ASHu2/goblinV1
LoRA-style vs openbmb/MiniCPM-V-4.6. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, in_proj_a, in_proj_b, in_proj_qkv, in_proj_z, k_proj, o_proj, out_proj, q_proj, up_proj, v_proj changed, in 24 of 27 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/4, Norms 0/195, Attention 114/392, MLP 72/184, other 0/4