aloobun/Meta-Llama-3-7B-29Layers

Partial fine-tune vs meta-llama/Meta-Llama-3-8B. Partial change: 15% of the weights changed (attention, mlp, norms), in 5 of 29 layers.

Tensors changed: Embeddings 0/1, Norms 10/59, MLP 15/87, Attention 20/116, lm_head 0/1

1 derivatives checked

• mradermacher/Meta-Llama-3-7B-29Layers-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse