AIInstituteLNCC/manaca-1b-instruct
Full fine-tune vs AIInstituteLNCC/manaca-1b-base. Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: Norms 1/49, MLP 144/144, Attention 192/192
Full fine-tune vs AIInstituteLNCC/manaca-1b-base. Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: Norms 1/49, MLP 144/144, Attention 192/192