Gryphe/Tiamat-24B-Magistral
Full fine-tune vs mistralai/Magistral-Small-2509. Fine-tune: 98% of the weights changed, across 40 of 40 layers.
Tensors changed: Embeddings 1/1, Norms 0/131, MLP 120/192, Attention 160/256, other 0/4, lm_head 1/1
5 derivatives checked
- bartowski/Gryphe_Tiamat-24B-Magistral-GGUF: Same weights, converted (fewer layers)
- mradermacher/Tiamat-24B-Magistral-i1-GGUF: Quantization (fewer layers)
- mradermacher/Tiamat-24B-Magistral-GGUF: Quantization (fewer layers)
- MuXodious/Tiamat-24B-Magistral-PaperWitch-heresy: Abliterated
- DarkArtsForge/Magistaroth-24B-v1: Full fine-tune (fewer layers)