zimage base vs turbo tests

zimage-turbo-vs-base-training/base/zbase_feliciaday_v1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d159a0ff71738cf2be23cc00e5f61231ff0f5511cdbd00e8e746a81cc5a945f
+size 170128296

zimage-turbo-vs-base-training/base/zbase_feliciaday_v1.yaml

@@ -0,0 +1,115 @@
+job: extension
+config:
+  name: zbase_feliciaday_v1
+  process:
+  - type: diffusion_trainer
+    training_folder: c:\Development\ai-toolkit\output
+    sqlite_db_path: c:\Development\ai-toolkit\aitk_db.db
+    device: cuda
+    trigger_word: sks woman
+    performance_log_every: 10
+    network:
+      type: lora
+      linear: 32
+      linear_alpha: 32
+      conv: 16
+      conv_alpha: 16
+      lokr_full_rank: true
+      lokr_factor: -1
+      network_kwargs:
+        ignore_if_contains: []
+    save:
+      dtype: bf16
+      save_every: 500
+      max_step_saves_to_keep: 18
+      save_format: diffusers
+      push_to_hub: false
+    datasets:
+    - folder_path: c:\Development\ai-toolkit\datasets/feliciaday
+      mask_path: null
+      mask_min_value: 0.1
+      default_caption: photo of a woman
+      caption_ext: txt
+      caption_dropout_rate: 0.05
+      cache_latents_to_disk: false
+      is_reg: false
+      network_weight: 1
+      resolution:
+      - 512
+      controls: []
+      shrink_video_to_frames: true
+      num_frames: 1
+      do_i2v: true
+      flip_x: false
+      flip_y: false
+    train:
+      batch_size: 1
+      bypass_guidance_embedding: false
+      steps: 5000
+      gradient_accumulation: 1
+      train_unet: true
+      train_text_encoder: false
+      gradient_checkpointing: true
+      noise_scheduler: flowmatch
+      optimizer: adamw8bit
+      timestep_type: weighted
+      content_or_style: balanced
+      optimizer_params:
+        weight_decay: 0.0001
+      unload_text_encoder: true
+      cache_text_embeddings: false
+      lr: 0.0001
+      ema_config:
+        use_ema: false
+        ema_decay: 0.99
+      skip_first_sample: true
+      force_first_sample: false
+      disable_sampling: false
+      dtype: bf16
+      diff_output_preservation: false
+      diff_output_preservation_multiplier: 1
+      diff_output_preservation_class: person
+      switch_boundary_every: 1
+      loss_type: mse
+      do_differential_guidance: true
+      differential_guidance_scale: 3
+    model:
+      name_or_path: Tongyi-MAI/Z-Image
+      quantize: true
+      qtype: qfloat8
+      quantize_te: true
+      qtype_te: qfloat8
+      arch: zimage
+      low_vram: true
+      model_kwargs: {}
+      layer_offloading: false
+      layer_offloading_text_encoder_percent: 1
+      layer_offloading_transformer_percent: 1
+    sample:
+      sampler: flowmatch
+      sample_every: 500
+      width: 512
+      height: 512
+      samples:
+      - prompt: woman with red hair, playing chess at the park, bomb going off in
+          the background
+      - prompt: a woman holding a coffee cup, in a beanie, sitting at a cafe
+      - prompt: a woman showing off her cool new t shirt at the beach, a shark is
+          jumping out of the water in the background
+      - prompt: woman playing the guitar, on stage, singing a song, laser lights,
+          punk rocker
+      - prompt: photo of a woman, white background, medium shot, modeling clothing,
+          studio lighting, white backdrop
+      neg: ''
+      seed: 42
+      walk_seed: true
+      guidance_scale: 1
+      sample_steps: 8
+      num_frames: 1
+      fps: 1
+    logging:
+      log_every: 1
+      use_ui_logger: true
+meta:
+  name: zbase_feliciaday_v1
+  version: '1.0'

zimage-turbo-vs-base-training/base/zimage_feliciaday_maltrainer.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c52474d63c811cc23e5ef1f6d6ccda62d3f7b698dea044f535298806be9eaa1
+size 190842896

zimage-turbo-vs-base-training/base/zimage_feliciaday_maltrainer.yaml

@@ -0,0 +1,69 @@
+# Z-Image LoRA Training Configuration Example
+
+# Model configuration
+model:
+  name_or_path: "Tongyi-MAI/Z-Image"  # or path to local model
+  
+# LoRA configuration
+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 configuration
+dataset:
+  path: "C:/Development/trainers/maltrainer/datasets/feliciaday"
+  trigger: "sks woman"
+  default_caption: "photo of a woman"
+  repeats: 1
+  resolution: 512
+  center_crop: true
+  random_flip: false
+  num_workers: 0
+
+# Training configuration
+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 configuration
+output:
+  path: "./output/feliciaday_lora"
+
+# Logging configuration
+logging:
+  level: "INFO"
+
+# Training settings
+dtype: "bfloat16"
+mixed_precision: "bf16"
+seed: 42

zimage-turbo-vs-base-training/base/zimage_feliciaday_onetrainer_adamw.json

@@ -0,0 +1,470 @@
+{
+    "__version": 10,
+    "training_method": "LORA",
+    "model_type": "Z_IMAGE",
+    "debug_mode": false,
+    "debug_dir": "debug",
+    "workspace_dir": "workspace/run",
+    "cache_dir": "workspace-cache/run",
+    "tensorboard": true,
+    "tensorboard_expose": false,
+    "tensorboard_always_on": false,
+    "tensorboard_port": 6006,
+    "validation": false,
+    "validate_after": 1,
+    "validate_after_unit": "EPOCH",
+    "continue_last_backup": false,
+    "include_train_config": "NONE",
+    "multi_gpu": false,
+    "device_indexes": "",
+    "gradient_reduce_precision": "FLOAT_32_STOCHASTIC",
+    "fused_gradient_reduce": true,
+    "async_gradient_reduce": true,
+    "async_gradient_reduce_buffer": 100,
+    "base_model_name": "Tongyi-MAI/Z-Image",
+    "output_dtype": "BFLOAT_16",
+    "output_model_format": "SAFETENSORS",
+    "output_model_destination": "models/model.safetensors",
+    "gradient_checkpointing": "ON",
+    "enable_async_offloading": true,
+    "enable_activation_offloading": true,
+    "layer_offload_fraction": 0.0,
+    "force_circular_padding": false,
+    "compile": true,
+    "concept_file_name": "training_concepts/concepts.json",
+    "concepts": null,
+    "aspect_ratio_bucketing": true,
+    "latent_caching": true,
+    "clear_cache_before_training": true,
+    "learning_rate_scheduler": "CONSTANT",
+    "custom_learning_rate_scheduler": null,
+    "scheduler_params": [],
+    "learning_rate": 0.0003,
+    "learning_rate_warmup_steps": 200.0,
+    "learning_rate_cycles": 1.0,
+    "learning_rate_min_factor": 0.0,
+    "epochs": 100,
+    "batch_size": 2,
+    "gradient_accumulation_steps": 1,
+    "ema": "OFF",
+    "ema_decay": 0.999,
+    "ema_update_step_interval": 5,
+    "dataloader_threads": 1,
+    "train_device": "cuda",
+    "temp_device": "cpu",
+    "train_dtype": "BFLOAT_16",
+    "fallback_train_dtype": "BFLOAT_16",
+    "enable_autocast_cache": true,
+    "only_cache": false,
+    "resolution": "512",
+    "frames": "25",
+    "mse_strength": 1.0,
+    "mae_strength": 0.0,
+    "log_cosh_strength": 0.0,
+    "huber_strength": 0.0,
+    "huber_delta": 1.0,
+    "vb_loss_strength": 1.0,
+    "loss_weight_fn": "CONSTANT",
+    "loss_weight_strength": 5.0,
+    "dropout_probability": 0.0,
+    "loss_scaler": "NONE",
+    "learning_rate_scaler": "NONE",
+    "clip_grad_norm": 1.0,
+    "offset_noise_weight": 0.0,
+    "generalized_offset_noise": false,
+    "perturbation_noise_weight": 0.0,
+    "rescale_noise_scheduler_to_zero_terminal_snr": false,
+    "force_v_prediction": false,
+    "force_epsilon_prediction": false,
+    "min_noising_strength": 0.0,
+    "max_noising_strength": 1.0,
+    "timestep_distribution": "LOGIT_NORMAL",
+    "noising_weight": 0.0,
+    "noising_bias": 0.0,
+    "timestep_shift": 1.0,
+    "dynamic_timestep_shifting": false,
+    "unet": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "prior": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "transformer": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "INT_W8A8",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "quantization": {
+        "__version": 0,
+        "layer_filter": "layers",
+        "layer_filter_preset": "blocks",
+        "layer_filter_regex": false,
+        "svd_dtype": "NONE",
+        "svd_rank": 16,
+        "cache_dir": "workspace-cache/run/quantization"
+    },
+    "text_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": false,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_8",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_layer_skip": 0,
+    "text_encoder_sequence_length": 512,
+    "text_encoder_2": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_2_layer_skip": 0,
+    "text_encoder_2_sequence_length": 77,
+    "text_encoder_3": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_3_layer_skip": 0,
+    "text_encoder_4": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_4_layer_skip": 0,
+    "vae": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "effnet_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder_text_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder_vqgan": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "masked_training": false,
+    "unmasked_probability": 0.1,
+    "unmasked_weight": 0.1,
+    "normalize_masked_area_loss": false,
+    "masked_prior_preservation_weight": 0.0,
+    "custom_conditioning_image": false,
+    "layer_filter": "^(?=.*attention)(?!.*refiner).*,^(?=.*feed_forward)(?!.*refiner).*",
+    "layer_filter_preset": "attn-mlp",
+    "layer_filter_regex": true,
+    "embedding_learning_rate": null,
+    "preserve_embedding_norm": false,
+    "embedding": {
+        "__version": 0,
+        "uuid": "356e3db1-1959-447f-933e-f1e055b200b7",
+        "model_name": "",
+        "placeholder": "<embedding>",
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "token_count": 1,
+        "initial_embedding_text": "*",
+        "is_output_embedding": false
+    },
+    "additional_embeddings": [],
+    "embedding_weight_dtype": "FLOAT_32",
+    "cloud": {
+        "__version": 0,
+        "enabled": false,
+        "type": "RUNPOD",
+        "file_sync": "NATIVE_SCP",
+        "create": true,
+        "name": "OneTrainer",
+        "tensorboard_tunnel": true,
+        "sub_type": "",
+        "gpu_type": "",
+        "volume_size": 100,
+        "min_download": 0,
+        "remote_dir": "/workspace",
+        "huggingface_cache_dir": "/workspace/huggingface_cache",
+        "onetrainer_dir": "/workspace/OneTrainer",
+        "install_cmd": "git clone https://github.com/Nerogar/OneTrainer",
+        "install_onetrainer": true,
+        "update_onetrainer": true,
+        "detach_trainer": false,
+        "run_id": "job1",
+        "download_samples": true,
+        "download_output_model": true,
+        "download_saves": true,
+        "download_backups": false,
+        "download_tensorboard": false,
+        "delete_workspace": false,
+        "on_finish": "NONE",
+        "on_error": "NONE",
+        "on_detached_finish": "NONE",
+        "on_detached_error": "NONE"
+    },
+    "peft_type": "LORA",
+    "lora_model_name": "",
+    "lora_rank": 128,
+    "lora_alpha": 128.0,
+    "lora_decompose": false,
+    "lora_decompose_norm_epsilon": true,
+    "lora_decompose_output_axis": false,
+    "lora_weight_dtype": "FLOAT_32",
+    "bundle_additional_embeddings": true,
+    "oft_block_size": 32,
+    "oft_coft": false,
+    "coft_eps": 0.0001,
+    "oft_block_share": false,
+    "optimizer": {
+        "__version": 0,
+        "optimizer": "PRODIGY",
+        "adam_w_mode": false,
+        "alpha": null,
+        "amsgrad": false,
+        "beta1": 0.9,
+        "beta2": 0.999,
+        "beta3": null,
+        "bias_correction": false,
+        "block_wise": false,
+        "capturable": false,
+        "centered": false,
+        "clip_threshold": null,
+        "d0": 1e-06,
+        "d_coef": 1.0,
+        "dampening": null,
+        "decay_rate": null,
+        "decouple": true,
+        "differentiable": false,
+        "eps": 1e-08,
+        "eps2": null,
+        "foreach": false,
+        "fsdp_in_use": false,
+        "fused": false,
+        "fused_back_pass": false,
+        "growth_rate": "inf",
+        "initial_accumulator_value": null,
+        "initial_accumulator": null,
+        "is_paged": false,
+        "log_every": null,
+        "lr_decay": null,
+        "max_unorm": null,
+        "maximize": false,
+        "min_8bit_size": null,
+        "quant_block_size": null,
+        "momentum": null,
+        "nesterov": false,
+        "no_prox": false,
+        "optim_bits": null,
+        "percentile_clipping": null,
+        "r": null,
+        "relative_step": false,
+        "safeguard_warmup": false,
+        "scale_parameter": false,
+        "stochastic_rounding": true,
+        "use_bias_correction": false,
+        "use_triton": false,
+        "warmup_init": false,
+        "weight_decay": 0.0,
+        "weight_lr_power": null,
+        "decoupled_decay": false,
+        "fixed_decay": false,
+        "rectify": false,
+        "degenerated_to_sgd": false,
+        "k": null,
+        "xi": null,
+        "n_sma_threshold": null,
+        "ams_bound": false,
+        "adanorm": false,
+        "adam_debias": false,
+        "slice_p": 11,
+        "cautious": false,
+        "weight_decay_by_lr": true,
+        "prodigy_steps": null,
+        "use_speed": false,
+        "split_groups": true,
+        "split_groups_mean": true,
+        "factored": true,
+        "factored_fp32": true,
+        "use_stableadamw": true,
+        "use_cautious": false,
+        "use_grams": false,
+        "use_adopt": false,
+        "d_limiter": true,
+        "use_schedulefree": true,
+        "use_orthograd": false,
+        "nnmf_factor": false,
+        "orthogonal_gradient": false,
+        "use_atan2": false,
+        "use_AdEMAMix": false,
+        "beta3_ema": null,
+        "alpha_grad": null,
+        "beta1_warmup": null,
+        "min_beta1": null,
+        "Simplified_AdEMAMix": false,
+        "cautious_mask": false,
+        "grams_moment": false,
+        "kourkoutas_beta": false,
+        "k_warmup_steps": null,
+        "schedulefree_c": null,
+        "ns_steps": null,
+        "MuonWithAuxAdam": false,
+        "muon_hidden_layers": null,
+        "muon_adam_regex": false,
+        "muon_adam_lr": null,
+        "muon_te1_adam_lr": null,
+        "muon_te2_adam_lr": null,
+        "muon_adam_config": null,
+        "rms_rescaling": true,
+        "normuon_variant": false,
+        "beta2_normuon": null,
+        "normuon_eps": null,
+        "low_rank_ortho": false,
+        "ortho_rank": null,
+        "accelerated_ns": false,
+        "cautious_wd": false,
+        "approx_mars": false,
+        "kappa_p": null,
+        "auto_kappa_p": false,
+        "compile": false
+    },
+    "optimizer_defaults": {},
+    "sample_definition_file_name": "training_samples/samples.json",
+    "samples": null,
+    "sample_after": 10,
+    "sample_after_unit": "NEVER",
+    "sample_skip_first": 0,
+    "sample_image_format": "JPG",
+    "sample_video_format": "MP4",
+    "sample_audio_format": "MP3",
+    "samples_to_tensorboard": true,
+    "non_ema_sampling": true,
+    "backup_after": 30,
+    "backup_after_unit": "MINUTE",
+    "rolling_backup": false,
+    "rolling_backup_count": 3,
+    "backup_before_save": true,
+    "save_every": 0,
+    "save_every_unit": "NEVER",
+    "save_skip_first": 0,
+    "save_filename_prefix": ""
+}

zimage-turbo-vs-base-training/base/zimage_feliciaday_onetrainer_adamw.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3e2afacd545236fe241bd9566e72531b8051582b514d071c85bbffba68a6263
+size 140162224

zimage-turbo-vs-base-training/base/zimage_feliciaday_onetrainer_prodigy.json

@@ -0,0 +1,470 @@
+{
+    "__version": 10,
+    "training_method": "LORA",
+    "model_type": "Z_IMAGE",
+    "debug_mode": false,
+    "debug_dir": "debug",
+    "workspace_dir": "workspace/run",
+    "cache_dir": "workspace-cache/run",
+    "tensorboard": true,
+    "tensorboard_expose": false,
+    "tensorboard_always_on": false,
+    "tensorboard_port": 6006,
+    "validation": false,
+    "validate_after": 1,
+    "validate_after_unit": "EPOCH",
+    "continue_last_backup": false,
+    "include_train_config": "NONE",
+    "multi_gpu": false,
+    "device_indexes": "",
+    "gradient_reduce_precision": "FLOAT_32_STOCHASTIC",
+    "fused_gradient_reduce": true,
+    "async_gradient_reduce": true,
+    "async_gradient_reduce_buffer": 100,
+    "base_model_name": "Tongyi-MAI/Z-Image",
+    "output_dtype": "BFLOAT_16",
+    "output_model_format": "SAFETENSORS",
+    "output_model_destination": "models/model.safetensors",
+    "gradient_checkpointing": "ON",
+    "enable_async_offloading": true,
+    "enable_activation_offloading": true,
+    "layer_offload_fraction": 0.0,
+    "force_circular_padding": false,
+    "compile": true,
+    "concept_file_name": "training_concepts/concepts.json",
+    "concepts": null,
+    "aspect_ratio_bucketing": true,
+    "latent_caching": true,
+    "clear_cache_before_training": true,
+    "learning_rate_scheduler": "COSINE",
+    "custom_learning_rate_scheduler": null,
+    "scheduler_params": [],
+    "learning_rate": 1.0,
+    "learning_rate_warmup_steps": 200.0,
+    "learning_rate_cycles": 1.0,
+    "learning_rate_min_factor": 0.0,
+    "epochs": 100,
+    "batch_size": 2,
+    "gradient_accumulation_steps": 1,
+    "ema": "OFF",
+    "ema_decay": 0.999,
+    "ema_update_step_interval": 5,
+    "dataloader_threads": 1,
+    "train_device": "cuda",
+    "temp_device": "cpu",
+    "train_dtype": "BFLOAT_16",
+    "fallback_train_dtype": "BFLOAT_16",
+    "enable_autocast_cache": true,
+    "only_cache": false,
+    "resolution": "512",
+    "frames": "25",
+    "mse_strength": 1.0,
+    "mae_strength": 0.0,
+    "log_cosh_strength": 0.0,
+    "huber_strength": 0.0,
+    "huber_delta": 1.0,
+    "vb_loss_strength": 1.0,
+    "loss_weight_fn": "CONSTANT",
+    "loss_weight_strength": 5.0,
+    "dropout_probability": 0.0,
+    "loss_scaler": "NONE",
+    "learning_rate_scaler": "NONE",
+    "clip_grad_norm": 1.0,
+    "offset_noise_weight": 0.0,
+    "generalized_offset_noise": false,
+    "perturbation_noise_weight": 0.0,
+    "rescale_noise_scheduler_to_zero_terminal_snr": false,
+    "force_v_prediction": false,
+    "force_epsilon_prediction": false,
+    "min_noising_strength": 0.0,
+    "max_noising_strength": 1.0,
+    "timestep_distribution": "LOGIT_NORMAL",
+    "noising_weight": 0.0,
+    "noising_bias": 0.0,
+    "timestep_shift": 1.0,
+    "dynamic_timestep_shifting": false,
+    "unet": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "prior": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "transformer": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "INT_W8A8",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "quantization": {
+        "__version": 0,
+        "layer_filter": "layers",
+        "layer_filter_preset": "blocks",
+        "layer_filter_regex": false,
+        "svd_dtype": "NONE",
+        "svd_rank": 16,
+        "cache_dir": "workspace-cache/run/quantization"
+    },
+    "text_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": false,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_8",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_layer_skip": 0,
+    "text_encoder_sequence_length": 512,
+    "text_encoder_2": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_2_layer_skip": 0,
+    "text_encoder_2_sequence_length": 77,
+    "text_encoder_3": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_3_layer_skip": 0,
+    "text_encoder_4": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_4_layer_skip": 0,
+    "vae": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "effnet_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder_text_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder_vqgan": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "masked_training": false,
+    "unmasked_probability": 0.1,
+    "unmasked_weight": 0.1,
+    "normalize_masked_area_loss": false,
+    "masked_prior_preservation_weight": 0.0,
+    "custom_conditioning_image": false,
+    "layer_filter": "^(?=.*attention)(?!.*refiner).*,^(?=.*feed_forward)(?!.*refiner).*",
+    "layer_filter_preset": "attn-mlp",
+    "layer_filter_regex": true,
+    "embedding_learning_rate": null,
+    "preserve_embedding_norm": false,
+    "embedding": {
+        "__version": 0,
+        "uuid": "356e3db1-1959-447f-933e-f1e055b200b7",
+        "model_name": "",
+        "placeholder": "<embedding>",
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "token_count": 1,
+        "initial_embedding_text": "*",
+        "is_output_embedding": false
+    },
+    "additional_embeddings": [],
+    "embedding_weight_dtype": "FLOAT_32",
+    "cloud": {
+        "__version": 0,
+        "enabled": false,
+        "type": "RUNPOD",
+        "file_sync": "NATIVE_SCP",
+        "create": true,
+        "name": "OneTrainer",
+        "tensorboard_tunnel": true,
+        "sub_type": "",
+        "gpu_type": "",
+        "volume_size": 100,
+        "min_download": 0,
+        "remote_dir": "/workspace",
+        "huggingface_cache_dir": "/workspace/huggingface_cache",
+        "onetrainer_dir": "/workspace/OneTrainer",
+        "install_cmd": "git clone https://github.com/Nerogar/OneTrainer",
+        "install_onetrainer": true,
+        "update_onetrainer": true,
+        "detach_trainer": false,
+        "run_id": "job1",
+        "download_samples": true,
+        "download_output_model": true,
+        "download_saves": true,
+        "download_backups": false,
+        "download_tensorboard": false,
+        "delete_workspace": false,
+        "on_finish": "NONE",
+        "on_error": "NONE",
+        "on_detached_finish": "NONE",
+        "on_detached_error": "NONE"
+    },
+    "peft_type": "LORA",
+    "lora_model_name": "",
+    "lora_rank": 128,
+    "lora_alpha": 128.0,
+    "lora_decompose": false,
+    "lora_decompose_norm_epsilon": true,
+    "lora_decompose_output_axis": false,
+    "lora_weight_dtype": "FLOAT_32",
+    "bundle_additional_embeddings": true,
+    "oft_block_size": 32,
+    "oft_coft": false,
+    "coft_eps": 0.0001,
+    "oft_block_share": false,
+    "optimizer": {
+        "__version": 0,
+        "optimizer": "PRODIGY",
+        "adam_w_mode": false,
+        "alpha": null,
+        "amsgrad": false,
+        "beta1": 0.9,
+        "beta2": 0.999,
+        "beta3": null,
+        "bias_correction": false,
+        "block_wise": false,
+        "capturable": false,
+        "centered": false,
+        "clip_threshold": null,
+        "d0": 1e-06,
+        "d_coef": 1.0,
+        "dampening": null,
+        "decay_rate": null,
+        "decouple": true,
+        "differentiable": false,
+        "eps": 1e-08,
+        "eps2": null,
+        "foreach": false,
+        "fsdp_in_use": false,
+        "fused": false,
+        "fused_back_pass": false,
+        "growth_rate": "inf",
+        "initial_accumulator_value": null,
+        "initial_accumulator": null,
+        "is_paged": false,
+        "log_every": null,
+        "lr_decay": null,
+        "max_unorm": null,
+        "maximize": false,
+        "min_8bit_size": null,
+        "quant_block_size": null,
+        "momentum": null,
+        "nesterov": false,
+        "no_prox": false,
+        "optim_bits": null,
+        "percentile_clipping": null,
+        "r": null,
+        "relative_step": false,
+        "safeguard_warmup": false,
+        "scale_parameter": false,
+        "stochastic_rounding": true,
+        "use_bias_correction": false,
+        "use_triton": false,
+        "warmup_init": false,
+        "weight_decay": 0.0,
+        "weight_lr_power": null,
+        "decoupled_decay": false,
+        "fixed_decay": false,
+        "rectify": false,
+        "degenerated_to_sgd": false,
+        "k": null,
+        "xi": null,
+        "n_sma_threshold": null,
+        "ams_bound": false,
+        "adanorm": false,
+        "adam_debias": false,
+        "slice_p": 11,
+        "cautious": false,
+        "weight_decay_by_lr": true,
+        "prodigy_steps": null,
+        "use_speed": false,
+        "split_groups": true,
+        "split_groups_mean": true,
+        "factored": true,
+        "factored_fp32": true,
+        "use_stableadamw": true,
+        "use_cautious": false,
+        "use_grams": false,
+        "use_adopt": false,
+        "d_limiter": true,
+        "use_schedulefree": true,
+        "use_orthograd": false,
+        "nnmf_factor": false,
+        "orthogonal_gradient": false,
+        "use_atan2": false,
+        "use_AdEMAMix": false,
+        "beta3_ema": null,
+        "alpha_grad": null,
+        "beta1_warmup": null,
+        "min_beta1": null,
+        "Simplified_AdEMAMix": false,
+        "cautious_mask": false,
+        "grams_moment": false,
+        "kourkoutas_beta": false,
+        "k_warmup_steps": null,
+        "schedulefree_c": null,
+        "ns_steps": null,
+        "MuonWithAuxAdam": false,
+        "muon_hidden_layers": null,
+        "muon_adam_regex": false,
+        "muon_adam_lr": null,
+        "muon_te1_adam_lr": null,
+        "muon_te2_adam_lr": null,
+        "muon_adam_config": null,
+        "rms_rescaling": true,
+        "normuon_variant": false,
+        "beta2_normuon": null,
+        "normuon_eps": null,
+        "low_rank_ortho": false,
+        "ortho_rank": null,
+        "accelerated_ns": false,
+        "cautious_wd": false,
+        "approx_mars": false,
+        "kappa_p": null,
+        "auto_kappa_p": false,
+        "compile": false
+    },
+    "optimizer_defaults": {},
+    "sample_definition_file_name": "training_samples/samples.json",
+    "samples": null,
+    "sample_after": 10,
+    "sample_after_unit": "NEVER",
+    "sample_skip_first": 0,
+    "sample_image_format": "JPG",
+    "sample_video_format": "MP4",
+    "sample_audio_format": "MP3",
+    "samples_to_tensorboard": true,
+    "non_ema_sampling": true,
+    "backup_after": 30,
+    "backup_after_unit": "MINUTE",
+    "rolling_backup": false,
+    "rolling_backup_count": 3,
+    "backup_before_save": true,
+    "save_every": 0,
+    "save_every_unit": "NEVER",
+    "save_skip_first": 0,
+    "save_filename_prefix": ""
+}

zimage-turbo-vs-base-training/base/zimage_feliciaday_onetrainer_prodigy.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a12ec9c9037b7e038232ecb11709b50a0e0e97ed8d37af1113e4162842e700b1
+size 1120746056

zimage-turbo-vs-base-training/readme.md

@@ -0,0 +1,41 @@
+# Z Image (Base) vs Z Image Turbo — Felicia Day LoRA Models
+
+## Dataset
+- Dataset: [dataset/](dataset/)
+
+## Model Trainers
+- **OneTrainer**: https://github.com/Nerogar/OneTrainer
+- **MalTrainer**: not released [yet] (by @malcolmrey)
+- **AI Toolkit**: https://github.com/ostris/ai-toolkit/
+
+---
+
+## Base (Z Image)
+
+### OneTrainer
+- [base/zimage_feliciaday_onetrainer_adamw.json](base/zimage_feliciaday_onetrainer_adamw.json)
+- [base/zimage_feliciaday_onetrainer_adamw.safetensors](base/zimage_feliciaday_onetrainer_adamw.safetensors)
+- [base/zimage_feliciaday_onetrainer_prodigy.json](base/zimage_feliciaday_onetrainer_prodigy.json)
+- [base/zimage_feliciaday_onetrainer_prodigy.safetensors](base/zimage_feliciaday_onetrainer_prodigy.safetensors)
+
+### maltrainer
+- [base/zimage_feliciaday_maltrainer.safetensors](base/zimage_feliciaday_maltrainer.safetensors)
+- [base/zimage_feliciaday_maltrainer.yaml](base/zimage_feliciaday_maltrainer.yaml)
+
+### AI Toolkit
+- [base/zbase_feliciaday_v1.safetensors](base/zbase_feliciaday_v1.safetensors)
+- [base/zbase_feliciaday_v1.yaml](base/zbase_feliciaday_v1.yaml)
+
+---
+
+## Turbo (Z Image Turbo)
+
+### OneTrainer
+- [turbo/zimage_feliciaday_onetrainer_adamw_turbo.json](turbo/zimage_feliciaday_onetrainer_adamw_turbo.json)
+- [turbo/zimage_feliciaday_onetrainer_adamw_turbo.safetensors](turbo/zimage_feliciaday_onetrainer_adamw_turbo.safetensors)
+
+### AI Toolkit
+- [turbo/zimage_feliciaday_v1.safetensors](turbo/zimage_feliciaday_v1.safetensors)
+- [turbo/zimage_feliciaday_v1.yaml](turbo/zimage_feliciaday_v1.yaml)
+- [turbo/zonetrainer_feliciaday_v1.safetensors](turbo/zonetrainer_feliciaday_v1.safetensors)
+- [turbo/zonetrainer_feliciaday_v1.yaml](turbo/zonetrainer_feliciaday_v1.yaml)

zimage-turbo-vs-base-training/turbo/zimage_feliciaday_onetrainer_adamw_turbo.json

@@ -0,0 +1,470 @@
+{
+    "__version": 10,
+    "training_method": "LORA",
+    "model_type": "Z_IMAGE",
+    "debug_mode": false,
+    "debug_dir": "debug",
+    "workspace_dir": "workspace/run",
+    "cache_dir": "workspace-cache/run",
+    "tensorboard": true,
+    "tensorboard_expose": false,
+    "tensorboard_always_on": false,
+    "tensorboard_port": 6006,
+    "validation": false,
+    "validate_after": 1,
+    "validate_after_unit": "EPOCH",
+    "continue_last_backup": false,
+    "include_train_config": "NONE",
+    "multi_gpu": false,
+    "device_indexes": "",
+    "gradient_reduce_precision": "FLOAT_32_STOCHASTIC",
+    "fused_gradient_reduce": true,
+    "async_gradient_reduce": true,
+    "async_gradient_reduce_buffer": 100,
+    "base_model_name": "Tongyi-MAI/Z-Image-Turbo",
+    "output_dtype": "BFLOAT_16",
+    "output_model_format": "SAFETENSORS",
+    "output_model_destination": "models/model.safetensors",
+    "gradient_checkpointing": "ON",
+    "enable_async_offloading": true,
+    "enable_activation_offloading": true,
+    "layer_offload_fraction": 0.0,
+    "force_circular_padding": false,
+    "compile": true,
+    "concept_file_name": "training_concepts/concepts.json",
+    "concepts": null,
+    "aspect_ratio_bucketing": true,
+    "latent_caching": true,
+    "clear_cache_before_training": true,
+    "learning_rate_scheduler": "CONSTANT",
+    "custom_learning_rate_scheduler": null,
+    "scheduler_params": [],
+    "learning_rate": 0.0003,
+    "learning_rate_warmup_steps": 200.0,
+    "learning_rate_cycles": 1.0,
+    "learning_rate_min_factor": 0.0,
+    "epochs": 100,
+    "batch_size": 2,
+    "gradient_accumulation_steps": 1,
+    "ema": "OFF",
+    "ema_decay": 0.999,
+    "ema_update_step_interval": 5,
+    "dataloader_threads": 1,
+    "train_device": "cuda",
+    "temp_device": "cpu",
+    "train_dtype": "BFLOAT_16",
+    "fallback_train_dtype": "BFLOAT_16",
+    "enable_autocast_cache": true,
+    "only_cache": false,
+    "resolution": "512",
+    "frames": "25",
+    "mse_strength": 1.0,
+    "mae_strength": 0.0,
+    "log_cosh_strength": 0.0,
+    "huber_strength": 0.0,
+    "huber_delta": 1.0,
+    "vb_loss_strength": 1.0,
+    "loss_weight_fn": "CONSTANT",
+    "loss_weight_strength": 5.0,
+    "dropout_probability": 0.0,
+    "loss_scaler": "NONE",
+    "learning_rate_scaler": "NONE",
+    "clip_grad_norm": 1.0,
+    "offset_noise_weight": 0.0,
+    "generalized_offset_noise": false,
+    "perturbation_noise_weight": 0.0,
+    "rescale_noise_scheduler_to_zero_terminal_snr": false,
+    "force_v_prediction": false,
+    "force_epsilon_prediction": false,
+    "min_noising_strength": 0.0,
+    "max_noising_strength": 1.0,
+    "timestep_distribution": "LOGIT_NORMAL",
+    "noising_weight": 0.0,
+    "noising_bias": 0.0,
+    "timestep_shift": 1.0,
+    "dynamic_timestep_shifting": false,
+    "unet": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "prior": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "transformer": {
+        "__version": 0,
+        "model_name": "https://huggingface.co/ostris/Z-Image-De-Turbo/blob/main/z_image_de_turbo_v1_bf16.safetensors",
+        "include": true,
+        "train": true,
+        "stop_training_after": 0,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "INT_W8A8",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "quantization": {
+        "__version": 0,
+        "layer_filter": "layers",
+        "layer_filter_preset": "blocks",
+        "layer_filter_regex": false,
+        "svd_dtype": "NONE",
+        "svd_rank": 16,
+        "cache_dir": "workspace-cache/run/quantization"
+    },
+    "text_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": false,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_8",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_layer_skip": 0,
+    "text_encoder_sequence_length": 512,
+    "text_encoder_2": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_2_layer_skip": 0,
+    "text_encoder_2_sequence_length": 77,
+    "text_encoder_3": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_3_layer_skip": 0,
+    "text_encoder_4": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": 30,
+        "stop_training_after_unit": "EPOCH",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "text_encoder_4_layer_skip": 0,
+    "vae": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "effnet_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder_text_encoder": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "decoder_vqgan": {
+        "__version": 0,
+        "model_name": "",
+        "include": true,
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "learning_rate": null,
+        "weight_dtype": "FLOAT_32",
+        "dropout_probability": 0.0,
+        "train_embedding": true,
+        "attention_mask": false,
+        "guidance_scale": 1.0
+    },
+    "masked_training": false,
+    "unmasked_probability": 0.1,
+    "unmasked_weight": 0.1,
+    "normalize_masked_area_loss": false,
+    "masked_prior_preservation_weight": 0.0,
+    "custom_conditioning_image": false,
+    "layer_filter": "^(?=.*attention)(?!.*refiner).*,^(?=.*feed_forward)(?!.*refiner).*",
+    "layer_filter_preset": "attn-mlp",
+    "layer_filter_regex": true,
+    "embedding_learning_rate": null,
+    "preserve_embedding_norm": false,
+    "embedding": {
+        "__version": 0,
+        "uuid": "356e3db1-1959-447f-933e-f1e055b200b7",
+        "model_name": "",
+        "placeholder": "<embedding>",
+        "train": true,
+        "stop_training_after": null,
+        "stop_training_after_unit": "NEVER",
+        "token_count": 1,
+        "initial_embedding_text": "*",
+        "is_output_embedding": false
+    },
+    "additional_embeddings": [],
+    "embedding_weight_dtype": "FLOAT_32",
+    "cloud": {
+        "__version": 0,
+        "enabled": false,
+        "type": "RUNPOD",
+        "file_sync": "NATIVE_SCP",
+        "create": true,
+        "name": "OneTrainer",
+        "tensorboard_tunnel": true,
+        "sub_type": "",
+        "gpu_type": "",
+        "volume_size": 100,
+        "min_download": 0,
+        "remote_dir": "/workspace",
+        "huggingface_cache_dir": "/workspace/huggingface_cache",
+        "onetrainer_dir": "/workspace/OneTrainer",
+        "install_cmd": "git clone https://github.com/Nerogar/OneTrainer",
+        "install_onetrainer": true,
+        "update_onetrainer": true,
+        "detach_trainer": false,
+        "run_id": "job1",
+        "download_samples": true,
+        "download_output_model": true,
+        "download_saves": true,
+        "download_backups": false,
+        "download_tensorboard": false,
+        "delete_workspace": false,
+        "on_finish": "NONE",
+        "on_error": "NONE",
+        "on_detached_finish": "NONE",
+        "on_detached_error": "NONE"
+    },
+    "peft_type": "LORA",
+    "lora_model_name": "",
+    "lora_rank": 16,
+    "lora_alpha": 1.0,
+    "lora_decompose": false,
+    "lora_decompose_norm_epsilon": true,
+    "lora_decompose_output_axis": false,
+    "lora_weight_dtype": "FLOAT_32",
+    "bundle_additional_embeddings": true,
+    "oft_block_size": 32,
+    "oft_coft": false,
+    "coft_eps": 0.0001,
+    "oft_block_share": false,
+    "optimizer": {
+        "__version": 0,
+        "optimizer": "ADAMW",
+        "adam_w_mode": false,
+        "alpha": null,
+        "amsgrad": false,
+        "beta1": 0.9,
+        "beta2": 0.999,
+        "beta3": null,
+        "bias_correction": false,
+        "block_wise": false,
+        "capturable": false,
+        "centered": false,
+        "clip_threshold": null,
+        "d0": null,
+        "d_coef": null,
+        "dampening": null,
+        "decay_rate": null,
+        "decouple": false,
+        "differentiable": false,
+        "eps": 1e-08,
+        "eps2": null,
+        "foreach": false,
+        "fsdp_in_use": false,
+        "fused": true,
+        "fused_back_pass": false,
+        "growth_rate": null,
+        "initial_accumulator_value": null,
+        "initial_accumulator": null,
+        "is_paged": false,
+        "log_every": null,
+        "lr_decay": null,
+        "max_unorm": null,
+        "maximize": false,
+        "min_8bit_size": null,
+        "quant_block_size": null,
+        "momentum": null,
+        "nesterov": false,
+        "no_prox": false,
+        "optim_bits": null,
+        "percentile_clipping": null,
+        "r": null,
+        "relative_step": false,
+        "safeguard_warmup": false,
+        "scale_parameter": false,
+        "stochastic_rounding": false,
+        "use_bias_correction": false,
+        "use_triton": false,
+        "warmup_init": false,
+        "weight_decay": 0.01,
+        "weight_lr_power": null,
+        "decoupled_decay": false,
+        "fixed_decay": false,
+        "rectify": false,
+        "degenerated_to_sgd": false,
+        "k": null,
+        "xi": null,
+        "n_sma_threshold": null,
+        "ams_bound": false,
+        "adanorm": false,
+        "adam_debias": false,
+        "slice_p": null,
+        "cautious": false,
+        "weight_decay_by_lr": true,
+        "prodigy_steps": null,
+        "use_speed": false,
+        "split_groups": true,
+        "split_groups_mean": true,
+        "factored": true,
+        "factored_fp32": true,
+        "use_stableadamw": true,
+        "use_cautious": false,
+        "use_grams": false,
+        "use_adopt": false,
+        "d_limiter": true,
+        "use_schedulefree": true,
+        "use_orthograd": false,
+        "nnmf_factor": false,
+        "orthogonal_gradient": false,
+        "use_atan2": false,
+        "use_AdEMAMix": false,
+        "beta3_ema": null,
+        "alpha_grad": null,
+        "beta1_warmup": null,
+        "min_beta1": null,
+        "Simplified_AdEMAMix": false,
+        "cautious_mask": false,
+        "grams_moment": false,
+        "kourkoutas_beta": false,
+        "k_warmup_steps": null,
+        "schedulefree_c": null,
+        "ns_steps": null,
+        "MuonWithAuxAdam": false,
+        "muon_hidden_layers": null,
+        "muon_adam_regex": false,
+        "muon_adam_lr": null,
+        "muon_te1_adam_lr": null,
+        "muon_te2_adam_lr": null,
+        "muon_adam_config": null,
+        "rms_rescaling": true,
+        "normuon_variant": false,
+        "beta2_normuon": null,
+        "normuon_eps": null,
+        "low_rank_ortho": false,
+        "ortho_rank": null,
+        "accelerated_ns": false,
+        "cautious_wd": false,
+        "approx_mars": false,
+        "kappa_p": null,
+        "auto_kappa_p": false,
+        "compile": false
+    },
+    "optimizer_defaults": {},
+    "sample_definition_file_name": "training_samples/samples.json",
+    "samples": null,
+    "sample_after": 10,
+    "sample_after_unit": "NEVER",
+    "sample_skip_first": 0,
+    "sample_image_format": "JPG",
+    "sample_video_format": "MP4",
+    "sample_audio_format": "MP3",
+    "samples_to_tensorboard": true,
+    "non_ema_sampling": true,
+    "backup_after": 30,
+    "backup_after_unit": "MINUTE",
+    "rolling_backup": false,
+    "rolling_backup_count": 3,
+    "backup_before_save": true,
+    "save_every": 0,
+    "save_every_unit": "NEVER",
+    "save_skip_first": 0,
+    "save_filename_prefix": ""
+}

zimage-turbo-vs-base-training/turbo/zimage_feliciaday_onetrainer_adamw_turbo.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a05cf6957b23d3ba4fbacba37be4d9aba456a3ef111710ef3514fa132c7ba522
+size 140162224

zimage-turbo-vs-base-training/turbo/zimage_feliciaday_v1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f560d7cb2b454992009ab2b307f55db99a7c3410f0eee62a316e66db0a23a7c
+size 170128296

zimage-turbo-vs-base-training/turbo/zimage_feliciaday_v1.yaml

@@ -0,0 +1,113 @@
+job: extension
+config:
+  name: zimage_feliciaday_v1
+  process:
+  - type: diffusion_trainer
+    training_folder: c:\Development\ai-toolkit\output
+    sqlite_db_path: c:\Development\ai-toolkit\aitk_db.db
+    device: cuda
+    trigger_word: sks woman
+    performance_log_every: 10
+    network:
+      type: lora
+      linear: 32
+      linear_alpha: 32
+      conv: 16
+      conv_alpha: 16
+      lokr_full_rank: true
+      lokr_factor: -1
+      network_kwargs:
+        ignore_if_contains: []
+    save:
+      dtype: bf16
+      save_every: 500
+      max_step_saves_to_keep: 8
+      save_format: diffusers
+      push_to_hub: false
+    datasets:
+    - folder_path: c:\Development\ai-toolkit\datasets/feliciaday
+      mask_path: null
+      mask_min_value: 0.1
+      default_caption: photo of a woman
+      caption_ext: txt
+      caption_dropout_rate: 0.05
+      cache_latents_to_disk: false
+      is_reg: false
+      network_weight: 1
+      resolution:
+      - 512
+      controls: []
+      shrink_video_to_frames: true
+      num_frames: 1
+      do_i2v: true
+      flip_x: false
+      flip_y: false
+    train:
+      batch_size: 1
+      bypass_guidance_embedding: false
+      steps: 2500
+      gradient_accumulation: 1
+      train_unet: true
+      train_text_encoder: false
+      gradient_checkpointing: true
+      noise_scheduler: flowmatch
+      optimizer: adamw8bit
+      timestep_type: weighted
+      content_or_style: balanced
+      optimizer_params:
+        weight_decay: 0.0001
+      unload_text_encoder: true
+      cache_text_embeddings: false
+      lr: 0.0001
+      ema_config:
+        use_ema: false
+        ema_decay: 0.99
+      skip_first_sample: false
+      force_first_sample: false
+      disable_sampling: true
+      dtype: bf16
+      diff_output_preservation: false
+      diff_output_preservation_multiplier: 1
+      diff_output_preservation_class: person
+      switch_boundary_every: 1
+      loss_type: mse
+      do_differential_guidance: true
+      differential_guidance_scale: 3
+    model:
+      name_or_path: Tongyi-MAI/Z-Image-Turbo
+      quantize: true
+      qtype: qfloat8
+      quantize_te: true
+      qtype_te: qfloat8
+      arch: zimage:turbo
+      low_vram: true
+      model_kwargs: {}
+      layer_offloading: false
+      layer_offloading_text_encoder_percent: 1
+      layer_offloading_transformer_percent: 1
+      assistant_lora_path: ostris/zimage_turbo_training_adapter/zimage_turbo_training_adapter_v1.safetensors
+    sample:
+      sampler: flowmatch
+      sample_every: 250
+      width: 512
+      height: 512
+      samples:
+      - prompt: woman with red hair, playing chess at the park, bomb going off in
+          the background
+      - prompt: a woman holding a coffee cup, in a beanie, sitting at a cafe
+      - prompt: a woman showing off her cool new t shirt at the beach, a shark is
+          jumping out of the water in the background
+      - prompt: woman playing the guitar, on stage, singing a song, laser lights,
+          punk rocker
+      - prompt: photo of a woman, white background, medium shot, modeling clothing,
+          studio lighting, white backdrop
+      neg: ''
+      seed: 42
+      walk_seed: true
+      guidance_scale: 1
+      sample_steps: 8
+      num_frames: 1
+      fps: 1
+meta:
+  name: zimage_feliciaday_v1
+  version: '1.0'

zimage-turbo-vs-base-training/turbo/zonetrainer_feliciaday_v1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b12c5fee8e22dd6ffe6d0a70a3906cc1e7a4227940db1b7d05c244c9ce1fdc34
+size 85094888

zimage-turbo-vs-base-training/turbo/zonetrainer_feliciaday_v1.yaml

@@ -0,0 +1,116 @@
+job: extension
+config:
+  name: zonetrainer_feliciaday_v1
+  process:
+  - type: diffusion_trainer
+    training_folder: c:\Development\ai-toolkit\output
+    sqlite_db_path: c:\Development\ai-toolkit\aitk_db.db
+    device: cuda
+    trigger_word: sks woman
+    performance_log_every: 10
+    network:
+      type: lora
+      linear: 16
+      linear_alpha: 1
+      conv: 16
+      conv_alpha: 1
+      lokr_full_rank: true
+      lokr_factor: -1
+      network_kwargs:
+        ignore_if_contains: []
+    save:
+      dtype: bf16
+      save_every: 2500
+      max_step_saves_to_keep: 8
+      save_format: safetensors
+      push_to_hub: false
+    datasets:
+    - folder_path: c:\Development\ai-toolkit\datasets/feliciaday
+      mask_path: null
+      mask_min_value: 0.1
+      default_caption: photo of a woman
+      caption_ext: txt
+      caption_dropout_rate: 0
+      cache_latents_to_disk: true
+      is_reg: false
+      network_weight: 1
+      resolution:
+      - 512
+      controls: []
+      shrink_video_to_frames: true
+      num_frames: 1
+      do_i2v: true
+      flip_x: false
+      flip_y: false
+    train:
+      batch_size: 1
+      bypass_guidance_embedding: false
+      steps: 2500
+      gradient_accumulation: 1
+      train_unet: true
+      train_text_encoder: false
+      gradient_checkpointing: true
+      noise_scheduler: flowmatch
+      optimizer: adamw
+      timestep_type: weighted
+      content_or_style: balanced
+      optimizer_params:
+        weight_decay: 0.01
+      unload_text_encoder: true
+      cache_text_embeddings: false
+      lr: 0.0003
+      ema_config:
+        use_ema: false
+        ema_decay: 0.999
+      skip_first_sample: false
+      force_first_sample: false
+      disable_sampling: true
+      dtype: bf16
+      diff_output_preservation: false
+      diff_output_preservation_multiplier: 1
+      diff_output_preservation_class: person
+      switch_boundary_every: 1
+      loss_type: mse
+      do_differential_guidance: true
+      differential_guidance_scale: 1
+    model:
+      name_or_path: Tongyi-MAI/Z-Image-Turbo
+      quantize: true
+      qtype: qfloat8
+      quantize_te: true
+      qtype_te: qfloat8
+      arch: zimage:turbo
+      low_vram: true
+      model_kwargs: {}
+      layer_offloading: true
+      layer_offloading_text_encoder_percent: 1
+      layer_offloading_transformer_percent: 1
+      assistant_lora_path: ostris/zimage_turbo_training_adapter/zimage_turbo_training_adapter_v1.safetensors
+    sample:
+      sampler: flowmatch
+      sample_every: 2500
+      width: 512
+      height: 512
+      samples:
+      - prompt: woman with red hair, playing chess at the park, bomb going off in
+          the background
+      - prompt: a woman holding a coffee cup, in a beanie, sitting at a cafe
+      - prompt: a woman showing off her cool new t shirt at the beach, a shark is
+          jumping out of the water in the background
+      - prompt: woman playing the guitar, on stage, singing a song, laser lights,
+          punk rocker
+      - prompt: photo of a woman, white background, medium shot, modeling clothing,
+          studio lighting, white backdrop
+      neg: ''
+      seed: 42
+      walk_seed: true
+      guidance_scale: 1
+      sample_steps: 8
+      num_frames: 1
+      fps: 1
+    logging:
+      log_every: 1
+      use_ui_logger: true
+meta:
+  name: zonetrainer_feliciaday_v1
+  version: '1.0'