Quickstart Guide¶

Note: For more advanced configurations, see the tutorial and options reference.

Model Quickstart Guides¶

Feature Compatibility¶

The complete compatibility matrix is split by feature area so each table stays readable.

✓ = supported, ✓ = supported but usually requires DeepSpeed/FSDP2 for full-rank training, ✗ = not supported. Ref Inputs marks existing reference/edit/I2V conditioning paths; opt means optional and req means the edit/I2V flavour requires it.* TwinFlow support is native when twinflow_enabled=true; diffusion models still require diff2flow_enabled=true plus twinflow_allow_diff2flow=true. Self-Flow refers to CREPA self-flow support. Internal Guidance covers diffusion transformers with an auxiliary denoising head. LayerSync marks backbones that expose hidden states for alignment. NextLat uses captured hidden states for the auxiliary next-latent objective. XM marks Explorative Modeling support: noise-candidate selection for diffusion models and route selection for autoregressive token models.

Fast paths: Z-Image Turbo & Flux Schnell¶

• Z-Image Turbo: Fully supported LoRA with TREAD; runs fast on NVIDIA and macOS even without quant (int8 works too). Often the bottleneck is just trainer setup.
• Flux Schnell: The quickstart config handles the fast noise schedule and assistant LoRA stack automatically; no extra flags needed to train Schnell LoRAs.

Advanced Experimental Features¶

• Diff2Flow: Allows training standard epsilon/v-prediction models (SD1.5, SDXL, DeepFloyd, etc.) using a Flow Matching loss objective. This bridges the gap between older architectures and modern flow-based training.
• Scheduled Sampling: Reduces exposure bias by letting the model generate its own intermediate noisy latents during training ("rollout"). This helps the model learn to recover from its own generation errors.

Common Issues¶

Dataset has fewer samples than expected¶

If your dataset ends up with fewer usable samples than you expected, files may have been filtered during processing. Common reasons include:

• Files too small: Images below minimum_image_size are filtered out
• Aspect ratio out of range: Images outside minimum_aspect_ratio/maximum_aspect_ratio bounds are excluded
• Duration limits: Audio/video files exceeding duration limits are skipped

Viewing filtering statistics: - In the WebUI, browse to your dataset directory and select it to see filtering statistics - Check the logs during dataset processing for statistics like: Sample processing statistics: {'total_processed': 100, 'skipped': {'too_small': 15, ...}}

For detailed troubleshooting, see Troubleshooting filtered datasets in the dataloader documentation.