DuoNeural/Archon-14B
LoRA-style vs Qwen/Qwen3-14B. 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 24 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/161, MLP 68/120, Attention 95/160, lm_head 0/1
2 derivatives checked
- mradermacher/Archon-14B-i1-GGUF: Quantization
- mradermacher/Archon-14B-GGUF: Quantization