Delta-Vector/Archaeo-32B-KTO
LoRA-style vs Delta-Vector/Archaeo-32B. 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 64 of 64 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/129, MLP 192/192, Attention 256/448
5 derivatives checked
- mradermacher/Archaeo-32B-KTO-i1-GGUF: Quantization
- mradermacher/Archaeo-32B-KTO-GGUF: Quantization
- mradermacher/Axo-Merge-Archaeo-V2-Lora-GGUF: Quantization
- Phr00t/Phr00tyMix-v4-32B: Full fine-tune
- Phr00t/Phr00tyMix-v3-32B: Full fine-tune