Aratako/Qwen3-8B-NSFW-JP
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 145/145, MLP 108/108, Attention 144/144, lm_head 1/1
4 derivatives checked
- DevQuasar/Aratako.Qwen3-8B-NSFW-JP-GGUF: Same weights, converted
- mradermacher/Qwen3-8B-NSFW-JP-i1-GGUF: Quantization
- mradermacher/Qwen3-8B-NSFW-JP-GGUF: Same weights, converted
- Aratako/Qwen3-8B-ERP-v0.1: Full fine-tune