agentscope-ai/QwenPaw-Flash-9B
Partial fine-tune vs Qwen/Qwen3.5-9B. Partial change: 84% of the weights changed (attention, lm_head, mlp, norms), in 32 of 32 layers.
Tensors changed: lm_head 1/1, Embeddings 0/4, MLP 96/208, Attention 187/332, Norms 102/215
8 derivatives checked
- bartowski/agentscope-ai_CoPaw-Flash-9B-GGUF: Same weights, converted
- mradermacher/QwenPaw-Flash-9B-i1-GGUF: Quantization
- mradermacher/CoPaw-Flash-9B-GGUF: Same weights, converted
- mradermacher/QwenPaw-Flash-9B-GGUF: Same weights, converted
- daydreamwarrior/QwenPaw-Flash-9B-heretic-coder3101-GGUF: Abliterated
- marksverdhei/CoPaw-Flash-9B-GGUF: Same weights, converted
- ykarout/Qwen3.5-9b-Opus-Openclaw-Distilled: LoRA-style
- coder3101/QwenPaw-Flash-9B-heretic: Abliterated
1 other repos hold byte-for-byte identical weight files.