Aleph Alpha releases German AI model with 78 GB of FP8 weights
Kolibri has 78.1 billion parameters, activates 3.46 billion per token and is released under Apache 2.0. Its model card calls for at least two A100 80GB GPUs.
By Ryan Merket · Published

Kolibri has 78.1 billion parameters, activates 3.46 billion per token and is released under Apache 2.0. Its model card calls for at least two A100 80GB GPUs.