Darkknight535/Theia-Unslop-Mell-21B
Full fine-tune vs Darkknight535/Patricide-21B-Unslop-Mell. Fine-tune: 100% of the weights changed, across 70 of 70 layers.
Tensors changed: Norms 138/141, MLP 210/210, Attention 280/280
2 derivatives checked
- mradermacher/Theia-Unslop-Mell-21B-i1-GGUF: Quantization
- mradermacher/Theia-Unslop-Mell-21B-GGUF: Quantization