CarrotAI/Rabbit-Ko-15B-Instruct
Full fine-tune (fewer layers) vs Qwen/Qwen2-7B-Instruct. Qwen/Qwen2-7B-Instruct matches part of CarrotAI/Rabbit-Ko-15B-Instruct (339 of its 723 tensors; the rest of CarrotAI/Rabbit-Ko-15B-Instruct has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 54/57, MLP 84/84, Attention 196/196, lm_head 1/1
2 derivatives checked
- mradermacher/Rabbit-Ko-15B-Instruct-i1-GGUF: Quantization
- mradermacher/Rabbit-Ko-15B-Instruct-GGUF: Quantization