Nexesenex/Llama_3.x_70b_SmarTrident_v1.02
Full fine-tune vs Nexesenex/Llama_3.x_70b_L3.3_Dolphin_128K_v1.02. Fine-tune: 100% of the weights changed, across 80 of 80 layers.
Tensors changed: Norms 160/161, MLP 240/240, Attention 320/320
2 derivatives checked
- mradermacher/Llama_3.x_70b_SmarTrident_v1.02-i1-GGUF: Quantization
- mradermacher/Llama_3.x_70b_SmarTrident_v1.02-GGUF: Quantization