Naphula/Cthulhu-70B-v1

Full fine-tune vs CrucibleLab/L3.3-70B-Loki-V2.0. Fine-tune: 100% of the weights changed, across 80 of 80 layers.

Tensors changed: Norms 161/161, MLP 240/240, Attention 320/320

2 derivatives checked

• mradermacher/Cthulhu-70B-v1-i1-GGUF: Quantization
• mradermacher/Cthulhu-70B-v1-GGUF: Quantization

On the Hub · Downloads on Model Pulse