Qwen/QVQ-72B-Preview
Full fine-tune vs Qwen/Qwen2-VL-72B. Fine-tune: 97% of the weights changed, across 80 of 80 layers.
Tensors changed: Embeddings 0/2, Norms 131/291, Attention 560/688, MLP 244/372, lm_head 1/1
7 derivatives checked
- bartowski/QVQ-72B-Preview-GGUF: Quantization
- mradermacher/QVQ-72B-Preview-i1-GGUF: Quantization
- mmnga/QVQ-72B-Preview-gguf: Quantization
- lmstudio-community/QVQ-72B-Preview-GGUF: Quantization
- tensorblock/QVQ-72B-Preview-GGUF: Quantization
- mradermacher/QVQ-72B-Preview-GGUF: Quantization
- DevQuasar-6/Qwen.QVQ-72B-Preview-GGUF: Quantization
1 other repos hold byte-for-byte identical weight files.