Delta-Vector/Rei-24B-KTO
LoRA-style vs Delta-Vector/Rei-24B-Base. 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/Rei-24B-KTO-i1-GGUF: Quantization
- DevQuasar/Delta-Vector.Rei-24B-KTO-GGUF: Same weights, converted
- mradermacher/Rei-24B-KTO-GGUF: Quantization
- OddTheGreat/Textolite_24B_V.2: Full fine-tune
- OddTheGreat/Circuitry_24B_V.2: LoRA-style
- OddTheGreat/Circuitry_24B_V.3: LoRA-style
- DoppelReflEx/MiniusLight-24B-v3: LoRA-style
- Sorihon/Celestial-Order-24B-V2.5: Full fine-tune
- Sorihon/CYDR-24B: Full fine-tune