nightmedia/Qwen3.6-27B-Seven

Full fine-tune vs DavidAU/Qwen3.5-27B-Claude-4.6-OS-INSTRUCT. Fine-tune: 100% of the weights changed, across 64 of 64 layers.

Tensors changed: lm_head 1/1, Embeddings 4/4, Norms 319/319, Attention 556/556, MLP 304/304

4 derivatives checked

• mradermacher/Qwen3.6-27B-Seven-GGUF: Quantization
• mradermacher/Qwen3.6-27B-Seven-i1-GGUF: Quantization
• nightmedia/Qwen3.8-27B-MindMeld-qx64-hi-mlx: Quantization
• nightmedia/Qwen3.8-27B-USS-Collective-qx64-hi-mlx: Quantization

On the Hub · Downloads on Model Pulse