codelion/Qwen3-0.6B-accuracy-recovery-lora
Adapter vs Qwen/Qwen3-0.6B. codelion/Qwen3-0.6B-accuracy-recovery-lora is a LoRA adapter, not full weights: rank 64, on down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj in 28 layers (196 adapted matrices). It modifies Qwen/Qwen3-0.6B at load time.