Announcement
OVHcloud unveils Olala mini 7A1B—and says the GPU is optional
OVHcloud says its seven-billion-parameter model was trained on European public supercomputers and is available on Hugging Face.
TLDR
Olala mini 7A1B kicks off OVHcloud’s first language-model family, the company says. It describes a seven-billion-parameter model that uses only a fraction of its parameters per request and can run on a standard CPU. OVHcloud says it is releasing model weights, a pre-trained base model and a technical report, with translation and summarisation variants also in the range.
Combined views
6.7K
2 Sources, first seen ago

Comments