jaspionjader/Kosmos-EVAA-gamma-v15-8B

Full fine-tune vs jaspionjader/Kosmos-EVAA-gamma-ultra-light-8B. Fine-tune: 93% of the weights changed, across 32 of 32 layers.

Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 65/65, MLP 88/96, Attention 119/128

1 derivatives checked

• mradermacher/Kosmos-EVAA-gamma-v15-8B-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse