hasanmdsadadul/PaddleOCR-VL-1.6-Bangla
LoRA-style vs PaddlePaddle/PaddleOCR-VL-1.6. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 18 of 27 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, MLP 54/170, Norms 0/151, Embeddings 0/5, Attention 72/292, other 0/1