rinna/nekomata-14b
Full fine-tune vs Qwen/Qwen-14B. Fine-tune: 100% of the weights changed, across 40 of 40 layers.
Tensors changed: Attention 120/120, Norms 81/81, MLP 120/120, Embeddings 1/1, lm_head 1/1
3 derivatives checked
- mradermacher/nekomata-14b-i1-GGUF: Quantization
- mradermacher/nekomata-14b-GGUF: Quantization
- rinna/nekomata-14b-instruction: Full fine-tune