Qwen3.5-9B-Qworus-V2
What is this? This is a 50/50 DARE-TIES merge of empero-ai/Qwen3.8-9B-Distill and ornith-ai/Ornith-1.5-9B.
The result is a hybrid reasoning model for coding, tool calls, planning, designing, and answering questions.
GGUF Quantizations are also provided with an imatrix trained on a mixture of code, business, math and wikitext. These come in either Q4_K_L for those with limited resources or Q6_K_L for those who want high precision. Both quant levels have the token embed and output layers bumped up to Q8_0 and ssm_* tensors bumped up to BF16/Q8_0 for higher accuracy on longer contexts.
Unlike the previous Qworus model, the 2 merge models were not uncensored, so this model is not uncensored by default.