MaziyarPanahi/Experiment28M7_Multi_verse_modelM7

Full fine-tune vs automerger/Experiment28M7-7B. Fine-tune: 94% of the weights changed, across 31 of 32 layers.

Tensors changed: Embeddings 1/1, Norms 18/65, MLP 90/96, Attention 120/128, lm_head 1/1

1 derivatives checked

• mradermacher/Experiment28M7_Multi_verse_modelM7-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse