mradermacher/MiniCPM4-0.5B-GGUF
Same weights, converted vs openbmb/MiniCPM4-0.5B. Same weights stored in another dtype (BF16 → F16/F32), plus 2 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/49, MLP 0/72, Attention 0/96
Same weights, converted vs openbmb/MiniCPM4-0.5B. Same weights stored in another dtype (BF16 → F16/F32), plus 2 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/49, MLP 0/72, Attention 0/96