Sakalti/ultiima-32B
Full fine-tune vs Qwen/Qwen2.5-32B. Fine-tune: 100% of the weights changed, across 64 of 64 layers.
Tensors changed: Embeddings 1/1, Norms 128/129, MLP 192/192, Attention 448/448, lm_head 1/1
5 derivatives checked
- mradermacher/ultiima-32B-i1-GGUF: Quantization
- mradermacher/ultiima-32B-GGUF: Quantization
- suayptalha/Qwenhancement-32B-v1: Partial fine-tune
- Youlln/ECE-MIRAGE-3: Partial fine-tune
- Sakalti/ultiima-52B: Partial fine-tune (fewer layers)