yamatazen/HMS-Fusion-12B
Partial fine-tune vs shisa-ai/shisa-v2-mistral-nemo-12b. Partial change: 78% of the weights changed (attention, embeddings, mlp, norms), in 40 of 40 layers.
Tensors changed: Embeddings 1/1, Norms 22/81, MLP 96/120, Attention 144/160, lm_head 0/1
2 derivatives checked
- mradermacher/HMS-Fusion-12B-i1-GGUF: Quantization
- mradermacher/HMS-Fusion-12B-GGUF: Quantization