mradermacher/GemmaX2-28-9B-v0.2-GGUF
Same weights, converted vs xiaomi-research/GemmaX2-28-9B-v0.2. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/169, MLP 0/126, Attention 0/168
Same weights, converted vs xiaomi-research/GemmaX2-28-9B-v0.2. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/169, MLP 0/126, Attention 0/168