mradermacher/Apertus-8B-2509-GGUF
Same weights, converted vs swiss-ai/Apertus-8B-2509. Same weights stored in another dtype (BF16 → F16/F32), plus 1 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/129, MLP 0/64, Attention 0/128, lm_head 0/1
Same weights, converted vs swiss-ai/Apertus-8B-2509. Same weights stored in another dtype (BF16 → F16/F32), plus 1 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/129, MLP 0/64, Attention 0/128, lm_head 0/1