End of training

.gitattributes

@@ -33,3 +33,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+image_0.png filter=lfs diff=lfs merge=lfs -text
+image_1.png filter=lfs diff=lfs merge=lfs -text
+image_2.png filter=lfs diff=lfs merge=lfs -text
+image_3.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,115 @@
+---
+base_model: black-forest-labs/FLUX.2-klein-base-9B
+library_name: diffusers
+license: other
+widget:
+- text: Relight the image to remove all existing lighting conditions and replace them
+    with neutral, uniform illumination. Apply soft, evenly distributed lighting with
+    no directional shadows, no harsh highlights, and no dramatic contrast. Maintain
+    the original identity of all subjects exactly—preserve facial structure, skin
+    tone, proportions, expressions, hair, clothing, and textures. Do not alter pose,
+    camera angle, background geometry, or image composition. Lighting should appear
+    balanced, and studio-neutral, similar to diffuse overcast or a soft lightbox setup.
+    Ensure consistent exposure across the entire image with realistic depth and subtle
+    shading only where necessary for form.
+  output:
+    url: image_0.png
+- text: Relight the image to remove all existing lighting conditions and replace them
+    with neutral, uniform illumination. Apply soft, evenly distributed lighting with
+    no directional shadows, no harsh highlights, and no dramatic contrast. Maintain
+    the original identity of all subjects exactly—preserve facial structure, skin
+    tone, proportions, expressions, hair, clothing, and textures. Do not alter pose,
+    camera angle, background geometry, or image composition. Lighting should appear
+    balanced, and studio-neutral, similar to diffuse overcast or a soft lightbox setup.
+    Ensure consistent exposure across the entire image with realistic depth and subtle
+    shading only where necessary for form.
+  output:
+    url: image_1.png
+- text: Relight the image to remove all existing lighting conditions and replace them
+    with neutral, uniform illumination. Apply soft, evenly distributed lighting with
+    no directional shadows, no harsh highlights, and no dramatic contrast. Maintain
+    the original identity of all subjects exactly—preserve facial structure, skin
+    tone, proportions, expressions, hair, clothing, and textures. Do not alter pose,
+    camera angle, background geometry, or image composition. Lighting should appear
+    balanced, and studio-neutral, similar to diffuse overcast or a soft lightbox setup.
+    Ensure consistent exposure across the entire image with realistic depth and subtle
+    shading only where necessary for form.
+  output:
+    url: image_2.png
+- text: Relight the image to remove all existing lighting conditions and replace them
+    with neutral, uniform illumination. Apply soft, evenly distributed lighting with
+    no directional shadows, no harsh highlights, and no dramatic contrast. Maintain
+    the original identity of all subjects exactly—preserve facial structure, skin
+    tone, proportions, expressions, hair, clothing, and textures. Do not alter pose,
+    camera angle, background geometry, or image composition. Lighting should appear
+    balanced, and studio-neutral, similar to diffuse overcast or a soft lightbox setup.
+    Ensure consistent exposure across the entire image with realistic depth and subtle
+    shading only where necessary for form.
+  output:
+    url: image_3.png
+tags:
+- text-to-image
+- diffusers-training
+- diffusers
+- lora
+- flux2
+- flux2-diffusers
+- template:sd-lora
+---
+
+<!-- This model card has been generated automatically according to the information the training script had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+
+# Flux.2 [Klein] DreamBooth LoRA - linoyts/Flux2-Klein-Delight-LoRA
+
+<Gallery />
+
+## Model description
+
+These are linoyts/Flux2-Klein-Delight-LoRA DreamBooth LoRA weights for black-forest-labs/FLUX.2-klein-base-9B.
+
+The weights were trained using [DreamBooth](https://dreambooth.github.io/) with the [Flux2 diffusers trainer](https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/README_flux2.md).
+
+FP8 training? True
+
+## Trigger words
+
+You should use `None` to trigger the image generation.
+
+## Download model
+
+[Download the *.safetensors LoRA](linoyts/Flux2-Klein-Delight-LoRA/tree/main) in the Files & versions tab.
+
+## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)
+
+```py
+from diffusers import AutoPipelineForText2Image
+import torch
+pipeline = AutoPipelineForText2Image.from_pretrained("black-forest-labs/FLUX.2", torch_dtype=torch.bfloat16).to('cuda')
+pipeline.load_lora_weights('linoyts/Flux2-Klein-Delight-LoRA', weight_name='pytorch_lora_weights.safetensors')
+image = pipeline('Relight the image to remove all existing lighting conditions and replace them with neutral, uniform illumination. Apply soft, evenly distributed lighting with no directional shadows, no harsh highlights, and no dramatic contrast. Maintain the original identity of all subjects exactly—preserve facial structure, skin tone, proportions, expressions, hair, clothing, and textures. Do not alter pose, camera angle, background geometry, or image composition. Lighting should appear balanced, and studio-neutral, similar to diffuse overcast or a soft lightbox setup. Ensure consistent exposure across the entire image with realistic depth and subtle shading only where necessary for form.').images[0]
+```
+
+For more details, including weighting, merging and fusing LoRAs, check the [documentation on loading LoRAs in diffusers](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading_adapters)
+
+## License
+
+Please adhere to the licensing terms as described [here](https://huggingface.co/black-forest-labs/FLUX.2/blob/main/LICENSE.md).
+
+
+## Intended uses & limitations
+
+#### How to use
+
+```python
+# TODO: add an example code snippet for running this diffusion pipeline
+```
+
+#### Limitations and bias
+
+[TODO: provide examples of latent issues and potential remediations]
+
+## Training details
+
+[TODO: describe the data used to train the model]

checkpoint-1000/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc92596742672358800b3e941739de20201971a6088d318624a8c11da1c20137
+size 8597707

checkpoint-1000/pytorch_lora_weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2313b5e548f982e25c687da66bb10e0ae5ca8a0ea67a71916cc1bb521b5b2fb
+size 16786816

checkpoint-1000/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a18b5161a6d6944c269fa1aea436a45f824536cca9274fb143995f47f88fa63b
+size 14757

checkpoint-1000/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5ca5fd1f3d50dfbd88c5f854e399c1c22a842d2231d41415da16bf924479e76
+size 1401

checkpoint-500/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d74f1ca028954cea47bd7260a483c937bf3dde42e061b99b02063967ba24099
+size 8597707

checkpoint-500/pytorch_lora_weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:244a08e3d024e6136eea44c69a0f8e5a491ab01669cccab2a6ff43b2a0311432
+size 16786816

checkpoint-500/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64611be5220d7c2fa7124836fcaec65a82637df14aa59605a5ad621af52fbddc
+size 14757

checkpoint-500/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8baa45cc00e6da1eefbbbc2d50fb568432ef1e20241f0a3571b49dcb8d62f7f2
+size 1401

logs/dreambooth-flux2-image2img-lora/1769679228.1513677/events.out.tfevents.1769679228.209-20-158-47.1224629.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c309ab77a2baec61a82441cbca71b36ab2d2a25f5693c550f1485dd3ec1d73e8
+size 4144

logs/dreambooth-flux2-image2img-lora/1769679228.1548688/hparams.yml

@@ -0,0 +1,86 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+adam_epsilon: 1.0e-08
+adam_weight_decay: 0.0001
+adam_weight_decay_text_encoder: 0.001
+allow_tf32: false
+aspect_ratio_buckets: null
+bnb_quantization_config_path: null
+cache_dir: null
+cache_latents: true
+caption_column: prompt
+center_crop: false
+checkpointing_steps: 500
+checkpoints_total_limit: null
+class_data_dir: null
+cond_image_column: input_image
+dataloader_num_workers: 0
+dataset_config_name: null
+dataset_name: linoyts/flux-delight-dataset
+do_fp8_training: true
+enable_npu_flash_attention: false
+final_validation_prompt: null
+fsdp_text_encoder: false
+gradient_accumulation_steps: 1
+gradient_checkpointing: true
+guidance_scale: 3.5
+hub_model_id: null
+hub_token: null
+image_column: output_image
+instance_data_dir: null
+instance_prompt: null
+learning_rate: 0.0001
+local_rank: -1
+logging_dir: logs
+logit_mean: 0.0
+logit_std: 1.0
+lora_alpha: 4
+lora_dropout: 0.0
+lora_layers: null
+lr_num_cycles: 1
+lr_power: 1.0
+lr_scheduler: constant_with_warmup
+lr_warmup_steps: 200
+max_grad_norm: 1.0
+max_sequence_length: 512
+max_train_steps: 1000
+mixed_precision: null
+mode_scale: 1.29
+num_train_epochs: 8
+num_validation_images: 4
+offload: false
+optimizer: adamw
+output_dir: Flux2-Klein-Delight-LoRA
+pretrained_model_name_or_path: black-forest-labs/FLUX.2-klein-base-9B
+prodigy_beta3: null
+prodigy_decouple: true
+prodigy_safeguard_warmup: true
+prodigy_use_bias_correction: true
+push_to_hub: true
+random_flip: false
+rank: 16
+repeats: 1
+report_to: tensorboard
+resolution: 1024
+resume_from_checkpoint: null
+revision: null
+sample_batch_size: 4
+scale_lr: false
+seed: 0
+skip_final_inference: false
+train_batch_size: 1
+upcast_before_saving: false
+use_8bit_adam: true
+validation_epochs: 50
+validation_image: wednesday.jpg
+validation_prompt: "Relight the image to remove all existing lighting conditions and\
+  \ replace them with neutral, uniform illumination. Apply soft, evenly distributed\
+  \ lighting with no directional shadows, no harsh highlights, and no dramatic contrast.\
+  \ Maintain the original identity of all subjects exactly\u2014preserve facial structure,\
+  \ skin tone, proportions, expressions, hair, clothing, and textures. Do not alter\
+  \ pose, camera angle, background geometry, or image composition. Lighting should\
+  \ appear balanced, and studio-neutral, similar to diffuse overcast or a soft lightbox\
+  \ setup. Ensure consistent exposure across the entire image with realistic depth\
+  \ and subtle shading only where necessary for form."
+variant: null
+weighting_scheme: none

logs/dreambooth-flux2-image2img-lora/events.out.tfevents.1769679228.209-20-158-47.1224629.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c47b947f3d28cb2470cc60a0989d6c55ce5392e7d34e329fda9f7d4e510661d
+size 5090

pytorch_lora_weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2313b5e548f982e25c687da66bb10e0ae5ca8a0ea67a71916cc1bb521b5b2fb
+size 16786816