26B-Suite/Goetia-26B-A4B-v1.4-LazyLora-heresy
Full fine-tune vs Naphula/Goetia-26B-A4B-v1.4. Fine-tune: 97% of the weights changed, across 30 of 30 layers.
Tensors changed: MLP 150/150, Norms 0/271, other 56/120, Attention 115/115, Embeddings 0/1
7 derivatives checked
- mradermacher/Goetia-26B-A4B-v1.4-LazyLora-heresy-i1-GGUF: Quantization
- mradermacher/Goetia-26B-A4B-v1.4-SOMPOA-heresy-i1-GGUF: Quantization
- mradermacher/Goetia-26B-A4B-v1.4-LazyLora-heresy-GGUF: Quantization
- alexokita/Goetia-26B-A4B-v1.4-LazyLora-heresy-i1-GGUF: Quantization
- mradermacher/Goetia-26B-A4B-v1.4-SOMPOA-heresy-GGUF: Quantization
- Phil2Sat/Goetia-26B-A4B-v1.4-LazyLora-heresy-I-Compact-MXFP4-GGUF: Quantization
- alexokita/Goetia-26B-A4B-v1.4-LazyLora-heresy-Heretic: Abliterated (fewer layers)