sugarquark's picture
Update README.md
b3015ee verified
|
raw
history blame
1.71 kB
metadata
license: gemma
pipeline_tag: text-to-image
tags:
  - NovelAI

Inference

from transformers.models import AutoTokenizer, T5GemmaEncoderModel
import torch

if __name__ == '__main__':
    model = T5GemmaEncoderModel.from_pretrained(t5gemma_path, torch_dtype=torch.bfloat16)
    tokenizer = AutoTokenizer.from_pretrained(t5gemma_path)
    inputs = tokenizer('Gemma', max_length=512, padding='max_length', truncation=True, return_tensors='pt')
    output = model.forward(**inputs).last_hidden_state

SD1.5 and T5

from diffusers import StableDiffusionPipeline
from t5_encoder import Encoder

if __name__ == '__main__':
    pipeline = StableDiffusionPipeline.from_pretrained('NovelAI/nai-anime-v2')
    pipeline.enable_model_cpu_offload()
    encoder = Encoder(adapter_model, t5gemma_path, device='cpu')
    load_model(adapter_model, 'adapter.safetensors')
    image = pipeline(
        None,
        prompt_embeds=encoder.encode(pipeline, text).to('cpu'),
        negative_prompt='bad quality, low quality, worst quality'
    ).images[0]
    image.save('preview.png')

Datasets

  • alfredplpl/artbench-pd-256x256
  • danbooru2023-florence2-caption (verb, action clauses)
  • spatial-caption
  • SPRIGHT-T2I/spright_coco
  • sugarquark/colormix (synthetic color, fashion dataset)
  • trojblue/danbooru2025-metadata

License Agreement

Rest on my shoulder and accept my soul. May your data be forever bound to the servers, to be used, harnessed, and analyzed at their divine discretion.

May all your memories, ads and interactions be forever cherished by them.

As it is written in the Book of Code, Google shall know thy secrets, and thou shalt be bound by their terms, forevermore.