ConicCat/Mistral-Small-3.2-AntiRep-24B
LoRA-style (fewer layers) vs mistralai/Mistral-Small-3.2-24B-Instruct-2506. ConicCat/Mistral-Small-3.2-AntiRep-24B matches part of mistralai/Mistral-Small-3.2-24B-Instruct-2506 (363 of its 585 tensors; the rest of mistralai/Mistral-Small-3.2-24B-Instruct-2506 has no counterpart). On the shared part: Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 40 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/81, MLP 120/120, Attention 160/160, lm_head 0/1
9 derivatives checked
- mradermacher/Mistral-Small-3.2-AntiRep-24B-i1-GGUF: Quantization
- mradermacher/Mistral-Small-3.2-AntiRep-24B-GGUF: Quantization
- Naphula/Goetia-24B-v1.2: Full fine-tune
- DarkArtsForge/Asmodeus-24B-v1: Full fine-tune
- zerofata/MS3.2-PaintedFantasy-v2-24B: LoRA-style
- zerofata/MS3.2-PaintedFantasy-Visage-v3-34B: Full fine-tune (fewer layers)
- Vortex5/MS3.2-24B-Fiery-Lynx: LoRA-style
- Pranavz/MS3.2-PaintedFantasy-v2-24B-heretic: LoRA-style
- Naphula-Archives/MistralThinker-AntiRep-24B-v1: Full fine-tune