DiffusionPredictionType
Enums · AiDotNet.Enums
Defines what the diffusion model predicts during the denoising process.
For Beginners
When a model looks at a noisy image, it can be trained to predict:
- Epsilon (noise): “What noise was added to make this blurry?”
- Sample: “What did the clean image look like?”
- V-prediction: A mathematical blend of both (more stable training)
Most models use Epsilon prediction as it’s the most common and well-studied approach.
How It Works
Different diffusion models can be trained to predict different targets. The prediction type affects how the scheduler interprets the model output and computes the denoised sample.
Fields
| Field | Summary |
|---|---|
DiffusionPredictionType Epsilon |
Model predicts the noise (epsilon) that was added to the clean sample. |
DiffusionPredictionType Sample |
Model directly predicts the clean sample (x_0). |
DiffusionPredictionType VPrediction |
Model predicts v = sqrt(alpha_cumprod) * epsilon - sqrt(1-alpha_cumprod) * x_0. |
int value__ |