samuelcardillo/Carnice-Qwen3.6-MoE-35B-A3B
LoRA-style vs Qwen/Qwen3.6-35B-A3B. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of k_proj, o_proj, q_proj, v_proj changed, in 10 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/5, Attention 40/392, MLP 0/399, Norms 0/247, lm_head 0/1, other 0/1
4 derivatives checked
- mudler/Carnice-Qwen3.6-MoE-35B-A3B-APEX-GGUF: Quantization
- mudler/Carnice-Qwen3.6-MoE-35B-A3B-APEX-MTP-GGUF: Same weights, converted
- mradermacher/Carnice-Qwen3.6-MoE-35B-A3B-i1-GGUF: Quantization
- mradermacher/Carnice-Qwen3.6-MoE-35B-A3B-GGUF: Quantization