Qwen3 VL 32B Instruct NVFP4 Model Overview Model Architecture: Qwen3VLForConditionalGeneration Input: Text, Image Output: Text Model Optimizations: Weight quantization: FP4 Activation quantization: FP4 Release Date: Version: 1.0 Model Developers: : Red Hat Quantized version of Qwen/Qwen3 VL 32B Instruct. Model Optimizations This model was obtained by quantizing the weights and activations of Qwen/Qwen3 VL 32B Instruct to FP8 data type. This optimization reduces the number of bits per parameter from 16 to 4, reducing the disk size and GPU memory requirements by approximately 75%. Only the weights and activations of the linear operators within transformers blocks of the language model are quantized. Deployment Use with vLLM 1. Initialize vLLM server: 2. Send requests to the server: Creation This model was quantized using the llm compressor library as shown below. Creation details Evaluation The model was evaluated on the OpenLLMv1 leaderboard task, using lm evaluation harness. vLLM was used for all evaluations. Evaluation details ChartQA MMLU Accuracy Comparison ChartQA Results Model Accuracy Recovery (%) Qwen/Qwen3 VL 32B Instruct 61.52 100.00 Qwen/Qwen3 VL 32B Instruct FP8 86.92 14…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy