← Back to Explorer
On-Chain

BERT-base-uncased (2018)

BERT-base-uncased is a large language model released by Google, released under the Apache-2.0 license, with a claimed knowledge cutoff of 2018-10-11. Its weights, configuration, and tokenizer files have been cryptographically hashed and permanently attested on Ethereum (Sepolia) by Weight Registry.

Description

Bidirectional Transformer encoder pretrained via masked-language-modeling on BooksCorpus and Wikipedia; made transfer learning standard practice across NLP.

Metadata

Type
LLM
Issuer
Google
License
Apache-2.0
Size
420.05 MB
Knowledge cutoff
2018-10-11
Source
https://huggingface.co/google-bert/bert-base-uncased
Revision
86b5e0934494bd15c9632b12f734a8a67f723594
Attester
0x6a94784135e413A474E6fA28d1A36C5d180B70c9

Hugging Face stats

43,907,375 downloads on Hugging Face · 3,375 likes · pipeline: fill-mask · library: transformers

transformerspytorchtfjaxrustcoremlonnxsafetensorsbertfill-mask

Weight Hashes

SHA-2560x68d45e234eb4a928074dfd868cead0219ab85354cc53d20e772753c6bb9169d3
SHA3-2560x5511e0d86cfc48410f856806c16035c65051cb3334468077947c4da232b4b7ff
BLAKE30x688f0d6b61a2234c2ed3af1aaf10610ac1980c9d5103afd766d85939317c4a69

Verify it yourself

Download the file(s) from the source below, then hash them locally (sha256sum <file> on Linux/macOS, certutil -hashfile <file> SHA256 on Windows) and compare the output against the hash(es) shown above.