← Back to Explorer
On-Chain

Jamba-v0.1 (2024)

Jamba-v0.1 is a large language model released by AI21 Labs, released under the apache-2.0 license, with a claimed knowledge cutoff of 2024-03-28. Its weights, configuration, and tokenizer files have been cryptographically hashed and permanently attested on Ethereum (Sepolia) by Weight Registry.

Description

AI21 Labs' hybrid Transformer-Mamba architecture, combining attention layers with more compute-efficient state-space blocks.

Metadata

Type
LLM
Issuer
AI21 Labs
License
apache-2.0
Size
96.06 GB
Knowledge cutoff
2024-03-28
Source
https://huggingface.co/ai21labs/Jamba-v0.1
Revision
9efd11575ba791d9e3d25d4c8b670e78506b2df7
Attester
0x6a94784135e413A474E6fA28d1A36C5d180B70c9

Hugging Face stats

8,409 downloads on Hugging Face · 1,194 likes · pipeline: text-generation · library: transformers

transformerssafetensorsjambatext-generationmambamoecustom_codeendpoints_compatible

Weight Hashes

Combined commitment: a Merkle root over the weights, config, and tokenizer hashes together (not a raw weights-only hash) -- see this project's v2 on-chain schema.
Combined SHA-2560x2c52f4dcdb554a70815a0fa4d1890685f3261e1bddf6490285b0961980525d3b
Combined SHA3-2560x749cc985e52e2246aed11fe407e6979768c26b211bc8f83335789c1d1420691b
Combined BLAKE30x756c2e6f53a57968f518effe5cf0589b432c3c5c97c9f03587632c191d86f60b

Per-file breakdown

Computed locally prior to attestation; the combined hash above is the canonical on-chain commitment.

Config file
config.json · 1.24 KB
SHA-25698d005b3bdb7b4d3104ed3f8e6e8e45fa484846703012a11010d43aad768e984
SHA3-256887b3c2b9263cddfb471ab8e2e0f48451c128d4ff5b4cb0cf94c0041420fcf5c
BLAKE32de9f122d9521b4ba2797dde85ae7ba79b0635bb60c06ad8f2a4c70780b7ab08
Tokenizer file
tokenizer.json · 4.05 MB
SHA-256168522c4bc2db021984eb9ec0534b2421adee443b7fc038ed2ff8c36e934096a
SHA3-2563bdd36f31a04ff4b93a1c31d8fb694b1704d6dda321b78555de8b984cf0edad6
BLAKE3b4985b0cc15462551ec36b703c34927fc9f17c8c717d3f9656e5b07af9f8cf33

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.