Finetuning - SushantGautam/kandi2-prior-medical-model

This pipeline was finetuned from kandinsky-community/kandinsky-2-2-prior on the waitwhoami/vqa_caption.dataset-full dataset. Below are some example images generated with the finetuned pipeline using the following prompts: ['The colonoscopy image contains a single, moderate-sized polyp that has not been removed, appearing in red and pink tones in the center and lower areas']:

Pipeline usage

You can use the pipeline like so:

from diffusers import DiffusionPipeline
import torch

pipe_prior = DiffusionPipeline.from_pretrained("SushantGautam/kandi2-prior-medical-model", torch_dtype=torch.float16)
pipe_t2i = DiffusionPipeline.from_pretrained("kandinsky-community/kandinsky-2-2-decoder", torch_dtype=torch.float16)
prompt = "The colonoscopy image contains a single, moderate-sized polyp that has not been removed, appearing in red and pink tones in the center and lower areas"
image_embeds, negative_image_embeds = pipe_prior(prompt, guidance_scale=1.0).to_tuple()
image = pipe_t2i(image_embeds=image_embeds, negative_image_embeds=negative_image_embeds).images[0]
image.save("my_image.png")

Training info

These are the key hyperparameters used during training:

• Epochs: 30
• Learning rate: 1e-05
• Batch size: 128
• Gradient accumulation steps: 1
• Image resolution: 768
• Mixed-precision: None

More information on all the CLI arguments and the environment are available on your wandb run page.