sherif1313/Arabic-GLM-OCR-v2
LoRA-style vs zai-org/GLM-OCR. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, k_proj, o_proj, proj, q_proj, qkv, v_proj changed, in 24 of 24 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/3, Norms 0/164, MLP 16/179, Attention 112/160, other 0/3
1 other repos hold byte-for-byte identical weight files.