simonlqy/SkillGate-9B

Partial fine-tune vs Qwen/Qwen3.5-9B. Partial change: 74% of the weights changed (attention, embeddings, mlp, norms, other), in 32 of 32 layers.

Tensors changed: lm_head 0/1, Embeddings 1/5, MLP 99/211, Attention 182/336, other 1/1, Norms 67/221

2 derivatives checked

• mradermacher/SkillGate-9B-GGUF: Same weights, converted
• mradermacher/SkillGate-9B-i1-GGUF: Quantization

On the Hub · Downloads on Model Pulse