YOYO-AI/Qwen3-30B-A3B-YOYO-AutoThink

Full fine-tune vs Qwen/Qwen3-30B-A3B-Instruct-2507. Fine-tune: 95% of the weights changed, across 48 of 48 layers.

Tensors changed: Norms 165/193, MLP 2603/2613, Attention 143/192

2 derivatives checked

• mradermacher/Qwen3-30B-A3B-YOYO-AutoThink-GGUF: Quantization
• mradermacher/Qwen3-30B-A3B-YOYO-AutoThink-i1-GGUF: Quantization

On the Hub · Downloads on Model Pulse