IPADS-SAI/MobiMind-Mixed-4B-1031

Full fine-tune vs Qwen/Qwen3-VL-4B-Instruct. Fine-tune: 93% of the weights changed, across 36 of 36 layers.

Tensors changed: Embeddings 2/4, Norms 115/249, MLP 120/220, Attention 144/240

2 derivatives checked

• mradermacher/MobiMind-Mixed-4B-1031-i1-GGUF: Quantization (fewer layers)
• mradermacher/MobiMind-Mixed-4B-1031-GGUF: Same weights, converted (fewer layers)

On the Hub · Downloads on Model Pulse