Project AURIONAI Use-Cases Research & Innovation
Hugging Face
language encoderfamily

DistilBERT

Smaller, faster BERT made by knowledge distillation during pre-training, keeping most of BERT's accuracy for on-device and low-cost text tasks.

Hugging FaceHugging FaceUSA / France
Figure from the DistilBERT paper: Parameter counts of several recently released pretrained language models.
Levelfamily
Classlanguage encoder
Parametersnot in the notes
Contextnot in the notes
Released2 Oct 2019
Opennessopen weights permissive
Runs onnot in the notes
Statusactive

Lineage

Hugging FaceDistilBERT base multilingual cased belongs to this familylanguage encoder135.45M
Hugging FaceDistilBERT base uncased belongs to this familylanguage encoder66.99M
Hugging FaceDistilRoBERTa base belongs to this familylanguage encoder82.76M
Where this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. Confidence and evidence are the notes' own.

R&D Topics

Compact text encoders through distillationexemplar

Identity

Developer
Hugging Face
Family
not in the notes
Released
2 Oct 2019 sourcesannouncement page
Weights published
–
Record checked
12 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
dense transformer
Total parameters
– sources
Active parameters
–
Checkpoint parameters
–
Context length
not in the notes
Input
text
Output
text
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights permissive
Licence
Apache License 2.0 sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
not in the notes
Footprint
not in the notes
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
listed by developer
Regional languages
Burmese, Javanese, Malay, Minangkabau, Sundanese, Tagalog, Thai, Vietnamese
Jakarta hosting
unknown
Languages
Checkpoint model cards list 103 languages, including Indonesian
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Benchmark Results

The notes record no benchmark result for this model.

Industry Use Cases

No industry use case links this model.

Sources

  1. arxiv.org/abs/1910.01108 https://arxiv.org/abs/1910.01108
  2. jumpstart-cache-prod-us-west-2.s3.us-west-2.amazonaws.com/models_manifest.json https://jumpstart-cache-prod-us-west-2.s3.us-west-2.amazonaws.com/models_manifest.json
  3. aihub.qualcomm.com/models/distil_bert_base_uncased_hf https://aihub.qualcomm.com/models/distil_bert_base_uncased_hf
  4. huggingface.co/distilbert/distilbert-base-uncased https://huggingface.co/distilbert/distilbert-base-uncased
  5. huggingface.co/distilbert/distilbert-base-multilingual-cased https://huggingface.co/distilbert/distilbert-base-multilingual-cased

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "DistilBERT", https://aurion.semai.id/models/hugging-face-distilbert/, records of October 2026.