jaredpalmer/kev-0.8b
Adapter vs Qwen/Qwen3.5-0.8B-Base. jaredpalmer/kev-0.8b 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 24 layers (186 adapted matrices). It modifies Qwen/Qwen3.5-0.8B-Base at load time.
3 derivatives checked
- ggml-org/Kev-0.8B-GGUF: Adapter
- mys/kev-0.8b-GGUF: Adapter
- DreamBlooms/kev-0.8b-GGUF: Adapter