shared-components-512x896/metadata.json

{
  "metadata_version": "0.2",
  "kind": "diffusion",
  "name": "Z-Image-Shared-512x896",
  "assets": {
    "text_encoder": "TextEncoder.aimodel",
    "vae_decoder_half": "VAEDecoder_half.aimodel",
    "vae_encoder_half": "VAEEncoder_half.aimodel"
  },
  "diffusion": {
    "type": "z-image",
    "prediction_type": "flow_matching",
    "encoder_scale_factor": 0.3611,
    "decoder_scale_factor": 0.3611,
    "decoder_shift_factor": 0.1159,
    "image_size": 512,
    "image_height": 896,
    "patch_size": 2,
    "latent_channels": 16,
    "cap_feat_dim": 2560,
    "max_sequence_length": 512,
    "guidance_embeds": false,
    "default_guidance_scale": 0.0,
    "default_steps": 8,
    "rope_axes_dims": [
      32,
      48,
      48
    ],
    "rope_theta": 256.0,
    "scheduler_shift": 3.0
  },
  "source": {
    "model_definition": "torch",
    "hf_model_id": "Tongyi-MAI/Z-Image-Turbo"
  },
  "compression": "4bit",
  "compilation": {
    "targets": []
  }
}