zenlm/zen-vl-4b-instruct

Full fine-tune vs Qwen/Qwen3-VL-4B-Instruct. Fine-tune: 91% of the weights changed, across 36 of 36 layers.

Tensors changed: Embeddings 1/4, Norms 89/249, MLP 108/220, Attention 144/240

2 derivatives checked

• mradermacher/zen-vl-4b-instruct-i1-GGUF: Quantization (fewer layers)
• mradermacher/zen-vl-4b-instruct-GGUF: Same weights, converted (fewer layers)

On the Hub · Downloads on Model Pulse