jaredpalmer/kev-4b
Adapter vs Qwen/Qwen3.5-4B-Base. jaredpalmer/kev-4b is a LoRA adapter, not full weights: rank 16, on down_proj, gate_proj, in_proj_a, in_proj_b, in_proj_qkv, in_proj_z, k_proj, o_proj… in 32 layers (248 adapted matrices). It modifies Qwen/Qwen3.5-4B-Base at load time.
7 derivatives checked
- ggml-org/Kev-4B-GGUF: Adapter
- mys/kev-4b-GGUF: Adapter
- RoderickQiu/kev-4b-mlx-8bit: Adapter
- DreamBlooms/kev-4b-GGUF: Adapter
- aselea/Kev-4B-MLX-Serve-8bit: Adapter
- Jakevin/kev-4b-ternary-mlx: Adapter
- alanhuangya/OpenJev-4B: Adapter