dslim / token-classification Open weights updated 2 years ago

bert-base-NER

Name: bert-base-NER
Author: dslim

If my open source models have been useful to you, please consider supporting me in building small, useful AI models for everyone (and help me afford med school / help out my parents financially). Thanks!

Params

110 M

Context

512

Downloads 30d

1.1 M

Likes

747

Commercial use: allowed mit Not gated SAFETENSORS 1 languages View on Hugging Face ↗

Download history

daily snapshots · 106 days

▲ 303 K in the last 30 days (22.3%)

2.1 M1.0 M

Jul 28Sep 1Oct 6Nov 10

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	f32	0.4 GB	1.0 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Specifications

Architecture: BertForTokenClassification
Parameters: 110 M
Tensor type: F32
Context length: 512
Vocabulary: 28,996
Layers / heads: 12 / 12
Licence: mit
First seen on the Hub: 2022-03-02
Training datasets: conll2003
conll2003 (reported): 0.48325642943382
Added to our catalog: 2026-07-28

Compare with any token-classification model

Popular comparisons

indonesian-roberta-base-posp-tagger 3.6 M stanford-deidentifier-base 1.5 M llmlingua-2-bert-base-multilingual-cased-meetingbank 905 K ner-german-large 754 K wikineural-multilingual-ner 649 K punctuate-all 570 K