Zichen1024/SAP-4B

Partial fine-tune vs Qwen/Qwen3-4B-Instruct-2507. Partial change: 90% of the weights changed (attention, mlp, norms), in 36 of 36 layers.

Tensors changed: Embeddings 0/1, Norms 55/145, MLP 107/108, Attention 144/144

2 derivatives checked

• mradermacher/SAP-4B-i1-GGUF: Quantization
• mradermacher/SAP-4B-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse