HA-Siala/Mamba-v0.2
Adapter vs mistralai/Mistral-7B-v0.3. HA-Siala/Mamba-v0.2 is a LoRA adapter, not full weights: rank 16, on down_proj, gate_proj, k_proj, lm_head, o_proj, q_proj, up_proj, v_proj in 32 layers (225 adapted matrices). It modifies mistralai/Mistral-7B-v0.3 at load time.