mradermacher/Apertus-v1.1-4B-GGUF
Same weights, converted vs swiss-ai/Apertus-v1.1-4B. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/97, MLP 0/48, Attention 0/96, lm_head 0/1
Same weights, converted vs swiss-ai/Apertus-v1.1-4B. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/97, MLP 0/48, Attention 0/96, lm_head 0/1