| Qwen is licensed under the Qwen RESEARCH LICENSE AGREEMENT, Copyright (c) 2026 Hangzhou Tongyi Laboratory Technology Co., Ltd. All Rights Reserved. | |
| Modified by mlx-serve (https://github.com/ddalcu/mlx-serve): the transformer/ and text_encoder/ | |
| weights are re-quantized to 8-bit affine (group 64) by tests/convert_qwen_image21_weights.py, | |
| and lm_head and the VAE's time_conv weights are dropped. Every other file is copied unchanged. | |