metadata

base_model: KHongJae/full_train_TE_D_0_10000to50000
library_name: diffusers
license: creativeml-openrail-m
tags:
  - text-to-image
  - dreambooth
  - diffusers-training
  - stable-diffusion
  - stable-diffusion-diffusers
inference: true

KHongJae/Chatting_Based_Emoji_Generation_Model

T-Academy ASAC 6기 DL 프로젝트에서 사용한 이모티콘 생성 모델입니다.

기존 묘사형 프롬프트에서만 작동 하던 Stable Diffusion을 대화형 프롬프트에서 작동할 수 있도록 시도한 모델입니다.

DreamBooth for the text encoder was enabled: True.

Intended uses & limitations

How to use

pipeline = DiffusionPipeline.from_pretrained(
    "KHongJae/Chatting_Based_Emoji_Generation_Model",
    torch_dtype=torch.float16
).to("cuda")
pipeline.scheduler = UniPCMultistepScheduler.from_config(pipeline.scheduler.config)
prompt = "Create your own prompt"
negative_prompt = "Create your own negative prompt"

pipeline(
  prompt=prompt,
  negative_prompt=negative_prompt,
  width=200,
  height=200,
  num_inference_steps=50,
  num_images_per_prompt=1,
  generator=torch.manual_seed(123456789),
).images[0]

Limitations and bias

[TODO: provide examples of latent issues and potential remediations]

Training details

[TODO: describe the data used to train the model]