mradermacher/Qwen3.5-9B-OPSA-GGUF
Same weights, converted vs Tuwhy/Qwen3.5-9B-OPSA. Same weights stored in another dtype (BF16 → F16/F32), plus 39 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/81, MLP 0/96, Attention 0/32, lm_head 0/1
Same weights, converted vs Tuwhy/Qwen3.5-9B-OPSA. Same weights stored in another dtype (BF16 → F16/F32), plus 39 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/81, MLP 0/96, Attention 0/32, lm_head 0/1