An update to the Jamba hybrid models, which use Attention and Mamba layers to support a long context of 256K tokens.
Specs
Params94B active / 398B total
LicenseJamba Open Model License
Tags
modelsmodels/llmsmultilingualartifacts/8
Related Models
An update to the Jamba hybrid models, which use Attention and Mamba layers to support a long context of 256K tokens.
