| --- |
| base_model: |
| - OpenMOSS-Team/MOSS-Image-Base |
| base_model_relation: quantized |
| license: apache-2.0 |
| language: |
| - en |
| - zh |
| pipeline_tag: text-to-image |
| tags: |
| - unsloth |
| - fp8 |
| - int8 |
| - quantized |
| - moss-image |
| - text-to-image |
| - diffusion-single-file |
| widget: |
| - text: 'a photoreal close-up portrait of an older fisherman, weathered skin, sharp catchlights in the eyes, natural window light, 85mm lens (INT8-ConvRot)' |
| output: |
| url: samples/int8_convrot_portrait.png |
| - text: 'a red vintage pickup truck parked beside a wheat field, clear blue sky, film photography look (INT8-ConvRot)' |
| output: |
| url: samples/int8_convrot_pickup_truck.png |
| --- |
| |
| <div> |
| <div style="display: flex; gap: 5px; align-items: center; "> |
| <a href="https://github.com/unslothai/unsloth/"> |
| <img src="https://github.com/unslothai/unsloth/raw/main/images/unsloth%20new%20logo.png" width="133"> |
| </a> |
| <a href="https://discord.gg/unsloth"> |
| <img src="https://github.com/unslothai/unsloth/raw/main/images/Discord%20button.png" width="173"> |
| </a> |
| <a href="https://docs.unsloth.ai/"> |
| <img src="https://raw.githubusercontent.com/unslothai/unsloth/refs/heads/main/images/documentation%20green%20button.png" width="143"> |
| </a> |
| </div> |
| </div> |
| |
| This is an FP8 / INT8 quantized version of [MOSS-Image-Base](https://huggingface.co/OpenMOSS-Team/MOSS-Image-Base) by OpenMOSS, the 5.1B text-to-image transformer (Qwen3.5-9B text encoder, FLUX.2 small-decoder VAE). <br> |
| - The transformer is 5.42 GB instead of 10.16 GB in bf16. |
| - Loads pre-quantized in [Unsloth Studio](https://github.com/unslothai/unsloth/), which skips the bf16 download and the on-the-fly quantization. |
| - Text encoder: MOSS-Image-Base and MOSS-Image-SFT ship byte-identical Qwen3.5-9B text encoders, so the INT8-ConvRot encoder lives once, in [unsloth/MOSS-Image-SFT-FP8](https://huggingface.co/unsloth/MOSS-Image-SFT-FP8) (`MOSS-Image-SFT-text_encoder-INT8-ConvRot.safetensors`, 11.49 GB, LPIPS 0.059 vs the bf16 encoder). Unsloth Studio fetches it from there for Base too. |
| - This repo also carries the upstream diffusers files from [OpenMOSS-Team/MOSS-Image-Base](https://huggingface.co/OpenMOSS-Team/MOSS-Image-Base) (commit `85975793`), unchanged (`model_index.json`, `pipeline_moss_image.py`, `scheduler/`, `text_encoder/`, `tokenizer/`, `transformer/`, `vae/`), so Unsloth Studio downloads every component, the bf16 transformer and text encoder included, from here. |
|
|
| **Recommended:** `MOSS-Image-Base-INT8-ConvRot.safetensors`, Unsloth Studio's default (LPIPS 0.096). On Base, plain INT8 scores the same on average (0.095) with a lower worst case (0.184 against 0.403, one prompt whose composition changed). FP8 (0.163) is for explicit picks only. |
|
|
| | bf16 | INT8-ConvRot | |
| |---|---| |
| | <img src="samples/bf16_portrait.png" width="300"> | <img src="samples/int8_convrot_portrait.png" width="300"> | |
| | <img src="samples/bf16_pickup_truck.png" width="300"> | <img src="samples/int8_convrot_pickup_truck.png" width="300"> | |
|
|
| ## Usage in Unsloth Studio |
|
|
| Pick **MOSS-Image-Base** in the Images tab. Unsloth Studio keeps the transformer in bf16 while it fits on your GPU and otherwise loads `MOSS-Image-Base-INT8-ConvRot.safetensors` from this repo; the text encoder loads the INT8-ConvRot file by default. INT8 and FP8 can be chosen explicitly in the precision setting. Defaults follow the reference workflow: 1024x1024, 50 steps, CFG 3, shift 5, Euler. MOSS-Image needs transformers 5.3 or newer, and float16 is not supported (bf16 or fp32 only). |
|
|
| ## Files |
|
|
| | File | Size | Scheme | LPIPS (alex) vs bf16, mean / max | SSIM vs bf16 | |
| |---|---|---|---|---| |
| | `MOSS-Image-Base-INT8-ConvRot.safetensors` | 5.42 GB | INT8 per-row weight with ConvRot rotation (group 256) on the 252 block Linears, dynamic per-token activation; text embedder, timestep MLP and AdaLN bf16. Unsloth Studio's default | 0.096 / 0.403 | 0.897 | |
| | `MOSS-Image-Base-INT8.safetensors` | 5.42 GB | INT8 per-row weight, dynamic per-token activation, same bf16 layers | 0.095 / 0.184 | 0.905 | |
| | `MOSS-Image-Base-FP8.safetensors` | 5.41 GB | FP8 e4m3, per-row weight scale, dynamic per-token activation, same bf16 layers | 0.163 / 0.415 | 0.835 | |
|
|
| LPIPS (alex, lower is better) and SSIM against the bf16 transformer: same prompt, seed, initial noise and bf16 text embeddings, 1024x1024, 50 steps, CFG 3, over 12 prompts (8 general prompts plus 4 held-out ones: a close-up portrait, a storefront sign with text, hands folding origami, a long scene description). A repeated bf16 render scores 0. MOSS-Image is sensitive to tiny numeric changes: changing a single bf16 rounding step in one layer already moves the renders by 0.053 to 0.063 LPIPS on average, so differences near that level are composition shifts rather than quality loss. |
|
|
| GGUF builds for CPU, Mac and smaller GPUs: [unsloth/MOSS-Image-Base-GGUF](https://huggingface.co/unsloth/MOSS-Image-Base-GGUF). |
|
|
| --- |
|
|
| # MOSS-Image-Base |
|
|
|  |
|
|
| MOSS-Image-Base is the base text-to-image model in the **MOSS-Image** family |
| from OpenMOSS. It supports natural-language prompts, tag sequences, and JSON |
| layouts for controlling text placement and object arrangement. |
|
|
| ## Base vs. SFT |
|
|
| In this example, SFT produces clearer text grouping and a more organized whiteboard layout. |
|
|
| | MOSS-Image-Base | MOSS-Image-SFT | |
| | --- | --- | |
| |  |  | |
|
|
| ## Quick start |
|
|
| Use Python 3.10 or later and an NVIDIA GPU with CUDA support. Install a PyTorch |
| 2.9 or later build compatible with your CUDA environment, then install: |
|
|
| ```bash |
| pip install "diffusers>=0.38" "transformers>=5.3" "accelerate>=1.10" "safetensors>=0.6" "Pillow>=10" |
| ``` |
|
|
| ```python |
| import torch |
| from diffusers import DiffusionPipeline |
| |
| pipe = DiffusionPipeline.from_pretrained( |
| "OpenMOSS-Team/MOSS-Image-Base", |
| custom_pipeline="pipeline_moss_image", |
| trust_remote_code=True, |
| torch_dtype={"default": torch.bfloat16, "vae": torch.float32}, |
| ) |
| pipe.scheduler = type(pipe.scheduler).from_config(pipe.scheduler.config) |
| pipe.to("cuda") |
| |
| prompt = '''A wide view of a refrigerator beside a kitchen counter with a fruit bowl. A small sticky note on the refrigerator door is handwritten in Chinese: |
| "记得买牛奶和鸡蛋！ |
| 牛奶两盒，鸡蛋一打。 |
| 顺便带一袋面包。"''' |
| |
| image = pipe( |
| prompt, |
| height=1024, |
| width=1024, |
| num_inference_steps=50, |
| guidance_scale=3.0, |
| uncond_skip_layers=[22], |
| generator=torch.Generator("cpu").manual_seed(42), |
| ).images[0] |
| image.save("refrigerator_note.png") |
| ``` |
|
|
| The model and custom pipeline code are downloaded and cached automatically. You |
| can also pass a complete local model directory to `from_pretrained`. |
| The transformer and text encoder use bfloat16; the VAE and initial noise use |
| float32. The bundled VAE uses the FLUX.2 small decoder. |
|
|
| For the Gradio interface, CPU offload, and compilation options, follow the |
| [project quick start](https://github.com/OpenMOSS/MOSS-Image#quick-start). |
| For ComfyUI, search for **MOSS-Image** in **ComfyUI Manager** to install the |
| native plugin. Integration into the official ComfyUI repository is in progress. |
| See the [installation and usage guide](https://github.com/OpenMOSS/MOSS-Image/blob/main/comfyui/README.md) |
| and [example workflow](https://github.com/OpenMOSS/MOSS-Image/blob/main/comfyui/moss-image/example_workflows/moss-image.json). |
|
|
| ## Sampling settings |
|
|
| | Setting | Default | |
| | --- | --- | |
| | Sampler | Flow-matching Euler | |
| | Steps | 50 | |
| | CFG | 3 | |
| | Shift | 5 | |
| | Layer skipping | Negative branch, layer index 22 (zero-based) | |
| | Negative prompt | Empty | |
|
|
| The default guidance uses two transformer forwards per step. Pass |
| `uncond_skip_layers=[]` to disable negative-branch skipping, or set |
| `negative_prompt` to provide a negative prompt. Python API image dimensions must |
| be positive multiples of 16. |
|
|
| See the [user guide](https://github.com/OpenMOSS/MOSS-Image/blob/main/USERGUIDE.md) |
| for resolution, sampling, prompt, and LoRA training recommendations. |
|
|
| ## Layout-to-image generation |
|
|
| Pass a JSON layout as the prompt to describe the scene, rendered text, objects, |
| and their positions. Layouts use `summary.caption` and `items`, with |
| `bbox_format: "yxyx"` and `bbox_coord_space: "abs_centralized"`. |
|
|
| With the Python API, serialize the layout using |
| `json.dumps(layout, ensure_ascii=False)` and pass it to `pipe`. In Gradio, paste |
| the JSON into the Prompt field and set the matching image dimensions. |
| See the [layout format and example](https://github.com/OpenMOSS/MOSS-Image/blob/main/USERGUIDE.md#layout-prompts-experimental). |
|
|
| Layouts can also support controlled editing and user-friendly layout tools. |
| Edit text, element descriptions, and bounding boxes to adjust image content and |
| composition in a structured way. |
|
|
|  |
|
|
| ## Showcase |
|
|
|  |
|
|
|  |
|
|
| ## Citation |
|
|
| ```bibtex |
| @misc{ai2026mossimagetechnicalreport, |
| title={MOSS-Image Technical Report}, |
| author={MOSI. AI}, |
| year={2026}, |
| eprint={2610.23374}, |
| archivePrefix={arXiv}, |
| primaryClass={cs.CV}, |
| url={https://arxiv.org/abs/2610.23374}, |
| } |
| ``` |
|
|
| ## License and acknowledgements |
|
|
| The model weights and bundled inference code are released under the |
| [Apache License, Version 2.0](LICENSE.md). |
|
|
| We thank the Diffusers, Qwen, and FLUX teams. |
|
|
