kai-os/Carnice-9b
LoRA-style (fewer layers) vs Qwen/Qwen3.5-9B. kai-os/Carnice-9b matches part of Qwen/Qwen3.5-9B (427 of its 775 tensors; the rest of Qwen/Qwen3.5-9B 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 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, MLP 96/96, Attention 32/224, Norms 0/105
5 derivatives checked
- kai-os/Carnice-9b-GGUF: Quantization
- ermiaazarkhalili/Carnice-9B-SFT-Fable5-Glint-GGUF: Full fine-tune
- ermiaazarkhalili/Carnice-9B-Function-Calling-xLAM-Unsloth: LoRA-style
- tripplet-research/suzhou3.1: LoRA-style
- ermiaazarkhalili/Carnice-9B-SFT-Claude-Opus-Reasoning-Unsloth: LoRA-style