NOTICE

NOTICE
======

Qwen is licensed under the Qwen RESEARCH LICENSE AGREEMENT, Copyright (c) 2026
Hangzhou Tongyi Laboratory Technology Co., Ltd. All Rights Reserved.

This repository is a derivative work of Qwen-Image-2.1. The full agreement is in
`LICENSE`, a copy of it is given to every recipient of these files.

Modified files, as required by section 2.b of the agreement:

  transformer/config.json    added the `text_fusion_config` key, `_class_name` changed to
                             QwenImage21FusionTransformer2DModel
  transformer/*.safetensors  re-saved in fp16, plus the 66 `text_fusion.*` tensors of the
                             adapter from the image21-08b-text-encoder-adapter project (adapter_v11)
  vae/                       copied unchanged (fp32)
  scheduler/                 modified: dynamic shifting off, plain static `shift` 5.0, `shift_terminal`
                             removed (was 0.02) — the sdxs-micro schedule
  text_encoder/              different model: Qwen3.5-0.8B instead of Qwen3-VL-8B. Upstream files are
                             re-saved in fp16 (bf16 -> fp16, single `model.safetensors` instead of
                             the sharded bf16 file); `config.json` is unchanged. Modified under the
                             Apache License 2.0, section 4.b.
  tokenizer/, processor/     the Qwen3.5-0.8B tokenizer and processor, byte-identical to upstream

Files added by this repository: pipeline.py, transformer.py, example.py, model_index.json,
media/, README.md.

Improved using Qwen.
The Qwen3.5-0.8B text encoder is redistributed under the Apache License 2.0, see
`LICENSE-Qwen3.5-0.8B`.