usermma/Holo-3.1-0.8B-abliterated-rdo
Partial fine-tune (fewer layers) vs Hcompany/Holo-3.1-0.8B. usermma/Holo-3.1-0.8B-abliterated-rdo matches part of Hcompany/Holo-3.1-0.8B (320 of its 474 tensors; the rest of Hcompany/Holo-3.1-0.8B has no counterpart). On the shared part: Partial change: 44% of the weights changed (attention, embeddings, mlp), in 4 of 24 layers.
Tensors changed: Embeddings 1/1, Norms 0/79, Attention 10/168, MLP 12/72
2 derivatives checked
- mradermacher/Holo-3.1-0.8B-abliterated-rdo-i1-GGUF: Quantization
- mradermacher/Holo-3.1-0.8B-abliterated-rdo-GGUF: Same weights, converted