jagilley/stable-diffusion-upscaler 🔢🖼️📝❓ → 🖼️

▶️ 3.6K runs 📅 Jan 2023 ⚙️ Cog 0.6.1 🔗 GitHub
image-editing image-upscaling

About

Upscale images with Stable Diffusion

Example Output

Output

Example output

Performance Metrics

18.11s Prediction Time
18.17s Total Time
All Input Parameters 
{
  "eta": 1,
  "image": "https://replicate.delivery/pbxt/IHlAm5lcJQkjWUh0cHS6L7SZff9ODH0TyFvplbFg8VOcOwmR/out-2.png",
  "scale": "4",
  "steps": 10,
  "decoder": "finetuned_840k",
  "sampler": "k_dpm_adaptive",
  "tol_scale": 0.25,
  "batch_size": 1,
  "num_samples": 1,
  "guidance_scale": 1,
  "noise_aug_type": "gaussian"
}
Input Parameters
eta Type: numberDefault: 1
ETA
image (required) Type: string
Image to be upscaled
scale Type: numberDefault: 1.5Range: 0 - 10
Factor to scale image by
steps Type: integerDefault: 10
Number of steps to take in the diffusion process
prompt Type: stringDefault:
Prompt. Not strictly required but can subtly affect the upscaling result.
decoder Default: finetuned_840k
Decoder to use
sampler Default: k_dpm_adaptive
Sampler to use
tol_scale Type: numberDefault: 0.25
Tolerance scale
batch_size Type: integerDefault: 1
Batch size
num_samples Type: integerDefault: 1
Number of samples to generate
guidance_scale Type: numberDefault: 1Range: 0 - 10
Scale factor for guidance
noise_aug_type Default: gaussian
Noise augmentation type
noise_aug_level Type: numberDefault: 0Range: 0 - 0.6
Noise augmentation level
Output Schema

Output

Type: string • Format: uri

Example Execution Logs 
No seed was provided, using the current time.
Generating with seed=1675968995
Global seed set to 1675968995
Version Details
Version ID
4d0aeee7387b1170b0f6f42bdf14c7d9d8e00a15430d95446ad7426dc61fc3d8
Version Created
January 10, 2023
Run on Replicate →