凉屋/涼香/Ryohka/Suzuya

Model Overview

This model was trained using the training environment and computing resources provided by the Model Training section of the ModelScope AIGC platform

• Model Type: LoRA

• Base Model: openmoss/MOSS-Image-SFT

• Training Framework: DiffSynth-Studio

• Training Dataset Size: 100

• Total Training Steps: 9,000

• Open-Source License: Apache-2.0

Inference Code

Install DiffSynth-Studio

pip install diffsynth

Run inference

from diffsynth.pipelines.moss_image import MOSSImagePipeline, ModelConfigimport torchpipe = MOSSImagePipeline.from_pretrained(    torch_dtype=torch.bfloat16,    device="cuda",    model_configs=[        ModelConfig(            model_id="openmoss/MOSS-Image-SFT",            origin_file_pattern="transformer/diffusion_pytorch_model*.safetensors"        ),        ModelConfig(            model_id="openmoss/MOSS-Image-SFT",            origin_file_pattern="text_encoder/model*.safetensors"        ),        ModelConfig(            model_id="openmoss/MOSS-Image-SFT",            origin_file_pattern="vae/diffusion_pytorch_model.safetensors",            computation_dtype=torch.float32        ),    ],    tokenizer_config=ModelConfig(        model_id="openmoss/MOSS-Image-SFT",        origin_file_pattern="tokenizer/"    ),)pipe.load_lora(    pipe.dit,    ModelConfig(        model_id="synthesia/Ryohka_Suzuya",        origin_file_pattern="Ryohka_Suzuya_c1-st9000.safetensors"    ))prompt = "a cat"image = pipe(    prompt,    height=1024,    width=1024,    seed=0,    num_inference_steps=50,    cfg_scale=3.0)image.save("image.jpg")