Qwen/Qwen3Guard-Gen-8B
Full fine-tune vs Qwen/Qwen3-8B. Fine-tune: 92% of the weights changed, across 36 of 36 layers.
Tensors changed: Embeddings 0/1, Norms 122/145, MLP 108/108, Attention 144/144, lm_head 1/1
11 derivatives checked
- mradermacher/Qwen3Guard-Gen-8B-GGUF: Same weights, converted
- Yunhao-Feng/HazardAuditor: Full fine-tune
- zenlm/zen-guard-gen-8b: Identical copy
- ZJU-Safety/DARWIN-Guard: Partial fine-tune
- DevQuasar/Qwen.Qwen3Guard-Gen-8B-GGUF: Same weights, converted
- ShahzebKhoso/Qwen3Guard-Gen-8B-GGUF: Same weights, converted
- Yunhao-Feng/AdaGuard-8B: Full fine-tune
- VISHNUDHAT/Qwen3Guard-Gen-8B-Q4_K_M-GGUF: Quantization
- geoffmunn/Qwen3Guard-Gen-8B-GGUF: Quantization
- Rakancorle1/FoodGuard_qwenguard_4k: Full fine-tune
- zenlm/zen-guard-gen: Identical copy
4 other repos hold byte-for-byte identical weight files.