mradermacher/Qwen-SEA-Guard-4B-2602-GGUF
Same weights, converted (fewer layers) vs aisingapore/Qwen-SEA-Guard-4B-2602. mradermacher/Qwen-SEA-Guard-4B-2602-GGUF matches part of aisingapore/Qwen-SEA-Guard-4B-2602 (398 of its 713 tensors; the rest of aisingapore/Qwen-SEA-Guard-4B-2602 has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/145, MLP 0/108, Attention 0/144