Upload 32 files

checkpoint-3200/README.md

@@ -0,0 +1,129 @@
+---
+license: other
+base_model: "black-forest-labs/FLUX.2-dev"
+tags:
+  - flux2
+  - flux2-diffusers
+  - text-to-image
+  - image-to-image
+  - diffusers
+  - simpletuner
+  - not-for-all-audiences
+  - lora
+
+  - template:sd-lora
+  - standard
+pipeline_tag: text-to-image
+inference: true
+
+---
+
+# quzo/fl2
+
+This is a PEFT LoRA derived from [black-forest-labs/FLUX.2-dev](https://huggingface.co/black-forest-labs/FLUX.2-dev).
+
+The main validation prompt used during training was:
+```
+bm82 man
+```
+
+
+## Validation settings
+- CFG: `7.5`
+- CFG Rescale: `0.0`
+- Steps: `20`
+- Sampler: `FlowMatchEulerDiscreteScheduler`
+- Seed: `None`
+- Resolution: `1024x1024`
+
+
+Note: The validation settings are not necessarily the same as the [training settings](#training-settings).
+
+
+
+
+<Gallery />
+
+The text encoder **was not** trained.
+You may reuse the base model text encoder for inference.
+
+
+## Training settings
+
+- Training epochs: 533
+- Training steps: 3200
+- Learning rate: 0.0001
+  - Learning rate schedule: constant_with_warmup
+  - Warmup steps: 0
+- Max grad value: 2.0
+- Effective batch size: 2
+  - Micro-batch size: 2
+  - Gradient accumulation steps: 1
+  - Number of GPUs: 1
+- Gradient checkpointing: True
+- Prediction type: flow_matching[]
+- Optimizer: adamw_bf16
+- Trainable parameter precision: Pure BF16
+- Base model precision: `no_change`
+- Caption dropout probability: 0.1%
+
+
+
+- LoRA Rank: 16
+- LoRA Alpha: 16.0
+- LoRA Dropout: 0.1
+- LoRA initialisation style: default
+- LoRA mode: Standard
+    
+
+## Datasets
+
+### training-images
+- Repeats: 0
+- Total number of images: 12
+- Total number of aspect buckets: 2
+- Resolution: 1.048576 megapixels
+- Cropped: False
+- Crop style: None
+- Crop aspect: None
+- Used for regularisation data: No
+
+
+## Inference
+
+
+```python
+import torch
+from diffusers import DiffusionPipeline
+
+model_id = 'black-forest-labs/FLUX.2-dev'
+adapter_id = 'quzo/fl2'
+pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
+pipeline.load_lora_weights(adapter_id)
+
+prompt = "bm82 man"
+negative_prompt = 'blurry, cropped, ugly'
+
+## Optional: quantise the model to save on vram.
+## Note: The model was not quantised during training, so it is not necessary to quantise it during inference time.
+#from optimum.quanto import quantize, freeze, qint8
+#quantize(pipeline.transformer, weights=qint8)
+#freeze(pipeline.transformer)
+    
+pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # the pipeline is already in its target precision level
+model_output = pipeline(
+    prompt=prompt,
+    negative_prompt=negative_prompt,
+    num_inference_steps=20,
+    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(42),
+    width=1024,
+    height=1024,
+    guidance_scale=7.5,
+).images[0]
+
+model_output.save("output.png", format="PNG")
+
+```
+
+
+

checkpoint-3200/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcd3a3a7c184a0031b7eef26313fd7020b0b6fcca1949725feca4d9993b0f896
+size 321008843

checkpoint-3200/pytorch_lora_weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88ade5a67008a9aba49b0b6aabf1dd60c57d072029937bf5fb092c49a212413f
+size 106979136

checkpoint-3200/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05c0092c4c618520653c8345b12f78775995d65375428b89ab85aaa646f7482d
+size 14757

checkpoint-3200/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce701db56c2af726940feb1b070907060c446d2e48225a0937aba9bfefcf32e6
+size 1401

checkpoint-3200/simpletuner_config.json

@@ -0,0 +1,387 @@
+{
+    "model_family": "flux2",
+    "model_flavour": "dev",
+    "controlnet": false,
+    "pretrained_model_name_or_path": "black-forest-labs/FLUX.2-dev",
+    "output_dir": "/home/Ubuntu/.simpletuner/output",
+    "logging_dir": "/home/Ubuntu/.simpletuner/output/logs",
+    "model_type": "lora",
+    "seed": null,
+    "resolution": 1024,
+    "resume_from_checkpoint": null,
+    "prediction_type": null,
+    "pretrained_vae_model_name_or_path": "black-forest-labs/FLUX.2-dev",
+    "vae_dtype": "bf16",
+    "vae_cache_ondemand": false,
+    "vae_cache_disable": false,
+    "accelerator_cache_clear_interval": null,
+    "aspect_bucket_rounding": 2,
+    "base_model_precision": "no_change",
+    "text_encoder_1_precision": "no_change",
+    "text_encoder_2_precision": "no_change",
+    "text_encoder_3_precision": "no_change",
+    "text_encoder_4_precision": "no_change",
+    "gradient_checkpointing_interval": null,
+    "offload_during_startup": false,
+    "quantize_via": "accelerator",
+    "wan_force_2_1_time_embedding": false,
+    "fuse_qkv_projections": false,
+    "rescale_betas_zero_snr": false,
+    "control": false,
+    "controlnet_custom_config": null,
+    "controlnet_model_name_or_path": null,
+    "tread_config": null,
+    "pretrained_transformer_model_name_or_path": null,
+    "pretrained_transformer_subfolder": "transformer",
+    "pretrained_unet_model_name_or_path": null,
+    "pretrained_unet_subfolder": "unet",
+    "pretrained_t5_model_name_or_path": null,
+    "revision": null,
+    "variant": null,
+    "base_model_default_dtype": "bf16",
+    "unet_attention_slice": false,
+    "num_train_epochs": 834,
+    "max_train_steps": 5000,
+    "train_batch_size": 2,
+    "learning_rate": 0.0001,
+    "optimizer": "adamw_bf16",
+    "lr_scheduler": "constant_with_warmup",
+    "gradient_accumulation_steps": 1,
+    "lr_warmup_steps": 0,
+    "checkpoints_total_limit": 5,
+    "gradient_checkpointing": true,
+    "enable_group_offload": false,
+    "ramtorch": false,
+    "ramtorch_target_modules": null,
+    "ramtorch_text_encoder": false,
+    "ramtorch_vae": false,
+    "ramtorch_controlnet": false,
+    "group_offload_type": "block_level",
+    "group_offload_blocks_per_group": 1,
+    "group_offload_use_stream": false,
+    "group_offload_to_disk_path": "",
+    "group_offload_text_encoder": false,
+    "group_offload_vae": false,
+    "offload_during_save": false,
+    "enable_chunked_feed_forward": false,
+    "feed_forward_chunk_size": null,
+    "train_text_encoder": false,
+    "text_encoder_lr": null,
+    "lyrics_embedder_train": false,
+    "lyrics_embedder_optimizer": null,
+    "lyrics_embedder_lr": null,
+    "lyrics_embedder_lr_scheduler": null,
+    "lr_num_cycles": 1,
+    "lr_power": 0.8,
+    "use_soft_min_snr": false,
+    "use_ema": false,
+    "ema_device": "cpu",
+    "ema_cpu_only": false,
+    "ema_update_interval": 1,
+    "ema_foreach_disable": false,
+    "ema_decay": 0.995,
+    "lora_rank": 16,
+    "lora_alpha": 16.0,
+    "lora_type": "standard",
+    "lora_dropout": 0.1,
+    "lora_init_type": "default",
+    "peft_lora_mode": "standard",
+    "singlora_ramp_up_steps": 0,
+    "init_lora": null,
+    "lycoris_config": "/home/Ubuntu/.simpletuner/configs/config/lycoris_config.json",
+    "init_lokr_norm": null,
+    "flux_lora_target": "all",
+    "acestep_lora_target": "attn_qkv+linear_qkv",
+    "use_dora": false,
+    "resolution_type": "pixel_area",
+    "data_backend_config": "/home/Ubuntu/.simpletuner/configs/111/multidatabackend-bartek_mizak.json",
+    "caption_strategy": "filename",
+    "conditioning_multidataset_sampling": "random",
+    "instance_prompt": null,
+    "parquet_caption_column": null,
+    "parquet_filename_column": null,
+    "ignore_missing_files": false,
+    "vae_cache_scan_behaviour": "recreate",
+    "vae_enable_slicing": false,
+    "vae_enable_tiling": false,
+    "vae_batch_size": 4,
+    "max_upscale_threshold": null,
+    "caption_dropout_probability": 0.1,
+    "tokenizer_max_length": 512,
+    "audio_max_duration_seconds": null,
+    "audio_min_duration_seconds": null,
+    "audio_channels": 1,
+    "audio_duration_interval": 3.0,
+    "audio_truncation_mode": "beginning",
+    "validation_step_interval": 100,
+    "validation_epoch_interval": null,
+    "disable_benchmark": false,
+    "validation_preview": false,
+    "validation_preview_steps": 1,
+    "validation_prompt": "bm82 man",
+    "validation_lyrics": null,
+    "validation_audio_duration": 30.0,
+    "num_validation_images": 1,
+    "num_eval_images": 4,
+    "eval_steps_interval": null,
+    "eval_timesteps": 28,
+    "eval_dataset_pooling": false,
+    "evaluation_type": "none",
+    "pretrained_evaluation_model_name_or_path": "openai/clip-vit-large-patch14-336",
+    "validation_guidance": 7.5,
+    "validation_num_inference_steps": 20,
+    "validation_on_startup": false,
+    "validation_method": "simpletuner-local",
+    "validation_external_script": null,
+    "validation_external_background": false,
+    "validation_using_datasets": false,
+    "validation_torch_compile": false,
+    "validation_guidance_real": 1.0,
+    "validation_no_cfg_until_timestep": 2,
+    "validation_negative_prompt": "blurry, cropped, ugly",
+    "validation_randomize": false,
+    "validation_seed": null,
+    "validation_multigpu": "batch-parallel",
+    "validation_disable": false,
+    "validation_prompt_library": false,
+    "user_prompt_library": null,
+    "eval_dataset_id": null,
+    "validation_stitch_input_location": "left",
+    "validation_guidance_rescale": 0.0,
+    "validation_disable_unconditional": false,
+    "validation_guidance_skip_layers": null,
+    "validation_guidance_skip_layers_start": 0.01,
+    "validation_guidance_skip_layers_stop": 0.2,
+    "validation_guidance_skip_scale": 2.8,
+    "validation_lycoris_strength": 1.0,
+    "validation_noise_scheduler": null,
+    "validation_num_video_frames": null,
+    "validation_resolution": "1024x1024",
+    "validation_seed_source": "cpu",
+    "validation_torch_compile_mode": "max-autotune",
+    "validation_adapter_path": null,
+    "validation_adapter_name": null,
+    "validation_adapter_strength": 1.0,
+    "validation_adapter_mode": "adapter_only",
+    "validation_adapter_config": null,
+    "i_know_what_i_am_doing": false,
+    "flow_sigmoid_scale": 1.0,
+    "flux_fast_schedule": false,
+    "flow_use_uniform_schedule": false,
+    "flow_use_beta_schedule": false,
+    "flow_beta_schedule_alpha": 2.0,
+    "flow_beta_schedule_beta": 2.0,
+    "flow_schedule_shift": 3.0,
+    "flow_schedule_auto_shift": false,
+    "flux_guidance_mode": "constant",
+    "flux_attention_masked_training": false,
+    "flux_guidance_value": 1.0,
+    "flux_guidance_min": 0.0,
+    "flux_guidance_max": 4.0,
+    "t5_padding": "unmodified",
+    "sd3_clip_uncond_behaviour": "empty_string",
+    "sd3_t5_uncond_behaviour": null,
+    "soft_min_snr_sigma_data": null,
+    "mixed_precision": "bf16",
+    "attention_mechanism": "diffusers",
+    "sla_config": null,
+    "sageattention_usage": {
+        "_value_": "inference",
+        "_name_": "INFERENCE",
+        "__objclass__": "{'_generate_next_value_': <staticmethod(<function Enum._generate_next_value_ at 0x7649ca9cdf80>)>, '__module__': 'simpletuner.helpers.training.attention_backend', 'allows_training': <property object at 0x764872f93880>, 'allows_inference': <property object at 0x764872f93790>, 'from_raw': <classmethod(<function AttentionBackendMode.from_raw at 0x764872fe85e0>)>, '_new_member_': <built-in method __new__ of type object at 0xa234e0>, '_use_args_': True, '_member_names_': ['TRAINING', 'INFERENCE', 'TRAINING_AND_INFERENCE'], '_member_map_': {'TRAINING': <AttentionBackendMode.TRAINING: 'training'>, 'INFERENCE': <AttentionBackendMode.INFERENCE: 'inference'>, 'TRAINING_AND_INFERENCE': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>}, '_value2member_map_': {'training': <AttentionBackendMode.TRAINING: 'training'>, 'inference': <AttentionBackendMode.INFERENCE: 'inference'>, 'training+inference': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>}, '_unhashable_values_': [], '_member_type_': <class 'str'>, '_value_repr_': <slot wrapper '__repr__' of 'str' objects>, '__dict__': <attribute '__dict__' of 'AttentionBackendMode' objects>, '__weakref__': <attribute '__weakref__' of 'AttentionBackendMode' objects>, '__doc__': None, 'TRAINING': <AttentionBackendMode.TRAINING: 'training'>, 'INFERENCE': <AttentionBackendMode.INFERENCE: 'inference'>, 'TRAINING_AND_INFERENCE': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>, '__repr__': <function Enum.__repr__ at 0x7649ca9ce0c0>, '__str__': <function Enum.__str__ at 0x7649ca9ce160>, '__format__': <function Enum.__format__ at 0x7649ca9ce2a0>, '__new__': <function Enum.__new__ at 0x7649ca9cde40>}",
+        "_sort_order_": 1
+    },
+    "disable_tf32": false,
+    "set_grads_to_none": false,
+    "noise_offset": 0.1,
+    "noise_offset_probability": 0.25,
+    "input_perturbation": 0.0,
+    "input_perturbation_steps": 0,
+    "lr_end": "4e-7",
+    "lr_scale": false,
+    "lr_scale_sqrt": false,
+    "ignore_final_epochs": false,
+    "freeze_encoder_before": 12,
+    "freeze_encoder_after": 17,
+    "freeze_encoder_strategy": "after",
+    "layer_freeze_strategy": null,
+    "fully_unload_text_encoder": false,
+    "save_text_encoder": false,
+    "text_encoder_limit": 100,
+    "prepend_instance_prompt": false,
+    "only_instance_prompt": false,
+    "data_aesthetic_score": 7.0,
+    "delete_unwanted_images": false,
+    "delete_problematic_images": false,
+    "disable_bucket_pruning": false,
+    "allow_dataset_oversubscription": false,
+    "disable_segmented_timestep_sampling": false,
+    "preserve_data_backend_cache": false,
+    "override_dataset_config": false,
+    "cache_dir": "/home/Ubuntu/.simpletuner/output/cache",
+    "cache_dir_text": "cache",
+    "cache_dir_vae": "",
+    "compress_disk_cache": true,
+    "aspect_bucket_disable_rebuild": false,
+    "keep_vae_loaded": false,
+    "skip_file_discovery": "",
+    "data_backend_sampling": "auto-weighting",
+    "image_processing_batch_size": 32,
+    "write_batch_size": 128,
+    "read_batch_size": 25,
+    "enable_multiprocessing": false,
+    "accelerate_config": null,
+    "deepspeed_config": null,
+    "fsdp_enable": false,
+    "fsdp_version": 2,
+    "fsdp_reshard_after_forward": false,
+    "fsdp_state_dict_type": "SHARDED_STATE_DICT",
+    "fsdp_cpu_ram_efficient_loading": false,
+    "fsdp_auto_wrap_policy": "TRANSFORMER_BASED_WRAP",
+    "fsdp_limit_all_gathers": false,
+    "fsdp_cpu_offload": false,
+    "fsdp_activation_checkpointing": false,
+    "fsdp_transformer_layer_cls_to_wrap": null,
+    "context_parallel_size": 1,
+    "context_parallel_comm_strategy": "allgather",
+    "num_processes": 1,
+    "num_machines": 1,
+    "accelerate_extra_args": null,
+    "main_process_ip": "127.0.0.1",
+    "main_process_port": 29500,
+    "machine_rank": 0,
+    "same_network": true,
+    "dynamo_backend": "no",
+    "dynamo_mode": "",
+    "dynamo_fullgraph": false,
+    "dynamo_dynamic": false,
+    "dynamo_use_regional_compilation": false,
+    "max_workers": 32,
+    "aws_max_pool_connections": 128,
+    "torch_num_threads": 8,
+    "dataloader_prefetch": false,
+    "dataloader_prefetch_qlen": 10,
+    "aspect_bucket_worker_count": 12,
+    "aspect_bucket_alignment": 16,
+    "minimum_image_size": null,
+    "maximum_image_size": null,
+    "target_downsample_size": null,
+    "metadata_update_interval": 3600,
+    "debug_aspect_buckets": false,
+    "debug_dataset_loader": false,
+    "print_filenames": false,
+    "print_sampler_statistics": false,
+    "timestep_bias_strategy": null,
+    "timestep_bias_begin": 0,
+    "timestep_bias_end": 1000,
+    "timestep_bias_multiplier": 1.0,
+    "timestep_bias_portion": 0.25,
+    "training_scheduler_timestep_spacing": "trailing",
+    "inference_scheduler_timestep_spacing": "trailing",
+    "loss_type": "l2",
+    "huber_schedule": "snr",
+    "huber_c": 0.1,
+    "snr_gamma": null,
+    "masked_loss_probability": 1.0,
+    "hidream_use_load_balancing_loss": false,
+    "hidream_load_balancing_loss_weight": null,
+    "adam_beta1": 0.9,
+    "adam_beta2": 0.999,
+    "optimizer_beta1": null,
+    "optimizer_beta2": null,
+    "optimizer_cpu_offload_method": null,
+    "gradient_precision": null,
+    "adam_weight_decay": 0.01,
+    "adam_epsilon": 1e-08,
+    "prodigy_steps": null,
+    "max_grad_norm": 2.0,
+    "optimizer_config": null,
+    "grad_clip_method": "value",
+    "optimizer_offload_gradients": false,
+    "fuse_optimizer": false,
+    "optimizer_release_gradients": false,
+    "push_to_hub": false,
+    "publishing_config": null,
+    "post_upload_script": null,
+    "push_checkpoints_to_hub": false,
+    "push_to_hub_background": false,
+    "hub_model_id": "quzo/fl2",
+    "model_card_private": false,
+    "model_card_safe_for_work": false,
+    "model_card_note": null,
+    "report_to": "none",
+    "custom_tracker": null,
+    "checkpoint_step_interval": 200,
+    "checkpoint_epoch_interval": null,
+    "checkpointing_rolling_steps": 0,
+    "checkpointing_use_tempdir": false,
+    "checkpoints_rolling_total_limit": 1,
+    "tracker_run_name": "simpletuner-testing",
+    "tracker_project_name": "simpletuner",
+    "tracker_image_layout": "gallery",
+    "enable_watermark": false,
+    "framerate": null,
+    "seed_for_each_device": true,
+    "snr_weight": 1.0,
+    "webhook_config": [
+        {
+            "webhook_type": "raw",
+            "callback_url": "http://0.0.0.0:8001/callback",
+            "log_level": "info",
+            "ssl_no_verify": false
+        }
+    ],
+    "webhook_reporting_interval": null,
+    "distillation_method": null,
+    "distillation_config": null,
+    "ema_validation": "comparison",
+    "local_rank": -1,
+    "ltx_train_mode": "i2v",
+    "ltx_i2v_prob": 0.1,
+    "ltx_partial_noise_fraction": 0.05,
+    "ltx_protect_first_frame": false,
+    "offload_param_path": null,
+    "offset_noise": false,
+    "quantize_activations": false,
+    "refiner_training": false,
+    "refiner_training_invert_schedule": false,
+    "refiner_training_strength": 0.2,
+    "sdxl_refiner_uses_full_range": false,
+    "sana_complex_human_instruction": "Given a user prompt, generate an 'Enhanced prompt' that provides detailed visual descriptions suitable for image generation. Evaluate the level of detail in the user prompt:\n- If the prompt is simple, focus on adding specifics about colors, shapes, sizes, textures, and spatial relationships to create vivid and concrete scenes.\n- Do not add detail that is not in the original prompt! Do not mention colors if the User Prompt does not!\n- If the prompt is already detailed, refine and enhance the existing details slightly without overcomplicating.\n- If the prompt contains explicit material, do not remove this information from the output, which would damage the performance of downstream classifiers.\nHere are examples of how to transform or refine prompts:\n- User Prompt: A cat sleeping -> Enhanced: A cat sleeping peacefully, showcasing the joy of pet ownership. Cute floof kitty cat gatto.\n- User Prompt: A busy city street -> Enhanced: A bustling city street scene featuring a crowd of people.\nPlease generate only the enhanced description for the prompt below and avoid including any additional commentary or evaluations:\nUser Prompt: ",
+    "vae_path": "black-forest-labs/FLUX.2-dev",
+    "accelerator_project_config": {
+        "project_dir": "/home/Ubuntu/.simpletuner/output",
+        "logging_dir": "/home/Ubuntu/.simpletuner/output/logs",
+        "automatic_checkpoint_naming": false,
+        "total_limit": null,
+        "iteration": 16,
+        "save_on_each_node": false
+    },
+    "process_group_kwargs": {
+        "backend": "nccl",
+        "init_method": null,
+        "timeout": "1:30:00"
+    },
+    "is_quantized": false,
+    "weight_dtype": "torch.bfloat16",
+    "disable_accelerator": false,
+    "lora_initialisation_style": true,
+    "checkpointing_steps": 200,
+    "use_fsdp": false,
+    "model_type_label": "Flux.2",
+    "use_deepspeed_optimizer": false,
+    "use_deepspeed_scheduler": false,
+    "base_weight_dtype": "torch.bfloat16",
+    "is_quanto": false,
+    "is_torchao": false,
+    "is_bnb": false,
+    "flow_matching": true,
+    "overrode_max_train_steps": false,
+    "total_num_batches": 6,
+    "num_update_steps_per_epoch": 6,
+    "total_batch_size": 2,
+    "enable_adamw_bf16": true,
+    "is_schedulefree": false,
+    "is_lr_scheduler_disabled": false,
+    "total_steps_remaining_at_start": 5000
+}

checkpoint-3200/training_state-training-images.json

@@ -0,0 +1 @@
+{"aspect_ratio_bucket_indices": {"1.0": ["/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im008.png", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im021.png"], "0.74": ["/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0014.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9743.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9827.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9935.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0003.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9796.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9890.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9865.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9260.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0090.jpg"]}, "buckets": ["1.0", "0.74"], "exhausted_buckets": [], "batch_size": 2, "current_bucket": 1, "seen_images": {"/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im008.png": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im021.png": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9743.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9796.jpg": true}, "current_epoch": 534}

checkpoint-3200/training_state.json

@@ -0,0 +1 @@
+{"global_step": 3200, "epoch_step": 3733, "epoch": 534, "exhausted_backends": [], "repeats": {"training-images": 0}}

checkpoint-3400/README.md

@@ -0,0 +1,129 @@
+---
+license: other
+base_model: "black-forest-labs/FLUX.2-dev"
+tags:
+  - flux2
+  - flux2-diffusers
+  - text-to-image
+  - image-to-image
+  - diffusers
+  - simpletuner
+  - not-for-all-audiences
+  - lora
+
+  - template:sd-lora
+  - standard
+pipeline_tag: text-to-image
+inference: true
+
+---
+
+# quzo/fl2
+
+This is a PEFT LoRA derived from [black-forest-labs/FLUX.2-dev](https://huggingface.co/black-forest-labs/FLUX.2-dev).
+
+The main validation prompt used during training was:
+```
+bm82 man
+```
+
+
+## Validation settings
+- CFG: `7.5`
+- CFG Rescale: `0.0`
+- Steps: `20`
+- Sampler: `FlowMatchEulerDiscreteScheduler`
+- Seed: `None`
+- Resolution: `1024x1024`
+
+
+Note: The validation settings are not necessarily the same as the [training settings](#training-settings).
+
+
+
+
+<Gallery />
+
+The text encoder **was not** trained.
+You may reuse the base model text encoder for inference.
+
+
+## Training settings
+
+- Training epochs: 566
+- Training steps: 3400
+- Learning rate: 0.0001
+  - Learning rate schedule: constant_with_warmup
+  - Warmup steps: 0
+- Max grad value: 2.0
+- Effective batch size: 2
+  - Micro-batch size: 2
+  - Gradient accumulation steps: 1
+  - Number of GPUs: 1
+- Gradient checkpointing: True
+- Prediction type: flow_matching[]
+- Optimizer: adamw_bf16
+- Trainable parameter precision: Pure BF16
+- Base model precision: `no_change`
+- Caption dropout probability: 0.1%
+
+
+
+- LoRA Rank: 16
+- LoRA Alpha: 16.0
+- LoRA Dropout: 0.1
+- LoRA initialisation style: default
+- LoRA mode: Standard
+    
+
+## Datasets
+
+### training-images
+- Repeats: 0
+- Total number of images: 12
+- Total number of aspect buckets: 1
+- Resolution: 1.048576 megapixels
+- Cropped: False
+- Crop style: None
+- Crop aspect: None
+- Used for regularisation data: No
+
+
+## Inference
+
+
+```python
+import torch
+from diffusers import DiffusionPipeline
+
+model_id = 'black-forest-labs/FLUX.2-dev'
+adapter_id = 'quzo/fl2'
+pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
+pipeline.load_lora_weights(adapter_id)
+
+prompt = "bm82 man"
+negative_prompt = 'blurry, cropped, ugly'
+
+## Optional: quantise the model to save on vram.
+## Note: The model was not quantised during training, so it is not necessary to quantise it during inference time.
+#from optimum.quanto import quantize, freeze, qint8
+#quantize(pipeline.transformer, weights=qint8)
+#freeze(pipeline.transformer)
+    
+pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # the pipeline is already in its target precision level
+model_output = pipeline(
+    prompt=prompt,
+    negative_prompt=negative_prompt,
+    num_inference_steps=20,
+    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(42),
+    width=1024,
+    height=1024,
+    guidance_scale=7.5,
+).images[0]
+
+model_output.save("output.png", format="PNG")
+
+```
+
+
+

checkpoint-3400/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a1cf4adf651ea45be3c0204a409ab22a3913476ee69840d1ffce3c8b45697c1
+size 321008843

checkpoint-3400/pytorch_lora_weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:414a5aff8b7277e413b9f104c7edd4bbcd12ab0d261beb22649b90a62d304fc4
+size 106979136

checkpoint-3400/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01b906f76c1ba1117bc03af59e66750f2205f8a2f01c5e6240e20dc436024aae
+size 14885

checkpoint-3400/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f02eb14703c391ff6b0683ceeea335d568ec18788de76c6f5aa899ec9eb225ae
+size 1401

checkpoint-3400/simpletuner_config.json

@@ -0,0 +1,387 @@
+{
+    "model_family": "flux2",
+    "model_flavour": "dev",
+    "controlnet": false,
+    "pretrained_model_name_or_path": "black-forest-labs/FLUX.2-dev",
+    "output_dir": "/home/Ubuntu/.simpletuner/output",
+    "logging_dir": "/home/Ubuntu/.simpletuner/output/logs",
+    "model_type": "lora",
+    "seed": null,
+    "resolution": 1024,
+    "resume_from_checkpoint": null,
+    "prediction_type": null,
+    "pretrained_vae_model_name_or_path": "black-forest-labs/FLUX.2-dev",
+    "vae_dtype": "bf16",
+    "vae_cache_ondemand": false,
+    "vae_cache_disable": false,
+    "accelerator_cache_clear_interval": null,
+    "aspect_bucket_rounding": 2,
+    "base_model_precision": "no_change",
+    "text_encoder_1_precision": "no_change",
+    "text_encoder_2_precision": "no_change",
+    "text_encoder_3_precision": "no_change",
+    "text_encoder_4_precision": "no_change",
+    "gradient_checkpointing_interval": null,
+    "offload_during_startup": false,
+    "quantize_via": "accelerator",
+    "wan_force_2_1_time_embedding": false,
+    "fuse_qkv_projections": false,
+    "rescale_betas_zero_snr": false,
+    "control": false,
+    "controlnet_custom_config": null,
+    "controlnet_model_name_or_path": null,
+    "tread_config": null,
+    "pretrained_transformer_model_name_or_path": null,
+    "pretrained_transformer_subfolder": "transformer",
+    "pretrained_unet_model_name_or_path": null,
+    "pretrained_unet_subfolder": "unet",
+    "pretrained_t5_model_name_or_path": null,
+    "revision": null,
+    "variant": null,
+    "base_model_default_dtype": "bf16",
+    "unet_attention_slice": false,
+    "num_train_epochs": 834,
+    "max_train_steps": 5000,
+    "train_batch_size": 2,
+    "learning_rate": 0.0001,
+    "optimizer": "adamw_bf16",
+    "lr_scheduler": "constant_with_warmup",
+    "gradient_accumulation_steps": 1,
+    "lr_warmup_steps": 0,
+    "checkpoints_total_limit": 5,
+    "gradient_checkpointing": true,
+    "enable_group_offload": false,
+    "ramtorch": false,
+    "ramtorch_target_modules": null,
+    "ramtorch_text_encoder": false,
+    "ramtorch_vae": false,
+    "ramtorch_controlnet": false,
+    "group_offload_type": "block_level",
+    "group_offload_blocks_per_group": 1,
+    "group_offload_use_stream": false,
+    "group_offload_to_disk_path": "",
+    "group_offload_text_encoder": false,
+    "group_offload_vae": false,
+    "offload_during_save": false,
+    "enable_chunked_feed_forward": false,
+    "feed_forward_chunk_size": null,
+    "train_text_encoder": false,
+    "text_encoder_lr": null,
+    "lyrics_embedder_train": false,
+    "lyrics_embedder_optimizer": null,
+    "lyrics_embedder_lr": null,
+    "lyrics_embedder_lr_scheduler": null,
+    "lr_num_cycles": 1,
+    "lr_power": 0.8,
+    "use_soft_min_snr": false,
+    "use_ema": false,
+    "ema_device": "cpu",
+    "ema_cpu_only": false,
+    "ema_update_interval": 1,
+    "ema_foreach_disable": false,
+    "ema_decay": 0.995,
+    "lora_rank": 16,
+    "lora_alpha": 16.0,
+    "lora_type": "standard",
+    "lora_dropout": 0.1,
+    "lora_init_type": "default",
+    "peft_lora_mode": "standard",
+    "singlora_ramp_up_steps": 0,
+    "init_lora": null,
+    "lycoris_config": "/home/Ubuntu/.simpletuner/configs/config/lycoris_config.json",
+    "init_lokr_norm": null,
+    "flux_lora_target": "all",
+    "acestep_lora_target": "attn_qkv+linear_qkv",
+    "use_dora": false,
+    "resolution_type": "pixel_area",
+    "data_backend_config": "/home/Ubuntu/.simpletuner/configs/111/multidatabackend-bartek_mizak.json",
+    "caption_strategy": "filename",
+    "conditioning_multidataset_sampling": "random",
+    "instance_prompt": null,
+    "parquet_caption_column": null,
+    "parquet_filename_column": null,
+    "ignore_missing_files": false,
+    "vae_cache_scan_behaviour": "recreate",
+    "vae_enable_slicing": false,
+    "vae_enable_tiling": false,
+    "vae_batch_size": 4,
+    "max_upscale_threshold": null,
+    "caption_dropout_probability": 0.1,
+    "tokenizer_max_length": 512,
+    "audio_max_duration_seconds": null,
+    "audio_min_duration_seconds": null,
+    "audio_channels": 1,
+    "audio_duration_interval": 3.0,
+    "audio_truncation_mode": "beginning",
+    "validation_step_interval": 100,
+    "validation_epoch_interval": null,
+    "disable_benchmark": false,
+    "validation_preview": false,
+    "validation_preview_steps": 1,
+    "validation_prompt": "bm82 man",
+    "validation_lyrics": null,
+    "validation_audio_duration": 30.0,
+    "num_validation_images": 1,
+    "num_eval_images": 4,
+    "eval_steps_interval": null,
+    "eval_timesteps": 28,
+    "eval_dataset_pooling": false,
+    "evaluation_type": "none",
+    "pretrained_evaluation_model_name_or_path": "openai/clip-vit-large-patch14-336",
+    "validation_guidance": 7.5,
+    "validation_num_inference_steps": 20,
+    "validation_on_startup": false,
+    "validation_method": "simpletuner-local",
+    "validation_external_script": null,
+    "validation_external_background": false,
+    "validation_using_datasets": false,
+    "validation_torch_compile": false,
+    "validation_guidance_real": 1.0,
+    "validation_no_cfg_until_timestep": 2,
+    "validation_negative_prompt": "blurry, cropped, ugly",
+    "validation_randomize": false,
+    "validation_seed": null,
+    "validation_multigpu": "batch-parallel",
+    "validation_disable": false,
+    "validation_prompt_library": false,
+    "user_prompt_library": null,
+    "eval_dataset_id": null,
+    "validation_stitch_input_location": "left",
+    "validation_guidance_rescale": 0.0,
+    "validation_disable_unconditional": false,
+    "validation_guidance_skip_layers": null,
+    "validation_guidance_skip_layers_start": 0.01,
+    "validation_guidance_skip_layers_stop": 0.2,
+    "validation_guidance_skip_scale": 2.8,
+    "validation_lycoris_strength": 1.0,
+    "validation_noise_scheduler": null,
+    "validation_num_video_frames": null,
+    "validation_resolution": "1024x1024",
+    "validation_seed_source": "cpu",
+    "validation_torch_compile_mode": "max-autotune",
+    "validation_adapter_path": null,
+    "validation_adapter_name": null,
+    "validation_adapter_strength": 1.0,
+    "validation_adapter_mode": "adapter_only",
+    "validation_adapter_config": null,
+    "i_know_what_i_am_doing": false,
+    "flow_sigmoid_scale": 1.0,
+    "flux_fast_schedule": false,
+    "flow_use_uniform_schedule": false,
+    "flow_use_beta_schedule": false,
+    "flow_beta_schedule_alpha": 2.0,
+    "flow_beta_schedule_beta": 2.0,
+    "flow_schedule_shift": 3.0,
+    "flow_schedule_auto_shift": false,
+    "flux_guidance_mode": "constant",
+    "flux_attention_masked_training": false,
+    "flux_guidance_value": 1.0,
+    "flux_guidance_min": 0.0,
+    "flux_guidance_max": 4.0,
+    "t5_padding": "unmodified",
+    "sd3_clip_uncond_behaviour": "empty_string",
+    "sd3_t5_uncond_behaviour": null,
+    "soft_min_snr_sigma_data": null,
+    "mixed_precision": "bf16",
+    "attention_mechanism": "diffusers",
+    "sla_config": null,
+    "sageattention_usage": {
+        "_value_": "inference",
+        "_name_": "INFERENCE",
+        "__objclass__": "{'_generate_next_value_': <staticmethod(<function Enum._generate_next_value_ at 0x7649ca9cdf80>)>, '__module__': 'simpletuner.helpers.training.attention_backend', 'allows_training': <property object at 0x764872f93880>, 'allows_inference': <property object at 0x764872f93790>, 'from_raw': <classmethod(<function AttentionBackendMode.from_raw at 0x764872fe85e0>)>, '_new_member_': <built-in method __new__ of type object at 0xa234e0>, '_use_args_': True, '_member_names_': ['TRAINING', 'INFERENCE', 'TRAINING_AND_INFERENCE'], '_member_map_': {'TRAINING': <AttentionBackendMode.TRAINING: 'training'>, 'INFERENCE': <AttentionBackendMode.INFERENCE: 'inference'>, 'TRAINING_AND_INFERENCE': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>}, '_value2member_map_': {'training': <AttentionBackendMode.TRAINING: 'training'>, 'inference': <AttentionBackendMode.INFERENCE: 'inference'>, 'training+inference': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>}, '_unhashable_values_': [], '_member_type_': <class 'str'>, '_value_repr_': <slot wrapper '__repr__' of 'str' objects>, '__dict__': <attribute '__dict__' of 'AttentionBackendMode' objects>, '__weakref__': <attribute '__weakref__' of 'AttentionBackendMode' objects>, '__doc__': None, 'TRAINING': <AttentionBackendMode.TRAINING: 'training'>, 'INFERENCE': <AttentionBackendMode.INFERENCE: 'inference'>, 'TRAINING_AND_INFERENCE': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>, '__repr__': <function Enum.__repr__ at 0x7649ca9ce0c0>, '__str__': <function Enum.__str__ at 0x7649ca9ce160>, '__format__': <function Enum.__format__ at 0x7649ca9ce2a0>, '__new__': <function Enum.__new__ at 0x7649ca9cde40>}",
+        "_sort_order_": 1
+    },
+    "disable_tf32": false,
+    "set_grads_to_none": false,
+    "noise_offset": 0.1,
+    "noise_offset_probability": 0.25,
+    "input_perturbation": 0.0,
+    "input_perturbation_steps": 0,
+    "lr_end": "4e-7",
+    "lr_scale": false,
+    "lr_scale_sqrt": false,
+    "ignore_final_epochs": false,
+    "freeze_encoder_before": 12,
+    "freeze_encoder_after": 17,
+    "freeze_encoder_strategy": "after",
+    "layer_freeze_strategy": null,
+    "fully_unload_text_encoder": false,
+    "save_text_encoder": false,
+    "text_encoder_limit": 100,
+    "prepend_instance_prompt": false,
+    "only_instance_prompt": false,
+    "data_aesthetic_score": 7.0,
+    "delete_unwanted_images": false,
+    "delete_problematic_images": false,
+    "disable_bucket_pruning": false,
+    "allow_dataset_oversubscription": false,
+    "disable_segmented_timestep_sampling": false,
+    "preserve_data_backend_cache": false,
+    "override_dataset_config": false,
+    "cache_dir": "/home/Ubuntu/.simpletuner/output/cache",
+    "cache_dir_text": "cache",
+    "cache_dir_vae": "",
+    "compress_disk_cache": true,
+    "aspect_bucket_disable_rebuild": false,
+    "keep_vae_loaded": false,
+    "skip_file_discovery": "",
+    "data_backend_sampling": "auto-weighting",
+    "image_processing_batch_size": 32,
+    "write_batch_size": 128,
+    "read_batch_size": 25,
+    "enable_multiprocessing": false,
+    "accelerate_config": null,
+    "deepspeed_config": null,
+    "fsdp_enable": false,
+    "fsdp_version": 2,
+    "fsdp_reshard_after_forward": false,
+    "fsdp_state_dict_type": "SHARDED_STATE_DICT",
+    "fsdp_cpu_ram_efficient_loading": false,
+    "fsdp_auto_wrap_policy": "TRANSFORMER_BASED_WRAP",
+    "fsdp_limit_all_gathers": false,
+    "fsdp_cpu_offload": false,
+    "fsdp_activation_checkpointing": false,
+    "fsdp_transformer_layer_cls_to_wrap": null,
+    "context_parallel_size": 1,
+    "context_parallel_comm_strategy": "allgather",
+    "num_processes": 1,
+    "num_machines": 1,
+    "accelerate_extra_args": null,
+    "main_process_ip": "127.0.0.1",
+    "main_process_port": 29500,
+    "machine_rank": 0,
+    "same_network": true,
+    "dynamo_backend": "no",
+    "dynamo_mode": "",
+    "dynamo_fullgraph": false,
+    "dynamo_dynamic": false,
+    "dynamo_use_regional_compilation": false,
+    "max_workers": 32,
+    "aws_max_pool_connections": 128,
+    "torch_num_threads": 8,
+    "dataloader_prefetch": false,
+    "dataloader_prefetch_qlen": 10,
+    "aspect_bucket_worker_count": 12,
+    "aspect_bucket_alignment": 16,
+    "minimum_image_size": null,
+    "maximum_image_size": null,
+    "target_downsample_size": null,
+    "metadata_update_interval": 3600,
+    "debug_aspect_buckets": false,
+    "debug_dataset_loader": false,
+    "print_filenames": false,
+    "print_sampler_statistics": false,
+    "timestep_bias_strategy": null,
+    "timestep_bias_begin": 0,
+    "timestep_bias_end": 1000,
+    "timestep_bias_multiplier": 1.0,
+    "timestep_bias_portion": 0.25,
+    "training_scheduler_timestep_spacing": "trailing",
+    "inference_scheduler_timestep_spacing": "trailing",
+    "loss_type": "l2",
+    "huber_schedule": "snr",
+    "huber_c": 0.1,
+    "snr_gamma": null,
+    "masked_loss_probability": 1.0,
+    "hidream_use_load_balancing_loss": false,
+    "hidream_load_balancing_loss_weight": null,
+    "adam_beta1": 0.9,
+    "adam_beta2": 0.999,
+    "optimizer_beta1": null,
+    "optimizer_beta2": null,
+    "optimizer_cpu_offload_method": null,
+    "gradient_precision": null,
+    "adam_weight_decay": 0.01,
+    "adam_epsilon": 1e-08,
+    "prodigy_steps": null,
+    "max_grad_norm": 2.0,
+    "optimizer_config": null,
+    "grad_clip_method": "value",
+    "optimizer_offload_gradients": false,
+    "fuse_optimizer": false,
+    "optimizer_release_gradients": false,
+    "push_to_hub": false,
+    "publishing_config": null,
+    "post_upload_script": null,
+    "push_checkpoints_to_hub": false,
+    "push_to_hub_background": false,
+    "hub_model_id": "quzo/fl2",
+    "model_card_private": false,
+    "model_card_safe_for_work": false,
+    "model_card_note": null,
+    "report_to": "none",
+    "custom_tracker": null,
+    "checkpoint_step_interval": 200,
+    "checkpoint_epoch_interval": null,
+    "checkpointing_rolling_steps": 0,
+    "checkpointing_use_tempdir": false,
+    "checkpoints_rolling_total_limit": 1,
+    "tracker_run_name": "simpletuner-testing",
+    "tracker_project_name": "simpletuner",
+    "tracker_image_layout": "gallery",
+    "enable_watermark": false,
+    "framerate": null,
+    "seed_for_each_device": true,
+    "snr_weight": 1.0,
+    "webhook_config": [
+        {
+            "webhook_type": "raw",
+            "callback_url": "http://0.0.0.0:8001/callback",
+            "log_level": "info",
+            "ssl_no_verify": false
+        }
+    ],
+    "webhook_reporting_interval": null,
+    "distillation_method": null,
+    "distillation_config": null,
+    "ema_validation": "comparison",
+    "local_rank": -1,
+    "ltx_train_mode": "i2v",
+    "ltx_i2v_prob": 0.1,
+    "ltx_partial_noise_fraction": 0.05,
+    "ltx_protect_first_frame": false,
+    "offload_param_path": null,
+    "offset_noise": false,
+    "quantize_activations": false,
+    "refiner_training": false,
+    "refiner_training_invert_schedule": false,
+    "refiner_training_strength": 0.2,
+    "sdxl_refiner_uses_full_range": false,
+    "sana_complex_human_instruction": "Given a user prompt, generate an 'Enhanced prompt' that provides detailed visual descriptions suitable for image generation. Evaluate the level of detail in the user prompt:\n- If the prompt is simple, focus on adding specifics about colors, shapes, sizes, textures, and spatial relationships to create vivid and concrete scenes.\n- Do not add detail that is not in the original prompt! Do not mention colors if the User Prompt does not!\n- If the prompt is already detailed, refine and enhance the existing details slightly without overcomplicating.\n- If the prompt contains explicit material, do not remove this information from the output, which would damage the performance of downstream classifiers.\nHere are examples of how to transform or refine prompts:\n- User Prompt: A cat sleeping -> Enhanced: A cat sleeping peacefully, showcasing the joy of pet ownership. Cute floof kitty cat gatto.\n- User Prompt: A busy city street -> Enhanced: A bustling city street scene featuring a crowd of people.\nPlease generate only the enhanced description for the prompt below and avoid including any additional commentary or evaluations:\nUser Prompt: ",
+    "vae_path": "black-forest-labs/FLUX.2-dev",
+    "accelerator_project_config": {
+        "project_dir": "/home/Ubuntu/.simpletuner/output",
+        "logging_dir": "/home/Ubuntu/.simpletuner/output/logs",
+        "automatic_checkpoint_naming": false,
+        "total_limit": null,
+        "iteration": 17,
+        "save_on_each_node": false
+    },
+    "process_group_kwargs": {
+        "backend": "nccl",
+        "init_method": null,
+        "timeout": "1:30:00"
+    },
+    "is_quantized": false,
+    "weight_dtype": "torch.bfloat16",
+    "disable_accelerator": false,
+    "lora_initialisation_style": true,
+    "checkpointing_steps": 200,
+    "use_fsdp": false,
+    "model_type_label": "Flux.2",
+    "use_deepspeed_optimizer": false,
+    "use_deepspeed_scheduler": false,
+    "base_weight_dtype": "torch.bfloat16",
+    "is_quanto": false,
+    "is_torchao": false,
+    "is_bnb": false,
+    "flow_matching": true,
+    "overrode_max_train_steps": false,
+    "total_num_batches": 6,
+    "num_update_steps_per_epoch": 6,
+    "total_batch_size": 2,
+    "enable_adamw_bf16": true,
+    "is_schedulefree": false,
+    "is_lr_scheduler_disabled": false,
+    "total_steps_remaining_at_start": 5000
+}

checkpoint-3400/training_state-training-images.json

@@ -0,0 +1 @@
+{"aspect_ratio_bucket_indices": {"1.0": ["/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im008.png", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im021.png"], "0.74": ["/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0014.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9743.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9827.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9935.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0003.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9796.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9890.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9865.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9260.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0090.jpg"]}, "buckets": ["0.74"], "exhausted_buckets": ["1.0"], "batch_size": 2, "current_bucket": 0, "seen_images": {"/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im021.png": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im008.png": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9935.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9890.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0003.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9743.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9827.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9260.jpg": true}, "current_epoch": 567}

checkpoint-3400/training_state.json

@@ -0,0 +1 @@
+{"global_step": 3400, "epoch_step": 3966, "epoch": 567, "exhausted_backends": [], "repeats": {"training-images": 0}}

checkpoint-3600/README.md

@@ -0,0 +1,129 @@
+---
+license: other
+base_model: "black-forest-labs/FLUX.2-dev"
+tags:
+  - flux2
+  - flux2-diffusers
+  - text-to-image
+  - image-to-image
+  - diffusers
+  - simpletuner
+  - not-for-all-audiences
+  - lora
+
+  - template:sd-lora
+  - standard
+pipeline_tag: text-to-image
+inference: true
+
+---
+
+# quzo/fl2
+
+This is a PEFT LoRA derived from [black-forest-labs/FLUX.2-dev](https://huggingface.co/black-forest-labs/FLUX.2-dev).
+
+The main validation prompt used during training was:
+```
+bm82 man
+```
+
+
+## Validation settings
+- CFG: `7.5`
+- CFG Rescale: `0.0`
+- Steps: `20`
+- Sampler: `FlowMatchEulerDiscreteScheduler`
+- Seed: `None`
+- Resolution: `1024x1024`
+
+
+Note: The validation settings are not necessarily the same as the [training settings](#training-settings).
+
+
+
+
+<Gallery />
+
+The text encoder **was not** trained.
+You may reuse the base model text encoder for inference.
+
+
+## Training settings
+
+- Training epochs: 599
+- Training steps: 3600
+- Learning rate: 0.0001
+  - Learning rate schedule: constant_with_warmup
+  - Warmup steps: 0
+- Max grad value: 2.0
+- Effective batch size: 2
+  - Micro-batch size: 2
+  - Gradient accumulation steps: 1
+  - Number of GPUs: 1
+- Gradient checkpointing: True
+- Prediction type: flow_matching[]
+- Optimizer: adamw_bf16
+- Trainable parameter precision: Pure BF16
+- Base model precision: `no_change`
+- Caption dropout probability: 0.1%
+
+
+
+- LoRA Rank: 16
+- LoRA Alpha: 16.0
+- LoRA Dropout: 0.1
+- LoRA initialisation style: default
+- LoRA mode: Standard
+    
+
+## Datasets
+
+### training-images
+- Repeats: 0
+- Total number of images: 12
+- Total number of aspect buckets: 1
+- Resolution: 1.048576 megapixels
+- Cropped: False
+- Crop style: None
+- Crop aspect: None
+- Used for regularisation data: No
+
+
+## Inference
+
+
+```python
+import torch
+from diffusers import DiffusionPipeline
+
+model_id = 'black-forest-labs/FLUX.2-dev'
+adapter_id = 'quzo/fl2'
+pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
+pipeline.load_lora_weights(adapter_id)
+
+prompt = "bm82 man"
+negative_prompt = 'blurry, cropped, ugly'
+
+## Optional: quantise the model to save on vram.
+## Note: The model was not quantised during training, so it is not necessary to quantise it during inference time.
+#from optimum.quanto import quantize, freeze, qint8
+#quantize(pipeline.transformer, weights=qint8)
+#freeze(pipeline.transformer)
+    
+pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # the pipeline is already in its target precision level
+model_output = pipeline(
+    prompt=prompt,
+    negative_prompt=negative_prompt,
+    num_inference_steps=20,
+    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(42),
+    width=1024,
+    height=1024,
+    guidance_scale=7.5,
+).images[0]
+
+model_output.save("output.png", format="PNG")
+
+```
+
+
+

checkpoint-3600/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f020ed312a541740949018b668e757324f91d58a2c4e8cf7be9b56b4b7ce1cb9
+size 321008843

checkpoint-3600/pytorch_lora_weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5daaa3075a66bf8aad8942dfd83df8683047f13a9fbdd8cbd35e3090e8789f7
+size 106979136

checkpoint-3600/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2890a0caf51b59c66e1dac8a4761f5003fbe092a9bd8d548920ac59abfae98c5
+size 14693

checkpoint-3600/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6322143ad2efc56c6c09c3066b83311a4b8987351a4fe633c0f1c92f289e9d4f
+size 1401

checkpoint-3600/simpletuner_config.json

@@ -0,0 +1,387 @@
+{
+    "model_family": "flux2",
+    "model_flavour": "dev",
+    "controlnet": false,
+    "pretrained_model_name_or_path": "black-forest-labs/FLUX.2-dev",
+    "output_dir": "/home/Ubuntu/.simpletuner/output",
+    "logging_dir": "/home/Ubuntu/.simpletuner/output/logs",
+    "model_type": "lora",
+    "seed": null,
+    "resolution": 1024,
+    "resume_from_checkpoint": null,
+    "prediction_type": null,
+    "pretrained_vae_model_name_or_path": "black-forest-labs/FLUX.2-dev",
+    "vae_dtype": "bf16",
+    "vae_cache_ondemand": false,
+    "vae_cache_disable": false,
+    "accelerator_cache_clear_interval": null,
+    "aspect_bucket_rounding": 2,
+    "base_model_precision": "no_change",
+    "text_encoder_1_precision": "no_change",
+    "text_encoder_2_precision": "no_change",
+    "text_encoder_3_precision": "no_change",
+    "text_encoder_4_precision": "no_change",
+    "gradient_checkpointing_interval": null,
+    "offload_during_startup": false,
+    "quantize_via": "accelerator",
+    "wan_force_2_1_time_embedding": false,
+    "fuse_qkv_projections": false,
+    "rescale_betas_zero_snr": false,
+    "control": false,
+    "controlnet_custom_config": null,
+    "controlnet_model_name_or_path": null,
+    "tread_config": null,
+    "pretrained_transformer_model_name_or_path": null,
+    "pretrained_transformer_subfolder": "transformer",
+    "pretrained_unet_model_name_or_path": null,
+    "pretrained_unet_subfolder": "unet",
+    "pretrained_t5_model_name_or_path": null,
+    "revision": null,
+    "variant": null,
+    "base_model_default_dtype": "bf16",
+    "unet_attention_slice": false,
+    "num_train_epochs": 834,
+    "max_train_steps": 5000,
+    "train_batch_size": 2,
+    "learning_rate": 0.0001,
+    "optimizer": "adamw_bf16",
+    "lr_scheduler": "constant_with_warmup",
+    "gradient_accumulation_steps": 1,
+    "lr_warmup_steps": 0,
+    "checkpoints_total_limit": 5,
+    "gradient_checkpointing": true,
+    "enable_group_offload": false,
+    "ramtorch": false,
+    "ramtorch_target_modules": null,
+    "ramtorch_text_encoder": false,
+    "ramtorch_vae": false,
+    "ramtorch_controlnet": false,
+    "group_offload_type": "block_level",
+    "group_offload_blocks_per_group": 1,
+    "group_offload_use_stream": false,
+    "group_offload_to_disk_path": "",
+    "group_offload_text_encoder": false,
+    "group_offload_vae": false,
+    "offload_during_save": false,
+    "enable_chunked_feed_forward": false,
+    "feed_forward_chunk_size": null,
+    "train_text_encoder": false,
+    "text_encoder_lr": null,
+    "lyrics_embedder_train": false,
+    "lyrics_embedder_optimizer": null,
+    "lyrics_embedder_lr": null,
+    "lyrics_embedder_lr_scheduler": null,
+    "lr_num_cycles": 1,
+    "lr_power": 0.8,
+    "use_soft_min_snr": false,
+    "use_ema": false,
+    "ema_device": "cpu",
+    "ema_cpu_only": false,
+    "ema_update_interval": 1,
+    "ema_foreach_disable": false,
+    "ema_decay": 0.995,
+    "lora_rank": 16,
+    "lora_alpha": 16.0,
+    "lora_type": "standard",
+    "lora_dropout": 0.1,
+    "lora_init_type": "default",
+    "peft_lora_mode": "standard",
+    "singlora_ramp_up_steps": 0,
+    "init_lora": null,
+    "lycoris_config": "/home/Ubuntu/.simpletuner/configs/config/lycoris_config.json",
+    "init_lokr_norm": null,
+    "flux_lora_target": "all",
+    "acestep_lora_target": "attn_qkv+linear_qkv",
+    "use_dora": false,
+    "resolution_type": "pixel_area",
+    "data_backend_config": "/home/Ubuntu/.simpletuner/configs/111/multidatabackend-bartek_mizak.json",
+    "caption_strategy": "filename",
+    "conditioning_multidataset_sampling": "random",
+    "instance_prompt": null,
+    "parquet_caption_column": null,
+    "parquet_filename_column": null,
+    "ignore_missing_files": false,
+    "vae_cache_scan_behaviour": "recreate",
+    "vae_enable_slicing": false,
+    "vae_enable_tiling": false,
+    "vae_batch_size": 4,
+    "max_upscale_threshold": null,
+    "caption_dropout_probability": 0.1,
+    "tokenizer_max_length": 512,
+    "audio_max_duration_seconds": null,
+    "audio_min_duration_seconds": null,
+    "audio_channels": 1,
+    "audio_duration_interval": 3.0,
+    "audio_truncation_mode": "beginning",
+    "validation_step_interval": 100,
+    "validation_epoch_interval": null,
+    "disable_benchmark": false,
+    "validation_preview": false,
+    "validation_preview_steps": 1,
+    "validation_prompt": "bm82 man",
+    "validation_lyrics": null,
+    "validation_audio_duration": 30.0,
+    "num_validation_images": 1,
+    "num_eval_images": 4,
+    "eval_steps_interval": null,
+    "eval_timesteps": 28,
+    "eval_dataset_pooling": false,
+    "evaluation_type": "none",
+    "pretrained_evaluation_model_name_or_path": "openai/clip-vit-large-patch14-336",
+    "validation_guidance": 7.5,
+    "validation_num_inference_steps": 20,
+    "validation_on_startup": false,
+    "validation_method": "simpletuner-local",
+    "validation_external_script": null,
+    "validation_external_background": false,
+    "validation_using_datasets": false,
+    "validation_torch_compile": false,
+    "validation_guidance_real": 1.0,
+    "validation_no_cfg_until_timestep": 2,
+    "validation_negative_prompt": "blurry, cropped, ugly",
+    "validation_randomize": false,
+    "validation_seed": null,
+    "validation_multigpu": "batch-parallel",
+    "validation_disable": false,
+    "validation_prompt_library": false,
+    "user_prompt_library": null,
+    "eval_dataset_id": null,
+    "validation_stitch_input_location": "left",
+    "validation_guidance_rescale": 0.0,
+    "validation_disable_unconditional": false,
+    "validation_guidance_skip_layers": null,
+    "validation_guidance_skip_layers_start": 0.01,
+    "validation_guidance_skip_layers_stop": 0.2,
+    "validation_guidance_skip_scale": 2.8,
+    "validation_lycoris_strength": 1.0,
+    "validation_noise_scheduler": null,
+    "validation_num_video_frames": null,
+    "validation_resolution": "1024x1024",
+    "validation_seed_source": "cpu",
+    "validation_torch_compile_mode": "max-autotune",
+    "validation_adapter_path": null,
+    "validation_adapter_name": null,
+    "validation_adapter_strength": 1.0,
+    "validation_adapter_mode": "adapter_only",
+    "validation_adapter_config": null,
+    "i_know_what_i_am_doing": false,
+    "flow_sigmoid_scale": 1.0,
+    "flux_fast_schedule": false,
+    "flow_use_uniform_schedule": false,
+    "flow_use_beta_schedule": false,
+    "flow_beta_schedule_alpha": 2.0,
+    "flow_beta_schedule_beta": 2.0,
+    "flow_schedule_shift": 3.0,
+    "flow_schedule_auto_shift": false,
+    "flux_guidance_mode": "constant",
+    "flux_attention_masked_training": false,
+    "flux_guidance_value": 1.0,
+    "flux_guidance_min": 0.0,
+    "flux_guidance_max": 4.0,
+    "t5_padding": "unmodified",
+    "sd3_clip_uncond_behaviour": "empty_string",
+    "sd3_t5_uncond_behaviour": null,
+    "soft_min_snr_sigma_data": null,
+    "mixed_precision": "bf16",
+    "attention_mechanism": "diffusers",
+    "sla_config": null,
+    "sageattention_usage": {
+        "_value_": "inference",
+        "_name_": "INFERENCE",
+        "__objclass__": "{'_generate_next_value_': <staticmethod(<function Enum._generate_next_value_ at 0x7649ca9cdf80>)>, '__module__': 'simpletuner.helpers.training.attention_backend', 'allows_training': <property object at 0x764872f93880>, 'allows_inference': <property object at 0x764872f93790>, 'from_raw': <classmethod(<function AttentionBackendMode.from_raw at 0x764872fe85e0>)>, '_new_member_': <built-in method __new__ of type object at 0xa234e0>, '_use_args_': True, '_member_names_': ['TRAINING', 'INFERENCE', 'TRAINING_AND_INFERENCE'], '_member_map_': {'TRAINING': <AttentionBackendMode.TRAINING: 'training'>, 'INFERENCE': <AttentionBackendMode.INFERENCE: 'inference'>, 'TRAINING_AND_INFERENCE': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>}, '_value2member_map_': {'training': <AttentionBackendMode.TRAINING: 'training'>, 'inference': <AttentionBackendMode.INFERENCE: 'inference'>, 'training+inference': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>}, '_unhashable_values_': [], '_member_type_': <class 'str'>, '_value_repr_': <slot wrapper '__repr__' of 'str' objects>, '__dict__': <attribute '__dict__' of 'AttentionBackendMode' objects>, '__weakref__': <attribute '__weakref__' of 'AttentionBackendMode' objects>, '__doc__': None, 'TRAINING': <AttentionBackendMode.TRAINING: 'training'>, 'INFERENCE': <AttentionBackendMode.INFERENCE: 'inference'>, 'TRAINING_AND_INFERENCE': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>, '__repr__': <function Enum.__repr__ at 0x7649ca9ce0c0>, '__str__': <function Enum.__str__ at 0x7649ca9ce160>, '__format__': <function Enum.__format__ at 0x7649ca9ce2a0>, '__new__': <function Enum.__new__ at 0x7649ca9cde40>}",
+        "_sort_order_": 1
+    },
+    "disable_tf32": false,
+    "set_grads_to_none": false,
+    "noise_offset": 0.1,
+    "noise_offset_probability": 0.25,
+    "input_perturbation": 0.0,
+    "input_perturbation_steps": 0,
+    "lr_end": "4e-7",
+    "lr_scale": false,
+    "lr_scale_sqrt": false,
+    "ignore_final_epochs": false,
+    "freeze_encoder_before": 12,
+    "freeze_encoder_after": 17,
+    "freeze_encoder_strategy": "after",
+    "layer_freeze_strategy": null,
+    "fully_unload_text_encoder": false,
+    "save_text_encoder": false,
+    "text_encoder_limit": 100,
+    "prepend_instance_prompt": false,
+    "only_instance_prompt": false,
+    "data_aesthetic_score": 7.0,
+    "delete_unwanted_images": false,
+    "delete_problematic_images": false,
+    "disable_bucket_pruning": false,
+    "allow_dataset_oversubscription": false,
+    "disable_segmented_timestep_sampling": false,
+    "preserve_data_backend_cache": false,
+    "override_dataset_config": false,
+    "cache_dir": "/home/Ubuntu/.simpletuner/output/cache",
+    "cache_dir_text": "cache",
+    "cache_dir_vae": "",
+    "compress_disk_cache": true,
+    "aspect_bucket_disable_rebuild": false,
+    "keep_vae_loaded": false,
+    "skip_file_discovery": "",
+    "data_backend_sampling": "auto-weighting",
+    "image_processing_batch_size": 32,
+    "write_batch_size": 128,
+    "read_batch_size": 25,
+    "enable_multiprocessing": false,
+    "accelerate_config": null,
+    "deepspeed_config": null,
+    "fsdp_enable": false,
+    "fsdp_version": 2,
+    "fsdp_reshard_after_forward": false,
+    "fsdp_state_dict_type": "SHARDED_STATE_DICT",
+    "fsdp_cpu_ram_efficient_loading": false,
+    "fsdp_auto_wrap_policy": "TRANSFORMER_BASED_WRAP",
+    "fsdp_limit_all_gathers": false,
+    "fsdp_cpu_offload": false,
+    "fsdp_activation_checkpointing": false,
+    "fsdp_transformer_layer_cls_to_wrap": null,
+    "context_parallel_size": 1,
+    "context_parallel_comm_strategy": "allgather",
+    "num_processes": 1,
+    "num_machines": 1,
+    "accelerate_extra_args": null,
+    "main_process_ip": "127.0.0.1",
+    "main_process_port": 29500,
+    "machine_rank": 0,
+    "same_network": true,
+    "dynamo_backend": "no",
+    "dynamo_mode": "",
+    "dynamo_fullgraph": false,
+    "dynamo_dynamic": false,
+    "dynamo_use_regional_compilation": false,
+    "max_workers": 32,
+    "aws_max_pool_connections": 128,
+    "torch_num_threads": 8,
+    "dataloader_prefetch": false,
+    "dataloader_prefetch_qlen": 10,
+    "aspect_bucket_worker_count": 12,
+    "aspect_bucket_alignment": 16,
+    "minimum_image_size": null,
+    "maximum_image_size": null,
+    "target_downsample_size": null,
+    "metadata_update_interval": 3600,
+    "debug_aspect_buckets": false,
+    "debug_dataset_loader": false,
+    "print_filenames": false,
+    "print_sampler_statistics": false,
+    "timestep_bias_strategy": null,
+    "timestep_bias_begin": 0,
+    "timestep_bias_end": 1000,
+    "timestep_bias_multiplier": 1.0,
+    "timestep_bias_portion": 0.25,
+    "training_scheduler_timestep_spacing": "trailing",
+    "inference_scheduler_timestep_spacing": "trailing",
+    "loss_type": "l2",
+    "huber_schedule": "snr",
+    "huber_c": 0.1,
+    "snr_gamma": null,
+    "masked_loss_probability": 1.0,
+    "hidream_use_load_balancing_loss": false,
+    "hidream_load_balancing_loss_weight": null,
+    "adam_beta1": 0.9,
+    "adam_beta2": 0.999,
+    "optimizer_beta1": null,
+    "optimizer_beta2": null,
+    "optimizer_cpu_offload_method": null,
+    "gradient_precision": null,
+    "adam_weight_decay": 0.01,
+    "adam_epsilon": 1e-08,
+    "prodigy_steps": null,
+    "max_grad_norm": 2.0,
+    "optimizer_config": null,
+    "grad_clip_method": "value",
+    "optimizer_offload_gradients": false,
+    "fuse_optimizer": false,
+    "optimizer_release_gradients": false,
+    "push_to_hub": false,
+    "publishing_config": null,
+    "post_upload_script": null,
+    "push_checkpoints_to_hub": false,
+    "push_to_hub_background": false,
+    "hub_model_id": "quzo/fl2",
+    "model_card_private": false,
+    "model_card_safe_for_work": false,
+    "model_card_note": null,
+    "report_to": "none",
+    "custom_tracker": null,
+    "checkpoint_step_interval": 200,
+    "checkpoint_epoch_interval": null,
+    "checkpointing_rolling_steps": 0,
+    "checkpointing_use_tempdir": false,
+    "checkpoints_rolling_total_limit": 1,
+    "tracker_run_name": "simpletuner-testing",
+    "tracker_project_name": "simpletuner",
+    "tracker_image_layout": "gallery",
+    "enable_watermark": false,
+    "framerate": null,
+    "seed_for_each_device": true,
+    "snr_weight": 1.0,
+    "webhook_config": [
+        {
+            "webhook_type": "raw",
+            "callback_url": "http://0.0.0.0:8001/callback",
+            "log_level": "info",
+            "ssl_no_verify": false
+        }
+    ],
+    "webhook_reporting_interval": null,
+    "distillation_method": null,
+    "distillation_config": null,
+    "ema_validation": "comparison",
+    "local_rank": -1,
+    "ltx_train_mode": "i2v",
+    "ltx_i2v_prob": 0.1,
+    "ltx_partial_noise_fraction": 0.05,
+    "ltx_protect_first_frame": false,
+    "offload_param_path": null,
+    "offset_noise": false,
+    "quantize_activations": false,
+    "refiner_training": false,
+    "refiner_training_invert_schedule": false,
+    "refiner_training_strength": 0.2,
+    "sdxl_refiner_uses_full_range": false,
+    "sana_complex_human_instruction": "Given a user prompt, generate an 'Enhanced prompt' that provides detailed visual descriptions suitable for image generation. Evaluate the level of detail in the user prompt:\n- If the prompt is simple, focus on adding specifics about colors, shapes, sizes, textures, and spatial relationships to create vivid and concrete scenes.\n- Do not add detail that is not in the original prompt! Do not mention colors if the User Prompt does not!\n- If the prompt is already detailed, refine and enhance the existing details slightly without overcomplicating.\n- If the prompt contains explicit material, do not remove this information from the output, which would damage the performance of downstream classifiers.\nHere are examples of how to transform or refine prompts:\n- User Prompt: A cat sleeping -> Enhanced: A cat sleeping peacefully, showcasing the joy of pet ownership. Cute floof kitty cat gatto.\n- User Prompt: A busy city street -> Enhanced: A bustling city street scene featuring a crowd of people.\nPlease generate only the enhanced description for the prompt below and avoid including any additional commentary or evaluations:\nUser Prompt: ",
+    "vae_path": "black-forest-labs/FLUX.2-dev",
+    "accelerator_project_config": {
+        "project_dir": "/home/Ubuntu/.simpletuner/output",
+        "logging_dir": "/home/Ubuntu/.simpletuner/output/logs",
+        "automatic_checkpoint_naming": false,
+        "total_limit": null,
+        "iteration": 18,
+        "save_on_each_node": false
+    },
+    "process_group_kwargs": {
+        "backend": "nccl",
+        "init_method": null,
+        "timeout": "1:30:00"
+    },
+    "is_quantized": false,
+    "weight_dtype": "torch.bfloat16",
+    "disable_accelerator": false,
+    "lora_initialisation_style": true,
+    "checkpointing_steps": 200,
+    "use_fsdp": false,
+    "model_type_label": "Flux.2",
+    "use_deepspeed_optimizer": false,
+    "use_deepspeed_scheduler": false,
+    "base_weight_dtype": "torch.bfloat16",
+    "is_quanto": false,
+    "is_torchao": false,
+    "is_bnb": false,
+    "flow_matching": true,
+    "overrode_max_train_steps": false,
+    "total_num_batches": 6,
+    "num_update_steps_per_epoch": 6,
+    "total_batch_size": 2,
+    "enable_adamw_bf16": true,
+    "is_schedulefree": false,
+    "is_lr_scheduler_disabled": false,
+    "total_steps_remaining_at_start": 5000
+}

checkpoint-3600/training_state-training-images.json

@@ -0,0 +1 @@
+{"aspect_ratio_bucket_indices": {"1.0": ["/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im008.png", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im021.png"], "0.74": ["/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0014.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9743.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9827.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9935.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0003.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9796.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9890.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9865.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9260.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0090.jpg"]}, "buckets": ["0.74"], "exhausted_buckets": ["1.0"], "batch_size": 2, "current_bucket": 0, "seen_images": {"/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im008.png": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im021.png": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0014.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0003.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9827.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9260.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0090.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9890.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9743.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9865.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9796.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9935.jpg": true}, "current_epoch": 600}

checkpoint-3600/training_state.json

@@ -0,0 +1 @@
+{"global_step": 3600, "epoch_step": 4199, "epoch": 600, "exhausted_backends": [], "repeats": {"training-images": 0}}

checkpoint-4000/README.md

@@ -0,0 +1,129 @@
+---
+license: other
+base_model: "black-forest-labs/FLUX.2-dev"
+tags:
+  - flux2
+  - flux2-diffusers
+  - text-to-image
+  - image-to-image
+  - diffusers
+  - simpletuner
+  - not-for-all-audiences
+  - lora
+
+  - template:sd-lora
+  - standard
+pipeline_tag: text-to-image
+inference: true
+
+---
+
+# quzo/fl2
+
+This is a PEFT LoRA derived from [black-forest-labs/FLUX.2-dev](https://huggingface.co/black-forest-labs/FLUX.2-dev).
+
+The main validation prompt used during training was:
+```
+bm82 man
+```
+
+
+## Validation settings
+- CFG: `7.5`
+- CFG Rescale: `0.0`
+- Steps: `20`
+- Sampler: `FlowMatchEulerDiscreteScheduler`
+- Seed: `None`
+- Resolution: `1024x1024`
+
+
+Note: The validation settings are not necessarily the same as the [training settings](#training-settings).
+
+
+
+
+<Gallery />
+
+The text encoder **was not** trained.
+You may reuse the base model text encoder for inference.
+
+
+## Training settings
+
+- Training epochs: 666
+- Training steps: 4000
+- Learning rate: 0.0001
+  - Learning rate schedule: constant_with_warmup
+  - Warmup steps: 0
+- Max grad value: 2.0
+- Effective batch size: 2
+  - Micro-batch size: 2
+  - Gradient accumulation steps: 1
+  - Number of GPUs: 1
+- Gradient checkpointing: True
+- Prediction type: flow_matching[]
+- Optimizer: adamw_bf16
+- Trainable parameter precision: Pure BF16
+- Base model precision: `no_change`
+- Caption dropout probability: 0.1%
+
+
+
+- LoRA Rank: 16
+- LoRA Alpha: 16.0
+- LoRA Dropout: 0.1
+- LoRA initialisation style: default
+- LoRA mode: Standard
+    
+
+## Datasets
+
+### training-images
+- Repeats: 0
+- Total number of images: 12
+- Total number of aspect buckets: 1
+- Resolution: 1.048576 megapixels
+- Cropped: False
+- Crop style: None
+- Crop aspect: None
+- Used for regularisation data: No
+
+
+## Inference
+
+
+```python
+import torch
+from diffusers import DiffusionPipeline
+
+model_id = 'black-forest-labs/FLUX.2-dev'
+adapter_id = 'quzo/fl2'
+pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
+pipeline.load_lora_weights(adapter_id)
+
+prompt = "bm82 man"
+negative_prompt = 'blurry, cropped, ugly'
+
+## Optional: quantise the model to save on vram.
+## Note: The model was not quantised during training, so it is not necessary to quantise it during inference time.
+#from optimum.quanto import quantize, freeze, qint8
+#quantize(pipeline.transformer, weights=qint8)
+#freeze(pipeline.transformer)
+    
+pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # the pipeline is already in its target precision level
+model_output = pipeline(
+    prompt=prompt,
+    negative_prompt=negative_prompt,
+    num_inference_steps=20,
+    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(42),
+    width=1024,
+    height=1024,
+    guidance_scale=7.5,
+).images[0]
+
+model_output.save("output.png", format="PNG")
+
+```
+
+
+

checkpoint-4000/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb5e08c3dd2f66e213afe2a25b1326f19804647e293077e94c7081620e01658e
+size 321008843

checkpoint-4000/pytorch_lora_weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:263d549f7e01a872a1f228806b722556cb672a1e30876f93eb1601f703299642
+size 106979136

checkpoint-4000/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:042666dd1082a430bd12fc8536141c84d36683e4390d2ea7c2f635ce006f9880
+size 14757

checkpoint-4000/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ceee27b48e7dcd059bd1aa420fb9860574384cf52a7228f1287f54609342ffe
+size 1401

checkpoint-4000/simpletuner_config.json

@@ -0,0 +1,387 @@
+{
+    "model_family": "flux2",
+    "model_flavour": "dev",
+    "controlnet": false,
+    "pretrained_model_name_or_path": "black-forest-labs/FLUX.2-dev",
+    "output_dir": "/home/Ubuntu/.simpletuner/output",
+    "logging_dir": "/home/Ubuntu/.simpletuner/output/logs",
+    "model_type": "lora",
+    "seed": null,
+    "resolution": 1024,
+    "resume_from_checkpoint": null,
+    "prediction_type": null,
+    "pretrained_vae_model_name_or_path": "black-forest-labs/FLUX.2-dev",
+    "vae_dtype": "bf16",
+    "vae_cache_ondemand": false,
+    "vae_cache_disable": false,
+    "accelerator_cache_clear_interval": null,
+    "aspect_bucket_rounding": 2,
+    "base_model_precision": "no_change",
+    "text_encoder_1_precision": "no_change",
+    "text_encoder_2_precision": "no_change",
+    "text_encoder_3_precision": "no_change",
+    "text_encoder_4_precision": "no_change",
+    "gradient_checkpointing_interval": null,
+    "offload_during_startup": false,
+    "quantize_via": "accelerator",
+    "wan_force_2_1_time_embedding": false,
+    "fuse_qkv_projections": false,
+    "rescale_betas_zero_snr": false,
+    "control": false,
+    "controlnet_custom_config": null,
+    "controlnet_model_name_or_path": null,
+    "tread_config": null,
+    "pretrained_transformer_model_name_or_path": null,
+    "pretrained_transformer_subfolder": "transformer",
+    "pretrained_unet_model_name_or_path": null,
+    "pretrained_unet_subfolder": "unet",
+    "pretrained_t5_model_name_or_path": null,
+    "revision": null,
+    "variant": null,
+    "base_model_default_dtype": "bf16",
+    "unet_attention_slice": false,
+    "num_train_epochs": 834,
+    "max_train_steps": 5000,
+    "train_batch_size": 2,
+    "learning_rate": 0.0001,
+    "optimizer": "adamw_bf16",
+    "lr_scheduler": "constant_with_warmup",
+    "gradient_accumulation_steps": 1,
+    "lr_warmup_steps": 0,
+    "checkpoints_total_limit": 5,
+    "gradient_checkpointing": true,
+    "enable_group_offload": false,
+    "ramtorch": false,
+    "ramtorch_target_modules": null,
+    "ramtorch_text_encoder": false,
+    "ramtorch_vae": false,
+    "ramtorch_controlnet": false,
+    "group_offload_type": "block_level",
+    "group_offload_blocks_per_group": 1,
+    "group_offload_use_stream": false,
+    "group_offload_to_disk_path": "",
+    "group_offload_text_encoder": false,
+    "group_offload_vae": false,
+    "offload_during_save": false,
+    "enable_chunked_feed_forward": false,
+    "feed_forward_chunk_size": null,
+    "train_text_encoder": false,
+    "text_encoder_lr": null,
+    "lyrics_embedder_train": false,
+    "lyrics_embedder_optimizer": null,
+    "lyrics_embedder_lr": null,
+    "lyrics_embedder_lr_scheduler": null,
+    "lr_num_cycles": 1,
+    "lr_power": 0.8,
+    "use_soft_min_snr": false,
+    "use_ema": false,
+    "ema_device": "cpu",
+    "ema_cpu_only": false,
+    "ema_update_interval": 1,
+    "ema_foreach_disable": false,
+    "ema_decay": 0.995,
+    "lora_rank": 16,
+    "lora_alpha": 16.0,
+    "lora_type": "standard",
+    "lora_dropout": 0.1,
+    "lora_init_type": "default",
+    "peft_lora_mode": "standard",
+    "singlora_ramp_up_steps": 0,
+    "init_lora": null,
+    "lycoris_config": "/home/Ubuntu/.simpletuner/configs/config/lycoris_config.json",
+    "init_lokr_norm": null,
+    "flux_lora_target": "all",
+    "acestep_lora_target": "attn_qkv+linear_qkv",
+    "use_dora": false,
+    "resolution_type": "pixel_area",
+    "data_backend_config": "/home/Ubuntu/.simpletuner/configs/111/multidatabackend-bartek_mizak.json",
+    "caption_strategy": "filename",
+    "conditioning_multidataset_sampling": "random",
+    "instance_prompt": null,
+    "parquet_caption_column": null,
+    "parquet_filename_column": null,
+    "ignore_missing_files": false,
+    "vae_cache_scan_behaviour": "recreate",
+    "vae_enable_slicing": false,
+    "vae_enable_tiling": false,
+    "vae_batch_size": 4,
+    "max_upscale_threshold": null,
+    "caption_dropout_probability": 0.1,
+    "tokenizer_max_length": 512,
+    "audio_max_duration_seconds": null,
+    "audio_min_duration_seconds": null,
+    "audio_channels": 1,
+    "audio_duration_interval": 3.0,
+    "audio_truncation_mode": "beginning",
+    "validation_step_interval": 100,
+    "validation_epoch_interval": null,
+    "disable_benchmark": false,
+    "validation_preview": false,
+    "validation_preview_steps": 1,
+    "validation_prompt": "bm82 man",
+    "validation_lyrics": null,
+    "validation_audio_duration": 30.0,
+    "num_validation_images": 1,
+    "num_eval_images": 4,
+    "eval_steps_interval": null,
+    "eval_timesteps": 28,
+    "eval_dataset_pooling": false,
+    "evaluation_type": "none",
+    "pretrained_evaluation_model_name_or_path": "openai/clip-vit-large-patch14-336",
+    "validation_guidance": 7.5,
+    "validation_num_inference_steps": 20,
+    "validation_on_startup": false,
+    "validation_method": "simpletuner-local",
+    "validation_external_script": null,
+    "validation_external_background": false,
+    "validation_using_datasets": false,
+    "validation_torch_compile": false,
+    "validation_guidance_real": 1.0,
+    "validation_no_cfg_until_timestep": 2,
+    "validation_negative_prompt": "blurry, cropped, ugly",
+    "validation_randomize": false,
+    "validation_seed": null,
+    "validation_multigpu": "batch-parallel",
+    "validation_disable": false,
+    "validation_prompt_library": false,
+    "user_prompt_library": null,
+    "eval_dataset_id": null,
+    "validation_stitch_input_location": "left",
+    "validation_guidance_rescale": 0.0,
+    "validation_disable_unconditional": false,
+    "validation_guidance_skip_layers": null,
+    "validation_guidance_skip_layers_start": 0.01,
+    "validation_guidance_skip_layers_stop": 0.2,
+    "validation_guidance_skip_scale": 2.8,
+    "validation_lycoris_strength": 1.0,
+    "validation_noise_scheduler": null,
+    "validation_num_video_frames": null,
+    "validation_resolution": "1024x1024",
+    "validation_seed_source": "cpu",
+    "validation_torch_compile_mode": "max-autotune",
+    "validation_adapter_path": null,
+    "validation_adapter_name": null,
+    "validation_adapter_strength": 1.0,
+    "validation_adapter_mode": "adapter_only",
+    "validation_adapter_config": null,
+    "i_know_what_i_am_doing": false,
+    "flow_sigmoid_scale": 1.0,
+    "flux_fast_schedule": false,
+    "flow_use_uniform_schedule": false,
+    "flow_use_beta_schedule": false,
+    "flow_beta_schedule_alpha": 2.0,
+    "flow_beta_schedule_beta": 2.0,
+    "flow_schedule_shift": 3.0,
+    "flow_schedule_auto_shift": false,
+    "flux_guidance_mode": "constant",
+    "flux_attention_masked_training": false,
+    "flux_guidance_value": 1.0,
+    "flux_guidance_min": 0.0,
+    "flux_guidance_max": 4.0,
+    "t5_padding": "unmodified",
+    "sd3_clip_uncond_behaviour": "empty_string",
+    "sd3_t5_uncond_behaviour": null,
+    "soft_min_snr_sigma_data": null,
+    "mixed_precision": "bf16",
+    "attention_mechanism": "diffusers",
+    "sla_config": null,
+    "sageattention_usage": {
+        "_value_": "inference",
+        "_name_": "INFERENCE",
+        "__objclass__": "{'_generate_next_value_': <staticmethod(<function Enum._generate_next_value_ at 0x7649ca9cdf80>)>, '__module__': 'simpletuner.helpers.training.attention_backend', 'allows_training': <property object at 0x764872f93880>, 'allows_inference': <property object at 0x764872f93790>, 'from_raw': <classmethod(<function AttentionBackendMode.from_raw at 0x764872fe85e0>)>, '_new_member_': <built-in method __new__ of type object at 0xa234e0>, '_use_args_': True, '_member_names_': ['TRAINING', 'INFERENCE', 'TRAINING_AND_INFERENCE'], '_member_map_': {'TRAINING': <AttentionBackendMode.TRAINING: 'training'>, 'INFERENCE': <AttentionBackendMode.INFERENCE: 'inference'>, 'TRAINING_AND_INFERENCE': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>}, '_value2member_map_': {'training': <AttentionBackendMode.TRAINING: 'training'>, 'inference': <AttentionBackendMode.INFERENCE: 'inference'>, 'training+inference': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>}, '_unhashable_values_': [], '_member_type_': <class 'str'>, '_value_repr_': <slot wrapper '__repr__' of 'str' objects>, '__dict__': <attribute '__dict__' of 'AttentionBackendMode' objects>, '__weakref__': <attribute '__weakref__' of 'AttentionBackendMode' objects>, '__doc__': None, 'TRAINING': <AttentionBackendMode.TRAINING: 'training'>, 'INFERENCE': <AttentionBackendMode.INFERENCE: 'inference'>, 'TRAINING_AND_INFERENCE': <AttentionBackendMode.TRAINING_AND_INFERENCE: 'training+inference'>, '__repr__': <function Enum.__repr__ at 0x7649ca9ce0c0>, '__str__': <function Enum.__str__ at 0x7649ca9ce160>, '__format__': <function Enum.__format__ at 0x7649ca9ce2a0>, '__new__': <function Enum.__new__ at 0x7649ca9cde40>}",
+        "_sort_order_": 1
+    },
+    "disable_tf32": false,
+    "set_grads_to_none": false,
+    "noise_offset": 0.1,
+    "noise_offset_probability": 0.25,
+    "input_perturbation": 0.0,
+    "input_perturbation_steps": 0,
+    "lr_end": "4e-7",
+    "lr_scale": false,
+    "lr_scale_sqrt": false,
+    "ignore_final_epochs": false,
+    "freeze_encoder_before": 12,
+    "freeze_encoder_after": 17,
+    "freeze_encoder_strategy": "after",
+    "layer_freeze_strategy": null,
+    "fully_unload_text_encoder": false,
+    "save_text_encoder": false,
+    "text_encoder_limit": 100,
+    "prepend_instance_prompt": false,
+    "only_instance_prompt": false,
+    "data_aesthetic_score": 7.0,
+    "delete_unwanted_images": false,
+    "delete_problematic_images": false,
+    "disable_bucket_pruning": false,
+    "allow_dataset_oversubscription": false,
+    "disable_segmented_timestep_sampling": false,
+    "preserve_data_backend_cache": false,
+    "override_dataset_config": false,
+    "cache_dir": "/home/Ubuntu/.simpletuner/output/cache",
+    "cache_dir_text": "cache",
+    "cache_dir_vae": "",
+    "compress_disk_cache": true,
+    "aspect_bucket_disable_rebuild": false,
+    "keep_vae_loaded": false,
+    "skip_file_discovery": "",
+    "data_backend_sampling": "auto-weighting",
+    "image_processing_batch_size": 32,
+    "write_batch_size": 128,
+    "read_batch_size": 25,
+    "enable_multiprocessing": false,
+    "accelerate_config": null,
+    "deepspeed_config": null,
+    "fsdp_enable": false,
+    "fsdp_version": 2,
+    "fsdp_reshard_after_forward": false,
+    "fsdp_state_dict_type": "SHARDED_STATE_DICT",
+    "fsdp_cpu_ram_efficient_loading": false,
+    "fsdp_auto_wrap_policy": "TRANSFORMER_BASED_WRAP",
+    "fsdp_limit_all_gathers": false,
+    "fsdp_cpu_offload": false,
+    "fsdp_activation_checkpointing": false,
+    "fsdp_transformer_layer_cls_to_wrap": null,
+    "context_parallel_size": 1,
+    "context_parallel_comm_strategy": "allgather",
+    "num_processes": 1,
+    "num_machines": 1,
+    "accelerate_extra_args": null,
+    "main_process_ip": "127.0.0.1",
+    "main_process_port": 29500,
+    "machine_rank": 0,
+    "same_network": true,
+    "dynamo_backend": "no",
+    "dynamo_mode": "",
+    "dynamo_fullgraph": false,
+    "dynamo_dynamic": false,
+    "dynamo_use_regional_compilation": false,
+    "max_workers": 32,
+    "aws_max_pool_connections": 128,
+    "torch_num_threads": 8,
+    "dataloader_prefetch": false,
+    "dataloader_prefetch_qlen": 10,
+    "aspect_bucket_worker_count": 12,
+    "aspect_bucket_alignment": 16,
+    "minimum_image_size": null,
+    "maximum_image_size": null,
+    "target_downsample_size": null,
+    "metadata_update_interval": 3600,
+    "debug_aspect_buckets": false,
+    "debug_dataset_loader": false,
+    "print_filenames": false,
+    "print_sampler_statistics": false,
+    "timestep_bias_strategy": null,
+    "timestep_bias_begin": 0,
+    "timestep_bias_end": 1000,
+    "timestep_bias_multiplier": 1.0,
+    "timestep_bias_portion": 0.25,
+    "training_scheduler_timestep_spacing": "trailing",
+    "inference_scheduler_timestep_spacing": "trailing",
+    "loss_type": "l2",
+    "huber_schedule": "snr",
+    "huber_c": 0.1,
+    "snr_gamma": null,
+    "masked_loss_probability": 1.0,
+    "hidream_use_load_balancing_loss": false,
+    "hidream_load_balancing_loss_weight": null,
+    "adam_beta1": 0.9,
+    "adam_beta2": 0.999,
+    "optimizer_beta1": null,
+    "optimizer_beta2": null,
+    "optimizer_cpu_offload_method": null,
+    "gradient_precision": null,
+    "adam_weight_decay": 0.01,
+    "adam_epsilon": 1e-08,
+    "prodigy_steps": null,
+    "max_grad_norm": 2.0,
+    "optimizer_config": null,
+    "grad_clip_method": "value",
+    "optimizer_offload_gradients": false,
+    "fuse_optimizer": false,
+    "optimizer_release_gradients": false,
+    "push_to_hub": false,
+    "publishing_config": null,
+    "post_upload_script": null,
+    "push_checkpoints_to_hub": false,
+    "push_to_hub_background": false,
+    "hub_model_id": "quzo/fl2",
+    "model_card_private": false,
+    "model_card_safe_for_work": false,
+    "model_card_note": null,
+    "report_to": "none",
+    "custom_tracker": null,
+    "checkpoint_step_interval": 200,
+    "checkpoint_epoch_interval": null,
+    "checkpointing_rolling_steps": 0,
+    "checkpointing_use_tempdir": false,
+    "checkpoints_rolling_total_limit": 1,
+    "tracker_run_name": "simpletuner-testing",
+    "tracker_project_name": "simpletuner",
+    "tracker_image_layout": "gallery",
+    "enable_watermark": false,
+    "framerate": null,
+    "seed_for_each_device": true,
+    "snr_weight": 1.0,
+    "webhook_config": [
+        {
+            "webhook_type": "raw",
+            "callback_url": "http://0.0.0.0:8001/callback",
+            "log_level": "info",
+            "ssl_no_verify": false
+        }
+    ],
+    "webhook_reporting_interval": null,
+    "distillation_method": null,
+    "distillation_config": null,
+    "ema_validation": "comparison",
+    "local_rank": -1,
+    "ltx_train_mode": "i2v",
+    "ltx_i2v_prob": 0.1,
+    "ltx_partial_noise_fraction": 0.05,
+    "ltx_protect_first_frame": false,
+    "offload_param_path": null,
+    "offset_noise": false,
+    "quantize_activations": false,
+    "refiner_training": false,
+    "refiner_training_invert_schedule": false,
+    "refiner_training_strength": 0.2,
+    "sdxl_refiner_uses_full_range": false,
+    "sana_complex_human_instruction": "Given a user prompt, generate an 'Enhanced prompt' that provides detailed visual descriptions suitable for image generation. Evaluate the level of detail in the user prompt:\n- If the prompt is simple, focus on adding specifics about colors, shapes, sizes, textures, and spatial relationships to create vivid and concrete scenes.\n- Do not add detail that is not in the original prompt! Do not mention colors if the User Prompt does not!\n- If the prompt is already detailed, refine and enhance the existing details slightly without overcomplicating.\n- If the prompt contains explicit material, do not remove this information from the output, which would damage the performance of downstream classifiers.\nHere are examples of how to transform or refine prompts:\n- User Prompt: A cat sleeping -> Enhanced: A cat sleeping peacefully, showcasing the joy of pet ownership. Cute floof kitty cat gatto.\n- User Prompt: A busy city street -> Enhanced: A bustling city street scene featuring a crowd of people.\nPlease generate only the enhanced description for the prompt below and avoid including any additional commentary or evaluations:\nUser Prompt: ",
+    "vae_path": "black-forest-labs/FLUX.2-dev",
+    "accelerator_project_config": {
+        "project_dir": "/home/Ubuntu/.simpletuner/output",
+        "logging_dir": "/home/Ubuntu/.simpletuner/output/logs",
+        "automatic_checkpoint_naming": false,
+        "total_limit": null,
+        "iteration": 20,
+        "save_on_each_node": false
+    },
+    "process_group_kwargs": {
+        "backend": "nccl",
+        "init_method": null,
+        "timeout": "1:30:00"
+    },
+    "is_quantized": false,
+    "weight_dtype": "torch.bfloat16",
+    "disable_accelerator": false,
+    "lora_initialisation_style": true,
+    "checkpointing_steps": 200,
+    "use_fsdp": false,
+    "model_type_label": "Flux.2",
+    "use_deepspeed_optimizer": false,
+    "use_deepspeed_scheduler": false,
+    "base_weight_dtype": "torch.bfloat16",
+    "is_quanto": false,
+    "is_torchao": false,
+    "is_bnb": false,
+    "flow_matching": true,
+    "overrode_max_train_steps": false,
+    "total_num_batches": 6,
+    "num_update_steps_per_epoch": 6,
+    "total_batch_size": 2,
+    "enable_adamw_bf16": true,
+    "is_schedulefree": false,
+    "is_lr_scheduler_disabled": false,
+    "total_steps_remaining_at_start": 5000
+}

checkpoint-4000/training_state-training-images.json

@@ -0,0 +1 @@
+{"aspect_ratio_bucket_indices": {"1.0": ["/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im008.png", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im021.png"], "0.74": ["/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0014.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9743.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9827.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9935.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0003.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9796.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9890.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9865.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9260.jpg", "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0090.jpg"]}, "buckets": ["0.74"], "exhausted_buckets": ["1.0"], "batch_size": 2, "current_bucket": 0, "seen_images": {"/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im008.png": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/im021.png": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9796.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0003.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9935.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_0014.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9827.jpg": true, "/home/Ubuntu/.simpletuner/datasets/bartek_mizak/img_9743.jpg": true}, "current_epoch": 667}

checkpoint-4000/training_state.json

@@ -0,0 +1 @@
+{"global_step": 4000, "epoch_step": 4666, "epoch": 667, "exhausted_backends": [], "repeats": {"training-images": 0}}