ddvd233/QoQ-Med-VL-32B

Full fine-tune vs Qwen/Qwen2.5-VL-32B-Instruct. Fine-tune: 98% of the weights changed, across 64 of 64 layers.

Tensors changed: Attention 576/576, MLP 367/388, Norms 133/194, Embeddings 1/2, lm_head 1/1

2 derivatives checked

• mradermacher/QoQ-Med-VL-32B-i1-GGUF: Quantization (fewer layers)
• mradermacher/QoQ-Med-VL-32B-GGUF: Quantization (fewer layers)

On the Hub · Downloads on Model Pulse