| |
|
|
| |
| model: |
| name_or_path: "Tongyi-MAI/Z-Image" |
| |
| |
| lora: |
| prefix: "diffusion_model." |
| rank: 32 |
| alpha: 32 |
| target_modules: |
| - "to_q" |
| - "to_k" |
| - "to_v" |
| - "to_out.0" |
| - "feed_forward.w1" |
| - "feed_forward.w2" |
| - "feed_forward.w3" |
| - "adaLN_modulation.0" |
|
|
| |
| dataset: |
| path: "[dataset_location]/[person_to_train]" |
| trigger: "[instance_token] [class_token]" |
| default_caption: "photo of a [class_token]" |
| repeats: 1 |
| resolution: 512 |
| center_crop: true |
| random_flip: false |
| num_workers: 0 |
|
|
| |
| train: |
| batch_size: 1 |
| gradient_accumulation_steps: 1 |
| num_epochs: 100 |
| optimizer: "adamw8bit" |
| learning_rate: 1e-4 |
| adam_beta1: 0.9 |
| adam_beta2: 0.999 |
| adam_epsilon: 1e-8 |
| weight_decay: 0.01 |
| weight_decay_exclude_lora: true |
| timestep_weighting: "none" |
| do_differential_guidance: true |
| differential_guidance_scale: 3.0 |
| unconditional_prompt: "" |
| dynamic_noise_offset: true |
| noise_multiplier: 1.0 |
| random_noise_multiplier: 0.0 |
| random_noise_shift: 0.0 |
| latent_multiplier: 1.0 |
| noisy_latent_multiplier: 1.0 |
| max_grad_norm: 1.0 |
| save_every: 500 |
|
|
| |
| output: |
| path: "./output/[person_to_train]" |
|
|
| |
| logging: |
| level: "INFO" |
|
|
| |
| dtype: "bfloat16" |
| mixed_precision: "bf16" |
| seed: 42 |
|
|
