300X with 192 GB at $1.85 an hour, the wrong size for this alone; pack it with other services.

Apache 2.0 permits commercial use, modification and redistribution, with notices kept, significant changes stated and a patent grant included. Access is open, and the weights ship in safetensors, pytorch, jax and tf, so the serving stack is your choice. The limit to check is context: 512 tokens over a 21,128 entry vocabulary, so long documents get chunked before they reach the model.

SAVRN Research, 2026-09-18

Model Card

By BERT community, published under apache-2.0, revision 8f23c25b06e1.

Table of Contents

  • Model Details
  • Uses
  • Risks, Limitations and Biases
  • Training
  • Evaluation
  • How to Get Started With the Model

Model Details

Model Description

This model has been pre-trained for Chinese, training and random input masking has been applied independently to word pieces (as in the original BERT paper).

  • Developed by: Google
  • Model Type: Fill-Mask
  • Language(s): Chinese
  • License: Apache 2.0
  • Parent Model: See the BERT base uncased model for more information about the BERT base model.

Model Sources

  • GitHub repo: https://github.com/google-research/bert/blob/master/multilingual.md
  • Paper: BERT

Uses

Direct Use

This model can be used for masked language modeling

Risks, Limitations and Biases

CONTENT WARNING: Readers should be aware this section contains content that is disturbing, offensive, and can propagate historical and current stereotypes.

Significant research has explored bias and fairness issues with language models (see, e.g., Sheng et al. (2021) and Bender et al. (2021)).

Training

Training Procedure

  • type_vocab_size: 2
  • vocab_size: 21128
  • num_hidden_layers: 12

Training Data

[More Information Needed]

Evaluation

Results

Read the full model card (180 words)

Configuration

Architecture
BertForMaskedLM
Context length (tokens)
512
Layers
12
Hidden size
768
Feed-forward size
3,072
Attention heads
12
Vocabulary size
21,128
Model type
bert

Identity and Version

Repository
google-bert/bert-base-chinese
Publisher
BERT community
Task
Fill mask
Modality
Text
Library
transformers
Parameters
103M parameters
Languages
zh
Revision
8f23c25b06e129b6c986331a13d8d025a92cf0ea
First published
2022-03-02
Last updated
2025-07-03

Files and Weights

10 files, 1.7 GB in total. The weights are 4 files totalling 1.7 GB in bin, h5, msgpack, safetensors.

Weights4 files · 1.7 GB
Configuration1 file · 624 B
Tokenizer3 files · 378.5 KB
Documentation1 file · 1.9 KB
Repository1 file · 445 B
Every file
FileTypeSizeSHA-256
flax_model.msgpackWeights409.2 MB 76df8425215f
model.safetensorsWeights411.6 MB 3404a1ffd8da
pytorch_model.binWeights411.6 MB 8a693db616ea
tf_model.h5Weights478.3 MB 612acd33db45
config.jsonConfiguration624 B —
README.mdDocumentation1.9 KB —
.gitattributesRepository445 B —
tokenizer.jsonTokenizer268.9 KB —
tokenizer_config.jsonTokenizer49 B —
vocab.txtTokenizer109.5 KB —

License and Download

License
apache-2.0
Access
Open weights, no gate
Download size
1.7 GB
Download from BERT community

Released by BERT community through its official repository on Hugging Face. Read the license.

Built From

Memory Requirements

PrecisionWeights in memory
As published1.7 GB
16-bit0.2 GB
8-bit0.1 GB
4-bit0.1 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Compare bert-base-chinese

Questions About bert-base-chinese

How much GPU memory does bert-base-chinese need?

About 0.2 GB at 16-bit and 0.1 GB at 4-bit: the weights (103M parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run bert-base-chinese on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use bert-base-chinese commercially?

Yes. bert-base-chinese is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

What is bert-base-chinese's context length?

512 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Fill mask

bert-base-cased

BERT community

Pretrained model on English language using a masked language modeling (MLM) objective. It was introduced in this paper and first released in this repository. This model is case-sensitive: it makes a difference between english and English. Disclaimer: The team releasing BERT did not write a model card for this model so this model card has been written by the Hugging Face team. BERT is a transformers model pretrained on a large corpus of English data in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from…

Open weights apache-2.0 109M parameters 512 tokens transformers
View model

Eval data: Conll03 (NER), GermEval14 (NER), GermEval18 (Classification), GNAD (Classification) Update April 3rd, 2020: we updated the vocabulary file on deepset's s3 to conform with the default tokenization of punctuation tokens. For details see the related FARM issue. If you want to use the old vocab we have also uploaded a "deepset/bert-base-german-cased-oldvocab" model. - We trained using Google's Tensorflow code on a single cloud TPU v2 with standard settings. - We trained 810k steps with a batch size of 1024 for sequence length 128 and 30k steps with sequence length 512. Training took about 9 days. - As training data we used the latest German Wikipedia dump (6GB of raw txt files), the…

Open weights mit 110M parameters 512 tokens transformers
View model

Model · Fill mask

bert-base-uncased

BERT community

Pretrained model on English language using a masked language modeling (MLM) objective. It was introduced in this paper and first released in this repository. This model is uncased: it does not make a difference between english and English. Disclaimer: The team releasing BERT did not write a model card for this model so this model card has been written by the Hugging Face team. BERT is a transformers model pretrained on a large corpus of English data in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labeling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from…

Open weights apache-2.0 110M parameters 512 tokens transformers
View model

Model · Fill mask

juribert-base

DaSciM

JuriBERT is a set of BERT models (tiny, mini, small and base) pre-trained from scratch on French legal-domain specific corpora. JuriBERT models are pretrained on 6.3GB of legal french raw text from two different sources: the first dataset is crawled from Légifrance and the other one consists of anonymized court’s decisions and the pleadings from the Court of Cassation (mémoires ampliatifs). The latter contains more than 100k long documents from different court cases. It is now on Hugging Face in four different versions with varying number of parameters. JuriBERT was trained and evaluated at École Polytechnique in collaboration with HEC Paris by Stella Douka, Hadi Abdine, Mihcalis…

Open weights mit 111M parameters 512 tokens transformers
View model

Model · Fill mask

camembert-base

ALMAnaCH (Inria)

CamemBERT is a state-of-the-art language model for French based on the RoBERTa model. It is now available on Hugging Face in 6 different versions with varying number of parameters, amount of pretraining data and pretraining data source domains. CamemBERT was trained and evaluated by Louis Martin\, Benjamin Muller\, Pedro Javier Ortiz Suárez\, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot. If you use our work, please cite

Open weights mit 111M parameters 514 tokens transformers
View model

Model · Fill mask

SecBERT

Kun

This is the pretrained model presented in SecBERT: A Pretrained Language Model for Cyber Security Text, which is a BERT model trained on cyber security text. The training corpus was papers taken from SecBERT has its own wordpiece vocabulary (secvocab) that's built to best match the training corpus. We trained SecBERT and SecRoBERTa versions. We proposed to build language model which work on cyber security text, as result, it can improve downstream tasks (NER, Text Classification, Semantic Understand, Q&A) in Cyber Security Domain. First, as below shows Fill-Mask pipeline in Google Bert, AllenAI SciBert and our SecBERT. The original repo can be found here.

Open weights apache-2.0 84M parameters 514 tokens transformers
View model