DavidAU/ERNIE-21B-A3B-Thinking-Gemini-3-Pro-High-Reasoning-V2
LoRA-style vs baidu/ERNIE-4.5-21B-A3B-Thinking. 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 28 of 28 layers; norms, biases and embeddings are untouched.
Tensors changed: MLP 2776/2830, Embeddings 0/1, Norms 0/57, Attention 112/112