mergekit-community/Ar_Mage_Deeplama_R31
Full fine-tune vs Omartificial-Intelligence-Space/Arabic-llama3.1-16bit-FT. Fine-tune: 97% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 1/1, Norms 65/65, MLP 93/96, Attention 124/128, lm_head 1/1
1 derivatives checked
- mradermacher/Ar_Mage_Deeplama_R31-GGUF: Same weights, converted