grimjim/kukulemon-v3-soul_mix-32k-7B
Partial fine-tune vs grimjim/kukulemon-32K-7B. Partial change: 62% of the weights changed (attention, embeddings, mlp, norms), in 32 of 32 layers.
Tensors changed: Embeddings 1/1, Norms 2/65, MLP 59/96, Attention 82/128, lm_head 0/1
1 derivatives checked
- grimjim/kukulemon-v3-soul_mix-32k-7B-GGUF: Quantization