realism images
Browse files- .gitattributes +29 -0
- .job_config.json +156 -0
- cartoon_fs.safetensors +3 -0
- cartoon_fs_000000500.safetensors +3 -0
- cartoon_fs_000001000.safetensors +3 -0
- cartoon_fs_000001500.safetensors +3 -0
- cartoon_fs_000002000.safetensors +3 -0
- cartoon_fs_000002500.safetensors +3 -0
- cartoon_fs_000003000.safetensors +3 -0
- config.yaml +162 -0
- log.txt +0 -0
- logs/0_log.txt +450 -0
- loss_log.db +3 -0
- optimizer.pt +3 -0
- pid.txt +1 -0
- samples/1771527536417__000000000_0.jpg +3 -0
- samples/1771527614382__000000000_1.jpg +3 -0
- samples/1771527699447__000000000_2.jpg +3 -0
- samples/1771527810299__000000000_3.jpg +3 -0
- samples/1771530259306__000000500_0.jpg +3 -0
- samples/1771530341964__000000500_1.jpg +3 -0
- samples/1771530431658__000000500_2.jpg +3 -0
- samples/1771530549515__000000500_3.jpg +3 -0
- samples/1771533082550__000001000_0.jpg +3 -0
- samples/1771533164969__000001000_1.jpg +3 -0
- samples/1771533254621__000001000_2.jpg +3 -0
- samples/1771533372296__000001000_3.jpg +3 -0
- samples/1771535855187__000001500_0.jpg +3 -0
- samples/1771535937534__000001500_1.jpg +3 -0
- samples/1771536026844__000001500_2.jpg +3 -0
- samples/1771536144451__000001500_3.jpg +3 -0
- samples/1771538638788__000002000_0.jpg +0 -0
- samples/1771538721118__000002000_1.jpg +3 -0
- samples/1771538810529__000002000_2.jpg +3 -0
- samples/1771538927765__000002000_3.jpg +3 -0
- samples/1771541399222__000002500_0.jpg +0 -0
- samples/1771541481365__000002500_1.jpg +3 -0
- samples/1771541570525__000002500_2.jpg +3 -0
- samples/1771541687531__000002500_3.jpg +3 -0
- samples/1771544172322__000003000_0.jpg +0 -0
- samples/1771544254495__000003000_1.jpg +3 -0
- samples/1771544343617__000003000_2.jpg +3 -0
- samples/1771544460288__000003000_3.jpg +3 -0
- samples/1771546958694__000003500_0.jpg +0 -0
- samples/1771547040776__000003500_1.jpg +3 -0
- samples/1771547129643__000003500_2.jpg +3 -0
- samples/1771547246344__000003500_3.jpg +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,32 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
loss_log.db filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
samples/1771527536417__000000000_0.jpg filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
samples/1771527614382__000000000_1.jpg filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
samples/1771527699447__000000000_2.jpg filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
samples/1771527810299__000000000_3.jpg filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
samples/1771530259306__000000500_0.jpg filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
samples/1771530341964__000000500_1.jpg filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
samples/1771530431658__000000500_2.jpg filter=lfs diff=lfs merge=lfs -text
|
| 44 |
+
samples/1771530549515__000000500_3.jpg filter=lfs diff=lfs merge=lfs -text
|
| 45 |
+
samples/1771533082550__000001000_0.jpg filter=lfs diff=lfs merge=lfs -text
|
| 46 |
+
samples/1771533164969__000001000_1.jpg filter=lfs diff=lfs merge=lfs -text
|
| 47 |
+
samples/1771533254621__000001000_2.jpg filter=lfs diff=lfs merge=lfs -text
|
| 48 |
+
samples/1771533372296__000001000_3.jpg filter=lfs diff=lfs merge=lfs -text
|
| 49 |
+
samples/1771535855187__000001500_0.jpg filter=lfs diff=lfs merge=lfs -text
|
| 50 |
+
samples/1771535937534__000001500_1.jpg filter=lfs diff=lfs merge=lfs -text
|
| 51 |
+
samples/1771536026844__000001500_2.jpg filter=lfs diff=lfs merge=lfs -text
|
| 52 |
+
samples/1771536144451__000001500_3.jpg filter=lfs diff=lfs merge=lfs -text
|
| 53 |
+
samples/1771538721118__000002000_1.jpg filter=lfs diff=lfs merge=lfs -text
|
| 54 |
+
samples/1771538810529__000002000_2.jpg filter=lfs diff=lfs merge=lfs -text
|
| 55 |
+
samples/1771538927765__000002000_3.jpg filter=lfs diff=lfs merge=lfs -text
|
| 56 |
+
samples/1771541481365__000002500_1.jpg filter=lfs diff=lfs merge=lfs -text
|
| 57 |
+
samples/1771541570525__000002500_2.jpg filter=lfs diff=lfs merge=lfs -text
|
| 58 |
+
samples/1771541687531__000002500_3.jpg filter=lfs diff=lfs merge=lfs -text
|
| 59 |
+
samples/1771544254495__000003000_1.jpg filter=lfs diff=lfs merge=lfs -text
|
| 60 |
+
samples/1771544343617__000003000_2.jpg filter=lfs diff=lfs merge=lfs -text
|
| 61 |
+
samples/1771544460288__000003000_3.jpg filter=lfs diff=lfs merge=lfs -text
|
| 62 |
+
samples/1771547040776__000003500_1.jpg filter=lfs diff=lfs merge=lfs -text
|
| 63 |
+
samples/1771547129643__000003500_2.jpg filter=lfs diff=lfs merge=lfs -text
|
| 64 |
+
samples/1771547246344__000003500_3.jpg filter=lfs diff=lfs merge=lfs -text
|
.job_config.json
ADDED
|
@@ -0,0 +1,156 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"job": "extension",
|
| 3 |
+
"config": {
|
| 4 |
+
"name": "cartoon_fs",
|
| 5 |
+
"process": [
|
| 6 |
+
{
|
| 7 |
+
"type": "diffusion_trainer",
|
| 8 |
+
"training_folder": "/home/segmindml/karthik/ai-toolkit/output",
|
| 9 |
+
"sqlite_db_path": "/home/segmindml/karthik/ai-toolkit/aitk_db.db",
|
| 10 |
+
"device": "cuda",
|
| 11 |
+
"trigger_word": null,
|
| 12 |
+
"performance_log_every": 10,
|
| 13 |
+
"network": {
|
| 14 |
+
"type": "lora",
|
| 15 |
+
"linear": 64,
|
| 16 |
+
"linear_alpha": 64,
|
| 17 |
+
"conv": 16,
|
| 18 |
+
"conv_alpha": 16,
|
| 19 |
+
"lokr_full_rank": true,
|
| 20 |
+
"lokr_factor": -1,
|
| 21 |
+
"network_kwargs": {
|
| 22 |
+
"ignore_if_contains": []
|
| 23 |
+
}
|
| 24 |
+
},
|
| 25 |
+
"save": {
|
| 26 |
+
"dtype": "bf16",
|
| 27 |
+
"save_every": 500,
|
| 28 |
+
"max_step_saves_to_keep": 40,
|
| 29 |
+
"save_format": "diffusers",
|
| 30 |
+
"push_to_hub": false
|
| 31 |
+
},
|
| 32 |
+
"datasets": [
|
| 33 |
+
{
|
| 34 |
+
"folder_path": "/home/segmindml/karthik/ai-toolkit/datasets/cn_output",
|
| 35 |
+
"mask_path": null,
|
| 36 |
+
"mask_min_value": 0.1,
|
| 37 |
+
"default_caption": "",
|
| 38 |
+
"caption_ext": "txt",
|
| 39 |
+
"caption_dropout_rate": 0.05,
|
| 40 |
+
"cache_latents_to_disk": false,
|
| 41 |
+
"is_reg": false,
|
| 42 |
+
"network_weight": 1,
|
| 43 |
+
"resolution": [
|
| 44 |
+
512,
|
| 45 |
+
768,
|
| 46 |
+
1024
|
| 47 |
+
],
|
| 48 |
+
"controls": [],
|
| 49 |
+
"shrink_video_to_frames": true,
|
| 50 |
+
"num_frames": 1,
|
| 51 |
+
"flip_x": false,
|
| 52 |
+
"flip_y": false,
|
| 53 |
+
"num_repeats": 1,
|
| 54 |
+
"control_path_1": "/home/segmindml/karthik/ai-toolkit/datasets/cn_target",
|
| 55 |
+
"control_path_2": "/home/segmindml/karthik/ai-toolkit/datasets/cn_source",
|
| 56 |
+
"control_path_3": null
|
| 57 |
+
}
|
| 58 |
+
],
|
| 59 |
+
"train": {
|
| 60 |
+
"batch_size": 1,
|
| 61 |
+
"bypass_guidance_embedding": false,
|
| 62 |
+
"steps": 3500,
|
| 63 |
+
"gradient_accumulation": 1,
|
| 64 |
+
"train_unet": true,
|
| 65 |
+
"train_text_encoder": false,
|
| 66 |
+
"gradient_checkpointing": true,
|
| 67 |
+
"noise_scheduler": "flowmatch",
|
| 68 |
+
"optimizer": "adamw8bit",
|
| 69 |
+
"timestep_type": "weighted",
|
| 70 |
+
"content_or_style": "balanced",
|
| 71 |
+
"optimizer_params": {
|
| 72 |
+
"weight_decay": 0.0001
|
| 73 |
+
},
|
| 74 |
+
"unload_text_encoder": false,
|
| 75 |
+
"cache_text_embeddings": true,
|
| 76 |
+
"lr": 0.0001,
|
| 77 |
+
"ema_config": {
|
| 78 |
+
"use_ema": false,
|
| 79 |
+
"ema_decay": 0.99
|
| 80 |
+
},
|
| 81 |
+
"skip_first_sample": false,
|
| 82 |
+
"force_first_sample": false,
|
| 83 |
+
"disable_sampling": false,
|
| 84 |
+
"dtype": "bf16",
|
| 85 |
+
"diff_output_preservation": false,
|
| 86 |
+
"diff_output_preservation_multiplier": 1,
|
| 87 |
+
"diff_output_preservation_class": "person",
|
| 88 |
+
"switch_boundary_every": 1,
|
| 89 |
+
"loss_type": "mse"
|
| 90 |
+
},
|
| 91 |
+
"logging": {
|
| 92 |
+
"log_every": 1,
|
| 93 |
+
"use_ui_logger": true
|
| 94 |
+
},
|
| 95 |
+
"model": {
|
| 96 |
+
"name_or_path": "black-forest-labs/FLUX.2-klein-base-9B",
|
| 97 |
+
"quantize": true,
|
| 98 |
+
"qtype": "qfloat8",
|
| 99 |
+
"quantize_te": true,
|
| 100 |
+
"qtype_te": "qfloat8",
|
| 101 |
+
"arch": "flux2_klein_9b",
|
| 102 |
+
"low_vram": true,
|
| 103 |
+
"model_kwargs": {
|
| 104 |
+
"match_target_res": true
|
| 105 |
+
},
|
| 106 |
+
"layer_offloading": false,
|
| 107 |
+
"layer_offloading_text_encoder_percent": 1,
|
| 108 |
+
"layer_offloading_transformer_percent": 1
|
| 109 |
+
},
|
| 110 |
+
"sample": {
|
| 111 |
+
"sampler": "flowmatch",
|
| 112 |
+
"sample_every": 500,
|
| 113 |
+
"width": 1024,
|
| 114 |
+
"height": 1024,
|
| 115 |
+
"samples": [
|
| 116 |
+
{
|
| 117 |
+
"prompt": "Edit Image 1 (target): Replace face and hair with those from Image 2 (source). Preserve source's facial features, expression, hair style, hair color, and length. Apply source's skin tone seamlessly to all visible skin regions in the target, including body if shown. Strictly maintain target's pose, body shape, clothes, lighting, background, composition, and colors (except skin tone). Ensure proper head-to-body proportions matching the target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high resolution, matching the artistic style, colors, outlines, shading, and overall aesthetic of the target image, with no distortions, no changes to non-skin elements.",
|
| 118 |
+
"ctrl_img_1": "/home/segmindml/karthik/ai-toolkit/data/images/afaffef6-6a68-42b6-9ce5-5aec078a37ca.png",
|
| 119 |
+
"width": 1152,
|
| 120 |
+
"height": 928,
|
| 121 |
+
"ctrl_img_2": "/home/segmindml/karthik/ai-toolkit/data/images/7c12ce6d-b6f0-4457-8b02-96a83bd77fe2.png"
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"prompt": "Edit Image 1 (target): Replace face and hair with those from Image 2 (source). Preserve source's facial features, expression, hair style, hair color, and length. Apply source's skin tone seamlessly to all visible skin regions in the target, including body if shown. Strictly maintain target's pose, body shape, clothes, lighting, background, composition, and colors (except skin tone). Ensure proper head-to-body proportions matching the target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high resolution, matching the artistic style, colors, outlines, shading, and overall aesthetic of the target image, with no distortions, no changes to non-skin elements.",
|
| 125 |
+
"ctrl_img_1": "/home/segmindml/karthik/ai-toolkit/data/images/d71fe06c-2566-4f4f-86e6-9c33d14f5685.png",
|
| 126 |
+
"width": 768,
|
| 127 |
+
"height": 1024,
|
| 128 |
+
"ctrl_img_2": "/home/segmindml/karthik/ai-toolkit/data/images/9adcc9ae-1077-49b3-8159-e24ca4dc3c5e.png"
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"prompt": "Edit Image 1 (target): Replace face and hair with those from Image 2 (source). Preserve source's facial features, expression, hair style, hair color, and length. Apply source's skin tone seamlessly to all visible skin regions in the target, including body if shown. Strictly maintain target's pose, body shape, clothes, lighting, background, composition, and colors (except skin tone). Ensure proper head-to-body proportions matching the target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high resolution, matching the artistic style, colors, outlines, shading, and overall aesthetic of the target image, with no distortions, no changes to non-skin elements.",
|
| 132 |
+
"ctrl_img_1": "/home/segmindml/karthik/ai-toolkit/data/images/aa1c1c71-0771-4e92-89ac-7bfaa93b530e.png",
|
| 133 |
+
"ctrl_img_2": "/home/segmindml/karthik/ai-toolkit/data/images/c76acded-0033-4b0d-a440-c84154ce8c23.png"
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"prompt": "Edit Image 1 (target): Replace face and hair with those from Image 2 (source). Preserve source's facial features, expression, hair style, hair color, and length. Apply source's skin tone seamlessly to all visible skin regions in the target, including body if shown. Strictly maintain target's pose, body shape, clothes, lighting, background, composition, and colors (except skin tone). Ensure proper head-to-body proportions matching the target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high resolution, matching the artistic style, colors, outlines, shading, and overall aesthetic of the target image, with no distortions, no changes to non-skin elements.",
|
| 137 |
+
"ctrl_img_1": "/home/segmindml/karthik/ai-toolkit/data/images/49ed4bee-ad97-49b6-a244-9a49aeb71f93.webp",
|
| 138 |
+
"ctrl_img_2": "/home/segmindml/karthik/ai-toolkit/data/images/42e389dd-788b-483f-81ec-feb26add5052.png"
|
| 139 |
+
}
|
| 140 |
+
],
|
| 141 |
+
"neg": "",
|
| 142 |
+
"seed": 42,
|
| 143 |
+
"walk_seed": true,
|
| 144 |
+
"guidance_scale": 4,
|
| 145 |
+
"sample_steps": 25,
|
| 146 |
+
"num_frames": 1,
|
| 147 |
+
"fps": 1
|
| 148 |
+
}
|
| 149 |
+
}
|
| 150 |
+
]
|
| 151 |
+
},
|
| 152 |
+
"meta": {
|
| 153 |
+
"name": "[name]",
|
| 154 |
+
"version": "1.0"
|
| 155 |
+
}
|
| 156 |
+
}
|
cartoon_fs.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:238c9591816d575ef797ee86e082ac63b47c9ba1c53e5482f41d5582d87c14f2
|
| 3 |
+
size 331379536
|
cartoon_fs_000000500.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4b58b25991ec64292379680f98df0abf26fd9ef58fea38ab0fffc6d52f56142
|
| 3 |
+
size 331379536
|
cartoon_fs_000001000.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce3a7af7eb6dd25f959f8f5a84b6acbca402fca350b0a3747c2671f5c03074c7
|
| 3 |
+
size 331379536
|
cartoon_fs_000001500.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:668578dbb7780c3bdd17a8ede1a2e2f8f39cb3f210e5e3e87c883176675bdc4d
|
| 3 |
+
size 331379536
|
cartoon_fs_000002000.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03d2f87c170972e7cc07cc69b62471d157225031953cea202437f4048c594369
|
| 3 |
+
size 331379536
|
cartoon_fs_000002500.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:845a6cddddc96a1c521bb9a9f37bc23a44c5f595cb296aefa8836ff9434b45b2
|
| 3 |
+
size 331379536
|
cartoon_fs_000003000.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee5b0f4dbdaf0e922b687a348d2643403244cf19402ce1294329fc510834e780
|
| 3 |
+
size 331379536
|
config.yaml
ADDED
|
@@ -0,0 +1,162 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
job: extension
|
| 2 |
+
config:
|
| 3 |
+
name: cartoon_fs
|
| 4 |
+
process:
|
| 5 |
+
- type: diffusion_trainer
|
| 6 |
+
training_folder: /home/segmindml/karthik/ai-toolkit/output
|
| 7 |
+
sqlite_db_path: /home/segmindml/karthik/ai-toolkit/aitk_db.db
|
| 8 |
+
device: cuda
|
| 9 |
+
trigger_word: null
|
| 10 |
+
performance_log_every: 10
|
| 11 |
+
network:
|
| 12 |
+
type: lora
|
| 13 |
+
linear: 64
|
| 14 |
+
linear_alpha: 64
|
| 15 |
+
conv: 16
|
| 16 |
+
conv_alpha: 16
|
| 17 |
+
lokr_full_rank: true
|
| 18 |
+
lokr_factor: -1
|
| 19 |
+
network_kwargs:
|
| 20 |
+
ignore_if_contains: []
|
| 21 |
+
save:
|
| 22 |
+
dtype: bf16
|
| 23 |
+
save_every: 500
|
| 24 |
+
max_step_saves_to_keep: 40
|
| 25 |
+
save_format: diffusers
|
| 26 |
+
push_to_hub: false
|
| 27 |
+
datasets:
|
| 28 |
+
- folder_path: /home/segmindml/karthik/ai-toolkit/datasets/cn_output
|
| 29 |
+
mask_path: null
|
| 30 |
+
mask_min_value: 0.1
|
| 31 |
+
default_caption: ''
|
| 32 |
+
caption_ext: txt
|
| 33 |
+
caption_dropout_rate: 0.05
|
| 34 |
+
cache_latents_to_disk: false
|
| 35 |
+
is_reg: false
|
| 36 |
+
network_weight: 1
|
| 37 |
+
resolution:
|
| 38 |
+
- 512
|
| 39 |
+
- 768
|
| 40 |
+
- 1024
|
| 41 |
+
controls: []
|
| 42 |
+
shrink_video_to_frames: true
|
| 43 |
+
num_frames: 1
|
| 44 |
+
flip_x: false
|
| 45 |
+
flip_y: false
|
| 46 |
+
num_repeats: 1
|
| 47 |
+
control_path_1: /home/segmindml/karthik/ai-toolkit/datasets/cn_target
|
| 48 |
+
control_path_2: /home/segmindml/karthik/ai-toolkit/datasets/cn_source
|
| 49 |
+
control_path_3: null
|
| 50 |
+
train:
|
| 51 |
+
batch_size: 1
|
| 52 |
+
bypass_guidance_embedding: false
|
| 53 |
+
steps: 3500
|
| 54 |
+
gradient_accumulation: 1
|
| 55 |
+
train_unet: true
|
| 56 |
+
train_text_encoder: false
|
| 57 |
+
gradient_checkpointing: true
|
| 58 |
+
noise_scheduler: flowmatch
|
| 59 |
+
optimizer: adamw8bit
|
| 60 |
+
timestep_type: weighted
|
| 61 |
+
content_or_style: balanced
|
| 62 |
+
optimizer_params:
|
| 63 |
+
weight_decay: 0.0001
|
| 64 |
+
unload_text_encoder: false
|
| 65 |
+
cache_text_embeddings: true
|
| 66 |
+
lr: 0.0001
|
| 67 |
+
ema_config:
|
| 68 |
+
use_ema: false
|
| 69 |
+
ema_decay: 0.99
|
| 70 |
+
skip_first_sample: false
|
| 71 |
+
force_first_sample: false
|
| 72 |
+
disable_sampling: false
|
| 73 |
+
dtype: bf16
|
| 74 |
+
diff_output_preservation: false
|
| 75 |
+
diff_output_preservation_multiplier: 1
|
| 76 |
+
diff_output_preservation_class: person
|
| 77 |
+
switch_boundary_every: 1
|
| 78 |
+
loss_type: mse
|
| 79 |
+
logging:
|
| 80 |
+
log_every: 1
|
| 81 |
+
use_ui_logger: true
|
| 82 |
+
model:
|
| 83 |
+
name_or_path: black-forest-labs/FLUX.2-klein-base-9B
|
| 84 |
+
quantize: true
|
| 85 |
+
qtype: qfloat8
|
| 86 |
+
quantize_te: true
|
| 87 |
+
qtype_te: qfloat8
|
| 88 |
+
arch: flux2_klein_9b
|
| 89 |
+
low_vram: true
|
| 90 |
+
model_kwargs:
|
| 91 |
+
match_target_res: true
|
| 92 |
+
layer_offloading: false
|
| 93 |
+
layer_offloading_text_encoder_percent: 1
|
| 94 |
+
layer_offloading_transformer_percent: 1
|
| 95 |
+
sample:
|
| 96 |
+
sampler: flowmatch
|
| 97 |
+
sample_every: 500
|
| 98 |
+
width: 1024
|
| 99 |
+
height: 1024
|
| 100 |
+
samples:
|
| 101 |
+
- prompt: 'Edit Image 1 (target): Replace face and hair with those from Image
|
| 102 |
+
2 (source). Preserve source''s facial features, expression, hair style,
|
| 103 |
+
hair color, and length. Apply source''s skin tone seamlessly to all visible
|
| 104 |
+
skin regions in the target, including body if shown. Strictly maintain target''s
|
| 105 |
+
pose, body shape, clothes, lighting, background, composition, and colors
|
| 106 |
+
(except skin tone). Ensure proper head-to-body proportions matching the
|
| 107 |
+
target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high
|
| 108 |
+
resolution, matching the artistic style, colors, outlines, shading, and
|
| 109 |
+
overall aesthetic of the target image, with no distortions, no changes to
|
| 110 |
+
non-skin elements.'
|
| 111 |
+
ctrl_img_1: /home/segmindml/karthik/ai-toolkit/data/images/afaffef6-6a68-42b6-9ce5-5aec078a37ca.png
|
| 112 |
+
width: 1152
|
| 113 |
+
height: 928
|
| 114 |
+
ctrl_img_2: /home/segmindml/karthik/ai-toolkit/data/images/7c12ce6d-b6f0-4457-8b02-96a83bd77fe2.png
|
| 115 |
+
- prompt: 'Edit Image 1 (target): Replace face and hair with those from Image
|
| 116 |
+
2 (source). Preserve source''s facial features, expression, hair style,
|
| 117 |
+
hair color, and length. Apply source''s skin tone seamlessly to all visible
|
| 118 |
+
skin regions in the target, including body if shown. Strictly maintain target''s
|
| 119 |
+
pose, body shape, clothes, lighting, background, composition, and colors
|
| 120 |
+
(except skin tone). Ensure proper head-to-body proportions matching the
|
| 121 |
+
target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high
|
| 122 |
+
resolution, matching the artistic style, colors, outlines, shading, and
|
| 123 |
+
overall aesthetic of the target image, with no distortions, no changes to
|
| 124 |
+
non-skin elements.'
|
| 125 |
+
ctrl_img_1: /home/segmindml/karthik/ai-toolkit/data/images/d71fe06c-2566-4f4f-86e6-9c33d14f5685.png
|
| 126 |
+
width: 768
|
| 127 |
+
height: 1024
|
| 128 |
+
ctrl_img_2: /home/segmindml/karthik/ai-toolkit/data/images/9adcc9ae-1077-49b3-8159-e24ca4dc3c5e.png
|
| 129 |
+
- prompt: 'Edit Image 1 (target): Replace face and hair with those from Image
|
| 130 |
+
2 (source). Preserve source''s facial features, expression, hair style,
|
| 131 |
+
hair color, and length. Apply source''s skin tone seamlessly to all visible
|
| 132 |
+
skin regions in the target, including body if shown. Strictly maintain target''s
|
| 133 |
+
pose, body shape, clothes, lighting, background, composition, and colors
|
| 134 |
+
(except skin tone). Ensure proper head-to-body proportions matching the
|
| 135 |
+
target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high
|
| 136 |
+
resolution, matching the artistic style, colors, outlines, shading, and
|
| 137 |
+
overall aesthetic of the target image, with no distortions, no changes to
|
| 138 |
+
non-skin elements.'
|
| 139 |
+
ctrl_img_1: /home/segmindml/karthik/ai-toolkit/data/images/aa1c1c71-0771-4e92-89ac-7bfaa93b530e.png
|
| 140 |
+
ctrl_img_2: /home/segmindml/karthik/ai-toolkit/data/images/c76acded-0033-4b0d-a440-c84154ce8c23.png
|
| 141 |
+
- prompt: 'Edit Image 1 (target): Replace face and hair with those from Image
|
| 142 |
+
2 (source). Preserve source''s facial features, expression, hair style,
|
| 143 |
+
hair color, and length. Apply source''s skin tone seamlessly to all visible
|
| 144 |
+
skin regions in the target, including body if shown. Strictly maintain target''s
|
| 145 |
+
pose, body shape, clothes, lighting, background, composition, and colors
|
| 146 |
+
(except skin tone). Ensure proper head-to-body proportions matching the
|
| 147 |
+
target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high
|
| 148 |
+
resolution, matching the artistic style, colors, outlines, shading, and
|
| 149 |
+
overall aesthetic of the target image, with no distortions, no changes to
|
| 150 |
+
non-skin elements.'
|
| 151 |
+
ctrl_img_1: /home/segmindml/karthik/ai-toolkit/data/images/49ed4bee-ad97-49b6-a244-9a49aeb71f93.webp
|
| 152 |
+
ctrl_img_2: /home/segmindml/karthik/ai-toolkit/data/images/42e389dd-788b-483f-81ec-feb26add5052.png
|
| 153 |
+
neg: ''
|
| 154 |
+
seed: 42
|
| 155 |
+
walk_seed: true
|
| 156 |
+
guidance_scale: 4
|
| 157 |
+
sample_steps: 25
|
| 158 |
+
num_frames: 1
|
| 159 |
+
fps: 1
|
| 160 |
+
meta:
|
| 161 |
+
name: cartoon_fs
|
| 162 |
+
version: '1.0'
|
log.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
logs/0_log.txt
ADDED
|
@@ -0,0 +1,450 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 0 |
0%| | 0/32 [00:00<?, ?it/s]
|
| 1 |
0%| | 0/32 [00:00<?, ?it/s]
|
| 2 |
3%|3 | 1/32 [00:00<00:04, 7.02it/s]
|
| 3 |
3%|3 | 1/32 [00:00<00:04, 7.02it/s]
|
| 4 |
9%|9 | 3/32 [00:00<00:03, 9.30it/s]
|
| 5 |
9%|9 | 3/32 [00:00<00:03, 9.30it/s]
|
| 6 |
16%|#5 | 5/32 [00:00<00:03, 8.14it/s]
|
| 7 |
16%|#5 | 5/32 [00:00<00:03, 8.14it/s]
|
| 8 |
22%|##1 | 7/32 [00:00<00:02, 8.96it/s]
|
| 9 |
22%|##1 | 7/32 [00:00<00:02, 8.96it/s]
|
| 10 |
28%|##8 | 9/32 [00:00<00:02, 10.18it/s]
|
| 11 |
28%|##8 | 9/32 [00:00<00:02, 10.18it/s]
|
| 12 |
34%|###4 | 11/32 [00:01<00:01, 11.86it/s]
|
| 13 |
34%|###4 | 11/32 [00:01<00:01, 11.86it/s]
|
| 14 |
41%|#### | 13/32 [00:01<00:01, 13.22it/s]
|
| 15 |
41%|#### | 13/32 [00:01<00:01, 13.22it/s]
|
| 16 |
47%|####6 | 15/32 [00:01<00:01, 14.26it/s]
|
| 17 |
47%|####6 | 15/32 [00:01<00:01, 14.26it/s]
|
| 18 |
53%|#####3 | 17/32 [00:01<00:00, 15.04it/s]
|
| 19 |
53%|#####3 | 17/32 [00:01<00:00, 15.04it/s]
|
| 20 |
59%|#####9 | 19/32 [00:01<00:00, 15.61it/s]
|
| 21 |
59%|#####9 | 19/32 [00:01<00:00, 15.61it/s]
|
| 22 |
66%|######5 | 21/32 [00:01<00:00, 16.04it/s]
|
| 23 |
66%|######5 | 21/32 [00:01<00:00, 16.04it/s]
|
| 24 |
72%|#######1 | 23/32 [00:01<00:00, 16.34it/s]
|
| 25 |
72%|#######1 | 23/32 [00:01<00:00, 16.34it/s]
|
| 26 |
78%|#######8 | 25/32 [00:01<00:00, 16.53it/s]
|
| 27 |
78%|#######8 | 25/32 [00:01<00:00, 16.53it/s]
|
| 28 |
84%|########4 | 27/32 [00:02<00:00, 16.64it/s]
|
| 29 |
84%|########4 | 27/32 [00:02<00:00, 16.64it/s]
|
| 30 |
91%|######### | 29/32 [00:02<00:00, 16.77it/s]
|
| 31 |
91%|######### | 29/32 [00:02<00:00, 16.77it/s]
|
| 32 |
97%|#########6| 31/32 [00:02<00:00, 16.87it/s]
|
| 33 |
97%|#########6| 31/32 [00:02<00:00, 16.87it/s]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 35 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 36 |
2%|1 | 2/125 [00:00<00:06, 17.98it/s]
|
| 37 |
2%|1 | 2/125 [00:00<00:06, 17.98it/s]
|
| 38 |
4%|4 | 5/125 [00:00<00:05, 20.29it/s]
|
| 39 |
4%|4 | 5/125 [00:00<00:05, 20.29it/s]
|
| 40 |
6%|6 | 8/125 [00:00<00:05, 19.55it/s]
|
| 41 |
6%|6 | 8/125 [00:00<00:05, 19.55it/s]
|
| 42 |
11%|#1 | 14/125 [00:00<00:03, 32.04it/s]
|
| 43 |
11%|#1 | 14/125 [00:00<00:03, 32.04it/s]
|
| 44 |
15%|#5 | 19/125 [00:00<00:03, 33.40it/s]
|
| 45 |
15%|#5 | 19/125 [00:00<00:03, 33.40it/s]
|
| 46 |
19%|#9 | 24/125 [00:00<00:02, 37.69it/s]
|
| 47 |
19%|#9 | 24/125 [00:00<00:02, 37.69it/s]
|
| 48 |
22%|##2 | 28/125 [00:01<00:03, 24.57it/s]
|
| 49 |
22%|##2 | 28/125 [00:01<00:03, 24.57it/s]
|
| 50 |
30%|### | 38/125 [00:01<00:02, 39.35it/s]
|
| 51 |
30%|### | 38/125 [00:01<00:02, 39.35it/s]
|
| 52 |
36%|###6 | 45/125 [00:01<00:01, 42.02it/s]
|
| 53 |
36%|###6 | 45/125 [00:01<00:01, 42.02it/s]
|
| 54 |
43%|####3 | 54/125 [00:01<00:01, 52.63it/s]
|
| 55 |
43%|####3 | 54/125 [00:01<00:01, 52.63it/s]
|
| 56 |
50%|##### | 63/125 [00:01<00:01, 54.80it/s]
|
| 57 |
50%|##### | 63/125 [00:01<00:01, 54.80it/s]
|
| 58 |
56%|#####6 | 70/125 [00:01<00:01, 38.81it/s]
|
| 59 |
56%|#####6 | 70/125 [00:01<00:01, 38.81it/s]
|
| 60 |
60%|###### | 75/125 [00:02<00:01, 39.09it/s]
|
| 61 |
60%|###### | 75/125 [00:02<00:01, 39.09it/s]
|
| 62 |
68%|######8 | 85/125 [00:02<00:00, 50.62it/s]
|
| 63 |
68%|######8 | 85/125 [00:02<00:00, 50.62it/s]
|
| 64 |
75%|#######5 | 94/125 [00:02<00:00, 59.26it/s]
|
| 65 |
75%|#######5 | 94/125 [00:02<00:00, 59.26it/s]
|
| 66 |
81%|######## | 101/125 [00:02<00:00, 51.47it/s]
|
| 67 |
81%|######## | 101/125 [00:02<00:00, 51.47it/s]
|
| 68 |
86%|########6 | 108/125 [00:02<00:00, 55.02it/s]
|
| 69 |
86%|########6 | 108/125 [00:02<00:00, 55.02it/s]
|
| 70 |
93%|#########2| 116/125 [00:02<00:00, 60.79it/s]
|
| 71 |
93%|#########2| 116/125 [00:02<00:00, 60.79it/s]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 73 |
0%| | 0/125 [00:00<?, ?it/s]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 75 |
0%| | 0/125 [00:00<?, ?it/s]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
|
| 77 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Running 1 job
|
| 2 |
+
{
|
| 3 |
+
"type": "diffusion_trainer",
|
| 4 |
+
"training_folder": "/home/segmindml/karthik/ai-toolkit/output",
|
| 5 |
+
"sqlite_db_path": "/home/segmindml/karthik/ai-toolkit/aitk_db.db",
|
| 6 |
+
"device": "cuda",
|
| 7 |
+
"trigger_word": null,
|
| 8 |
+
"performance_log_every": 10,
|
| 9 |
+
"network": {
|
| 10 |
+
"type": "lora",
|
| 11 |
+
"linear": 128,
|
| 12 |
+
"linear_alpha": 128,
|
| 13 |
+
"conv": 16,
|
| 14 |
+
"conv_alpha": 16,
|
| 15 |
+
"lokr_full_rank": true,
|
| 16 |
+
"lokr_factor": -1,
|
| 17 |
+
"network_kwargs": {
|
| 18 |
+
"ignore_if_contains": []
|
| 19 |
+
}
|
| 20 |
+
},
|
| 21 |
+
"save": {
|
| 22 |
+
"dtype": "bf16",
|
| 23 |
+
"save_every": 500,
|
| 24 |
+
"max_step_saves_to_keep": 40,
|
| 25 |
+
"save_format": "diffusers",
|
| 26 |
+
"push_to_hub": false
|
| 27 |
+
},
|
| 28 |
+
"datasets": [
|
| 29 |
+
{
|
| 30 |
+
"folder_path": "/home/segmindml/karthik/ai-toolkit/datasets/cn_output",
|
| 31 |
+
"mask_path": null,
|
| 32 |
+
"mask_min_value": 0.1,
|
| 33 |
+
"default_caption": "",
|
| 34 |
+
"caption_ext": "txt",
|
| 35 |
+
"caption_dropout_rate": 0.05,
|
| 36 |
+
"cache_latents_to_disk": false,
|
| 37 |
+
"is_reg": false,
|
| 38 |
+
"network_weight": 1,
|
| 39 |
+
"resolution": [
|
| 40 |
+
512,
|
| 41 |
+
768,
|
| 42 |
+
1024
|
| 43 |
+
],
|
| 44 |
+
"controls": [],
|
| 45 |
+
"shrink_video_to_frames": true,
|
| 46 |
+
"num_frames": 1,
|
| 47 |
+
"flip_x": false,
|
| 48 |
+
"flip_y": false,
|
| 49 |
+
"num_repeats": 1,
|
| 50 |
+
"control_path_1": "/home/segmindml/karthik/ai-toolkit/datasets/cn_target",
|
| 51 |
+
"control_path_2": "/home/segmindml/karthik/ai-toolkit/datasets/cn_source",
|
| 52 |
+
"control_path_3": null
|
| 53 |
+
}
|
| 54 |
+
],
|
| 55 |
+
"train": {
|
| 56 |
+
"batch_size": 1,
|
| 57 |
+
"bypass_guidance_embedding": false,
|
| 58 |
+
"steps": 3500,
|
| 59 |
+
"gradient_accumulation": 1,
|
| 60 |
+
"train_unet": true,
|
| 61 |
+
"train_text_encoder": false,
|
| 62 |
+
"gradient_checkpointing": true,
|
| 63 |
+
"noise_scheduler": "flowmatch",
|
| 64 |
+
"optimizer": "adamw8bit",
|
| 65 |
+
"timestep_type": "weighted",
|
| 66 |
+
"content_or_style": "balanced",
|
| 67 |
+
"optimizer_params": {
|
| 68 |
+
"weight_decay": 0.0001
|
| 69 |
+
},
|
| 70 |
+
"unload_text_encoder": false,
|
| 71 |
+
"cache_text_embeddings": false,
|
| 72 |
+
"lr": 0.0001,
|
| 73 |
+
"ema_config": {
|
| 74 |
+
"use_ema": false,
|
| 75 |
+
"ema_decay": 0.99
|
| 76 |
+
},
|
| 77 |
+
"skip_first_sample": false,
|
| 78 |
+
"force_first_sample": false,
|
| 79 |
+
"disable_sampling": false,
|
| 80 |
+
"dtype": "bf16",
|
| 81 |
+
"diff_output_preservation": false,
|
| 82 |
+
"diff_output_preservation_multiplier": 1,
|
| 83 |
+
"diff_output_preservation_class": "person",
|
| 84 |
+
"switch_boundary_every": 1,
|
| 85 |
+
"loss_type": "mse"
|
| 86 |
+
},
|
| 87 |
+
"logging": {
|
| 88 |
+
"log_every": 1,
|
| 89 |
+
"use_ui_logger": true
|
| 90 |
+
},
|
| 91 |
+
"model": {
|
| 92 |
+
"name_or_path": "black-forest-labs/FLUX.2-klein-base-9B",
|
| 93 |
+
"quantize": true,
|
| 94 |
+
"qtype": "qfloat8",
|
| 95 |
+
"quantize_te": true,
|
| 96 |
+
"qtype_te": "qfloat8",
|
| 97 |
+
"arch": "flux2_klein_9b",
|
| 98 |
+
"low_vram": true,
|
| 99 |
+
"model_kwargs": {
|
| 100 |
+
"match_target_res": true
|
| 101 |
+
},
|
| 102 |
+
"layer_offloading": false,
|
| 103 |
+
"layer_offloading_text_encoder_percent": 1,
|
| 104 |
+
"layer_offloading_transformer_percent": 1
|
| 105 |
+
},
|
| 106 |
+
"sample": {
|
| 107 |
+
"sampler": "flowmatch",
|
| 108 |
+
"sample_every": 500,
|
| 109 |
+
"width": 1024,
|
| 110 |
+
"height": 1024,
|
| 111 |
+
"samples": [
|
| 112 |
+
{
|
| 113 |
+
"prompt": "Edit Image 1 (target): Replace face and hair with those from Image 2 (source). Preserve source's facial features, expression, hair style, hair color, and length. Apply source's skin tone seamlessly to all visible skin regions in the target, including body if shown. Strictly maintain target's pose, body shape, clothes, lighting, background, composition, and colors (except skin tone). Ensure proper head-to-body proportions matching the target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high resolution, matching the artistic style, colors, outlines, shading, and overall aesthetic of the target image, with no distortions, no changes to non-skin elements.",
|
| 114 |
+
"ctrl_img_1": "/home/segmindml/karthik/ai-toolkit/data/images/afaffef6-6a68-42b6-9ce5-5aec078a37ca.png",
|
| 115 |
+
"width": 1152,
|
| 116 |
+
"height": 928,
|
| 117 |
+
"ctrl_img_2": "/home/segmindml/karthik/ai-toolkit/data/images/7c12ce6d-b6f0-4457-8b02-96a83bd77fe2.png"
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"prompt": "Edit Image 1 (target): Replace face and hair with those from Image 2 (source). Preserve source's facial features, expression, hair style, hair color, and length. Apply source's skin tone seamlessly to all visible skin regions in the target, including body if shown. Strictly maintain target's pose, body shape, clothes, lighting, background, composition, and colors (except skin tone). Ensure proper head-to-body proportions matching the target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high resolution, matching the artistic style, colors, outlines, shading, and overall aesthetic of the target image, with no distortions, no changes to non-skin elements.",
|
| 121 |
+
"ctrl_img_1": "/home/segmindml/karthik/ai-toolkit/data/images/d71fe06c-2566-4f4f-86e6-9c33d14f5685.png",
|
| 122 |
+
"width": 768,
|
| 123 |
+
"height": 1024,
|
| 124 |
+
"ctrl_img_2": "/home/segmindml/karthik/ai-toolkit/data/images/9adcc9ae-1077-49b3-8159-e24ca4dc3c5e.png"
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"prompt": "Edit Image 1 (target): Replace face and hair with those from Image 2 (source). Preserve source's facial features, expression, hair style, hair color, and length. Apply source's skin tone seamlessly to all visible skin regions in the target, including body if shown. Strictly maintain target's pose, body shape, clothes, lighting, background, composition, and colors (except skin tone). Ensure proper head-to-body proportions matching the target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high resolution, matching the artistic style, colors, outlines, shading, and overall aesthetic of the target image, with no distortions, no changes to non-skin elements.",
|
| 128 |
+
"ctrl_img_1": "/home/segmindml/karthik/ai-toolkit/data/images/aa1c1c71-0771-4e92-89ac-7bfaa93b530e.png",
|
| 129 |
+
"ctrl_img_2": "/home/segmindml/karthik/ai-toolkit/data/images/c76acded-0033-4b0d-a440-c84154ce8c23.png"
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"prompt": "Edit Image 1 (target): Replace face and hair with those from Image 2 (source). Preserve source's facial features, expression, hair style, hair color, and length. Apply source's skin tone seamlessly to all visible skin regions in the target, including body if shown. Strictly maintain target's pose, body shape, clothes, lighting, background, composition, and colors (except skin tone). Ensure proper head-to-body proportions matching the target; avoid enlarged, shrunken, or disproportionate heads. Output: Ultra-high resolution, matching the artistic style, colors, outlines, shading, and overall aesthetic of the target image, with no distortions, no changes to non-skin elements.",
|
| 133 |
+
"ctrl_img_1": "/home/segmindml/karthik/ai-toolkit/data/images/49ed4bee-ad97-49b6-a244-9a49aeb71f93.webp",
|
| 134 |
+
"ctrl_img_2": "/home/segmindml/karthik/ai-toolkit/data/images/42e389dd-788b-483f-81ec-feb26add5052.png"
|
| 135 |
+
}
|
| 136 |
+
],
|
| 137 |
+
"neg": "",
|
| 138 |
+
"seed": 42,
|
| 139 |
+
"walk_seed": true,
|
| 140 |
+
"guidance_scale": 4,
|
| 141 |
+
"sample_steps": 25,
|
| 142 |
+
"num_frames": 1,
|
| 143 |
+
"fps": 1
|
| 144 |
+
}
|
| 145 |
+
}
|
| 146 |
+
Using SQLite database at /home/segmindml/karthik/ai-toolkit/aitk_db.db
|
| 147 |
+
Job ID: "88cb137e-d1e7-4042-bbdf-bee20bc208c3"
|
| 148 |
+
|
| 149 |
+
#############################################
|
| 150 |
+
# Running job: cartoon_fs
|
| 151 |
+
#############################################
|
| 152 |
+
|
| 153 |
+
|
| 154 |
+
Running 1 process
|
| 155 |
+
Loading Flux2 model
|
| 156 |
+
Loading transformer
|
| 157 |
+
Quantizing Transformer
|
| 158 |
+
- quantizing 32 transformer blocks
|
| 159 |
+
|
| 160 |
0%| | 0/32 [00:00<?, ?it/s]
|
| 161 |
0%| | 0/32 [00:00<?, ?it/s]
|
| 162 |
3%|3 | 1/32 [00:00<00:04, 7.02it/s]
|
| 163 |
3%|3 | 1/32 [00:00<00:04, 7.02it/s]
|
| 164 |
9%|9 | 3/32 [00:00<00:03, 9.30it/s]
|
| 165 |
9%|9 | 3/32 [00:00<00:03, 9.30it/s]
|
| 166 |
16%|#5 | 5/32 [00:00<00:03, 8.14it/s]
|
| 167 |
16%|#5 | 5/32 [00:00<00:03, 8.14it/s]
|
| 168 |
22%|##1 | 7/32 [00:00<00:02, 8.96it/s]
|
| 169 |
22%|##1 | 7/32 [00:00<00:02, 8.96it/s]
|
| 170 |
28%|##8 | 9/32 [00:00<00:02, 10.18it/s]
|
| 171 |
28%|##8 | 9/32 [00:00<00:02, 10.18it/s]
|
| 172 |
34%|###4 | 11/32 [00:01<00:01, 11.86it/s]
|
| 173 |
34%|###4 | 11/32 [00:01<00:01, 11.86it/s]
|
| 174 |
41%|#### | 13/32 [00:01<00:01, 13.22it/s]
|
| 175 |
41%|#### | 13/32 [00:01<00:01, 13.22it/s]
|
| 176 |
47%|####6 | 15/32 [00:01<00:01, 14.26it/s]
|
| 177 |
47%|####6 | 15/32 [00:01<00:01, 14.26it/s]
|
| 178 |
53%|#####3 | 17/32 [00:01<00:00, 15.04it/s]
|
| 179 |
53%|#####3 | 17/32 [00:01<00:00, 15.04it/s]
|
| 180 |
59%|#####9 | 19/32 [00:01<00:00, 15.61it/s]
|
| 181 |
59%|#####9 | 19/32 [00:01<00:00, 15.61it/s]
|
| 182 |
66%|######5 | 21/32 [00:01<00:00, 16.04it/s]
|
| 183 |
66%|######5 | 21/32 [00:01<00:00, 16.04it/s]
|
| 184 |
72%|#######1 | 23/32 [00:01<00:00, 16.34it/s]
|
| 185 |
72%|#######1 | 23/32 [00:01<00:00, 16.34it/s]
|
| 186 |
78%|#######8 | 25/32 [00:01<00:00, 16.53it/s]
|
| 187 |
78%|#######8 | 25/32 [00:01<00:00, 16.53it/s]
|
| 188 |
84%|########4 | 27/32 [00:02<00:00, 16.64it/s]
|
| 189 |
84%|########4 | 27/32 [00:02<00:00, 16.64it/s]
|
| 190 |
91%|######### | 29/32 [00:02<00:00, 16.77it/s]
|
| 191 |
91%|######### | 29/32 [00:02<00:00, 16.77it/s]
|
| 192 |
97%|#########6| 31/32 [00:02<00:00, 16.87it/s]
|
| 193 |
97%|#########6| 31/32 [00:02<00:00, 16.87it/s]
|
| 194 |
+
|
| 195 |
+
- quantizing extras
|
| 196 |
+
Moving transformer to CPU
|
| 197 |
+
Loading Qwen3
|
| 198 |
+
`torch_dtype` is deprecated! Use `dtype` instead!
|
| 199 |
+
`torch_dtype` is deprecated! Use `dtype` instead!
|
| 200 |
+
|
| 201 |
+
|
| 202 |
+
Quantizing Qwen3
|
| 203 |
+
Loading VAE
|
| 204 |
+
Making pipe
|
| 205 |
+
Preparing Model
|
| 206 |
+
Model Loaded
|
| 207 |
+
create LoRA network. base dim (rank): 128, alpha: 128
|
| 208 |
+
neuron dropout: p=None, rank dropout: p=None, module dropout: p=None
|
| 209 |
+
apply LoRA to Conv2d with kernel size (3,3). dim (rank): 16, alpha: 16
|
| 210 |
+
create LoRA for Text Encoder: 0 modules.
|
| 211 |
+
create LoRA for U-Net: 112 modules.
|
| 212 |
+
enable LoRA for U-Net
|
| 213 |
+
Dataset: /home/segmindml/karthik/ai-toolkit/datasets/cn_output
|
| 214 |
+
- Preprocessing image dimensions
|
| 215 |
+
|
| 216 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 217 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 218 |
2%|1 | 2/125 [00:00<00:06, 17.98it/s]
|
| 219 |
2%|1 | 2/125 [00:00<00:06, 17.98it/s]
|
| 220 |
4%|4 | 5/125 [00:00<00:05, 20.29it/s]
|
| 221 |
4%|4 | 5/125 [00:00<00:05, 20.29it/s]
|
| 222 |
6%|6 | 8/125 [00:00<00:05, 19.55it/s]
|
| 223 |
6%|6 | 8/125 [00:00<00:05, 19.55it/s]
|
| 224 |
11%|#1 | 14/125 [00:00<00:03, 32.04it/s]
|
| 225 |
11%|#1 | 14/125 [00:00<00:03, 32.04it/s]
|
| 226 |
15%|#5 | 19/125 [00:00<00:03, 33.40it/s]
|
| 227 |
15%|#5 | 19/125 [00:00<00:03, 33.40it/s]
|
| 228 |
19%|#9 | 24/125 [00:00<00:02, 37.69it/s]
|
| 229 |
19%|#9 | 24/125 [00:00<00:02, 37.69it/s]
|
| 230 |
22%|##2 | 28/125 [00:01<00:03, 24.57it/s]
|
| 231 |
22%|##2 | 28/125 [00:01<00:03, 24.57it/s]
|
| 232 |
30%|### | 38/125 [00:01<00:02, 39.35it/s]
|
| 233 |
30%|### | 38/125 [00:01<00:02, 39.35it/s]
|
| 234 |
36%|###6 | 45/125 [00:01<00:01, 42.02it/s]
|
| 235 |
36%|###6 | 45/125 [00:01<00:01, 42.02it/s]
|
| 236 |
43%|####3 | 54/125 [00:01<00:01, 52.63it/s]
|
| 237 |
43%|####3 | 54/125 [00:01<00:01, 52.63it/s]
|
| 238 |
50%|##### | 63/125 [00:01<00:01, 54.80it/s]
|
| 239 |
50%|##### | 63/125 [00:01<00:01, 54.80it/s]
|
| 240 |
56%|#####6 | 70/125 [00:01<00:01, 38.81it/s]
|
| 241 |
56%|#####6 | 70/125 [00:01<00:01, 38.81it/s]
|
| 242 |
60%|###### | 75/125 [00:02<00:01, 39.09it/s]
|
| 243 |
60%|###### | 75/125 [00:02<00:01, 39.09it/s]
|
| 244 |
68%|######8 | 85/125 [00:02<00:00, 50.62it/s]
|
| 245 |
68%|######8 | 85/125 [00:02<00:00, 50.62it/s]
|
| 246 |
75%|#######5 | 94/125 [00:02<00:00, 59.26it/s]
|
| 247 |
75%|#######5 | 94/125 [00:02<00:00, 59.26it/s]
|
| 248 |
81%|######## | 101/125 [00:02<00:00, 51.47it/s]
|
| 249 |
81%|######## | 101/125 [00:02<00:00, 51.47it/s]
|
| 250 |
86%|########6 | 108/125 [00:02<00:00, 55.02it/s]
|
| 251 |
86%|########6 | 108/125 [00:02<00:00, 55.02it/s]
|
| 252 |
93%|#########2| 116/125 [00:02<00:00, 60.79it/s]
|
| 253 |
93%|#########2| 116/125 [00:02<00:00, 60.79it/s]
|
| 254 |
+
|
| 255 |
+
- Found 125 images
|
| 256 |
+
Bucket sizes for /home/segmindml/karthik/ai-toolkit/datasets/cn_output:
|
| 257 |
+
512x512: 80 files
|
| 258 |
+
736x352: 2 files
|
| 259 |
+
704x352: 5 files
|
| 260 |
+
416x576: 3 files
|
| 261 |
+
672x384: 1 files
|
| 262 |
+
400x400: 3 files
|
| 263 |
+
480x512: 6 files
|
| 264 |
+
384x416: 1 files
|
| 265 |
+
400x448: 1 files
|
| 266 |
+
768x320: 3 files
|
| 267 |
+
512x480: 3 files
|
| 268 |
+
496x496: 1 files
|
| 269 |
+
448x576: 1 files
|
| 270 |
+
448x448: 1 files
|
| 271 |
+
304x288: 1 files
|
| 272 |
+
416x608: 1 files
|
| 273 |
+
544x448: 1 files
|
| 274 |
+
192x192: 1 files
|
| 275 |
+
208x208: 2 files
|
| 276 |
+
336x416: 1 files
|
| 277 |
+
432x496: 1 files
|
| 278 |
+
464x432: 2 files
|
| 279 |
+
416x496: 1 files
|
| 280 |
+
176x176: 1 files
|
| 281 |
+
448x544: 1 files
|
| 282 |
+
384x352: 1 files
|
| 283 |
+
26 buckets made
|
| 284 |
+
Dataset: /home/segmindml/karthik/ai-toolkit/datasets/cn_output
|
| 285 |
+
- Preprocessing image dimensions
|
| 286 |
+
|
| 287 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 288 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 289 |
+
|
| 290 |
+
- Found 125 images
|
| 291 |
+
Bucket sizes for /home/segmindml/karthik/ai-toolkit/datasets/cn_output:
|
| 292 |
+
768x768: 48 files
|
| 293 |
+
1104x528: 2 files
|
| 294 |
+
1056x528: 5 files
|
| 295 |
+
624x864: 3 files
|
| 296 |
+
672x672: 3 files
|
| 297 |
+
1008x576: 1 files
|
| 298 |
+
400x400: 3 files
|
| 299 |
+
720x768: 4 files
|
| 300 |
+
656x656: 5 files
|
| 301 |
+
704x704: 10 files
|
| 302 |
+
384x416: 1 files
|
| 303 |
+
592x592: 3 files
|
| 304 |
+
400x448: 1 files
|
| 305 |
+
1152x480: 3 files
|
| 306 |
+
768x720: 2 files
|
| 307 |
+
496x496: 1 files
|
| 308 |
+
608x784: 1 files
|
| 309 |
+
688x688: 7 files
|
| 310 |
+
672x720: 1 files
|
| 311 |
+
448x448: 1 files
|
| 312 |
+
512x512: 1 files
|
| 313 |
+
304x288: 1 files
|
| 314 |
+
560x816: 1 files
|
| 315 |
+
816x672: 1 files
|
| 316 |
+
192x192: 1 files
|
| 317 |
+
640x640: 2 files
|
| 318 |
+
208x208: 2 files
|
| 319 |
+
336x416: 1 files
|
| 320 |
+
432x496: 1 files
|
| 321 |
+
464x432: 2 files
|
| 322 |
+
416x496: 1 files
|
| 323 |
+
688x656: 1 files
|
| 324 |
+
544x576: 1 files
|
| 325 |
+
624x624: 1 files
|
| 326 |
+
176x176: 1 files
|
| 327 |
+
640x784: 1 files
|
| 328 |
+
384x352: 1 files
|
| 329 |
+
37 buckets made
|
| 330 |
+
Dataset: /home/segmindml/karthik/ai-toolkit/datasets/cn_output
|
| 331 |
+
- Preprocessing image dimensions
|
| 332 |
+
|
| 333 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 334 |
0%| | 0/125 [00:00<?, ?it/s]
|
| 335 |
+
|
| 336 |
+
- Found 125 images
|
| 337 |
+
Bucket sizes for /home/segmindml/karthik/ai-toolkit/datasets/cn_output:
|
| 338 |
+
1024x1024: 39 files
|
| 339 |
+
1264x608: 2 files
|
| 340 |
+
1360x672: 4 files
|
| 341 |
+
768x992: 1 files
|
| 342 |
+
672x672: 3 files
|
| 343 |
+
1344x768: 1 files
|
| 344 |
+
400x400: 3 files
|
| 345 |
+
928x992: 1 files
|
| 346 |
+
656x656: 5 files
|
| 347 |
+
704x704: 10 files
|
| 348 |
+
1408x704: 1 files
|
| 349 |
+
1008x1008: 3 files
|
| 350 |
+
384x416: 1 files
|
| 351 |
+
592x592: 3 files
|
| 352 |
+
400x448: 1 files
|
| 353 |
+
1520x624: 3 files
|
| 354 |
+
1024x960: 2 files
|
| 355 |
+
496x496: 1 files
|
| 356 |
+
608x784: 1 files
|
| 357 |
+
688x688: 7 files
|
| 358 |
+
672x720: 1 files
|
| 359 |
+
448x448: 1 files
|
| 360 |
+
512x512: 1 files
|
| 361 |
+
304x288: 1 files
|
| 362 |
+
560x816: 1 files
|
| 363 |
+
992x992: 4 files
|
| 364 |
+
1088x896: 1 files
|
| 365 |
+
960x1024: 1 files
|
| 366 |
+
192x192: 1 files
|
| 367 |
+
640x640: 2 files
|
| 368 |
+
208x208: 2 files
|
| 369 |
+
336x416: 1 files
|
| 370 |
+
832x1152: 2 files
|
| 371 |
+
432x496: 1 files
|
| 372 |
+
464x432: 2 files
|
| 373 |
+
416x496: 1 files
|
| 374 |
+
768x816: 1 files
|
| 375 |
+
688x656: 1 files
|
| 376 |
+
960x960: 1 files
|
| 377 |
+
544x576: 1 files
|
| 378 |
+
624x624: 1 files
|
| 379 |
+
176x176: 1 files
|
| 380 |
+
640x784: 1 files
|
| 381 |
+
848x848: 1 files
|
| 382 |
+
384x352: 1 files
|
| 383 |
+
928x928: 1 files
|
| 384 |
+
46 buckets made
|
| 385 |
+
Generating baseline samples before training
|
| 386 |
+
|
| 387 |
|
| 388 |
|
| 389 |
+
|
| 390 |
+
========================================
|
| 391 |
+
Result:
|
| 392 |
+
- 0 completed jobs
|
| 393 |
+
- 1 failure
|
| 394 |
+
========================================
|
| 395 |
+
Traceback (most recent call last):
|
| 396 |
+
Traceback (most recent call last):
|
| 397 |
+
File "/home/segmindml/karthik/ai-toolkit/run.py", line 120, in <module>
|
| 398 |
+
File "/home/segmindml/karthik/ai-toolkit/run.py", line 120, in <module>
|
| 399 |
+
main()main()
|
| 400 |
+
|
| 401 |
+
File "/home/segmindml/karthik/ai-toolkit/run.py", line 108, in main
|
| 402 |
+
File "/home/segmindml/karthik/ai-toolkit/run.py", line 108, in main
|
| 403 |
+
raise eraise e
|
| 404 |
+
|
| 405 |
+
File "/home/segmindml/karthik/ai-toolkit/run.py", line 96, in main
|
| 406 |
+
File "/home/segmindml/karthik/ai-toolkit/run.py", line 96, in main
|
| 407 |
+
job.run()job.run()
|
| 408 |
+
|
| 409 |
+
File "/home/segmindml/karthik/ai-toolkit/jobs/ExtensionJob.py", line 22, in run
|
| 410 |
+
File "/home/segmindml/karthik/ai-toolkit/jobs/ExtensionJob.py", line 22, in run
|
| 411 |
+
process.run()process.run()
|
| 412 |
+
|
| 413 |
+
File "/home/segmindml/karthik/ai-toolkit/jobs/process/BaseSDTrainProcess.py", line 2039, in run
|
| 414 |
+
File "/home/segmindml/karthik/ai-toolkit/jobs/process/BaseSDTrainProcess.py", line 2039, in run
|
| 415 |
+
self.sample(self.step_num)self.sample(self.step_num)
|
| 416 |
+
|
| 417 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/sd_trainer/DiffusionTrainer.py", line 306, in sample
|
| 418 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/sd_trainer/DiffusionTrainer.py", line 306, in sample
|
| 419 |
+
super().sample(step, is_first)super().sample(step, is_first)
|
| 420 |
+
|
| 421 |
+
File "/home/segmindml/karthik/ai-toolkit/jobs/process/BaseSDTrainProcess.py", line 368, in sample
|
| 422 |
+
File "/home/segmindml/karthik/ai-toolkit/jobs/process/BaseSDTrainProcess.py", line 368, in sample
|
| 423 |
+
self.sd.generate_images(gen_img_config_list, sampler=sample_config.sampler)self.sd.generate_images(gen_img_config_list, sampler=sample_config.sampler)
|
| 424 |
+
|
| 425 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
|
| 426 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
|
| 427 |
+
return func(*args, **kwargs)return func(*args, **kwargs)
|
| 428 |
+
|
| 429 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 430 |
+
|
| 431 |
+
File "/home/segmindml/karthik/ai-toolkit/toolkit/models/base_model.py", line 654, in generate_images
|
| 432 |
+
File "/home/segmindml/karthik/ai-toolkit/toolkit/models/base_model.py", line 654, in generate_images
|
| 433 |
+
img = self.generate_single_image(img = self.generate_single_image(
|
| 434 |
+
|
| 435 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 436 |
+
|
| 437 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/diffusion_models/flux2/flux2_model.py", line 305, in generate_single_image
|
| 438 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/diffusion_models/flux2/flux2_model.py", line 305, in generate_single_image
|
| 439 |
+
img = pipeline(img = pipeline(
|
| 440 |
+
|
| 441 |
+
^^^^^^^^^^^^^^^^^^
|
| 442 |
+
|
| 443 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
|
| 444 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
|
| 445 |
+
return func(*args, **kwargs)return func(*args, **kwargs)
|
| 446 |
+
|
| 447 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 448 |
+
|
| 449 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/diffusion_models/flux2/src/pipeline.py", line 406, in __call__
|
| 450 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/diffusion_models/flux2/src/pipeline.py", line 406, in __call__
|
| 451 |
+
pred = self.transformer(pred = self.transformer(
|
| 452 |
+
|
| 453 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 454 |
+
|
| 455 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
|
| 456 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
|
| 457 |
+
return self._call_impl(*args, **kwargs)return self._call_impl(*args, **kwargs)
|
| 458 |
+
|
| 459 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 460 |
+
|
| 461 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
|
| 462 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
|
| 463 |
+
return forward_call(*args, **kwargs)return forward_call(*args, **kwargs)
|
| 464 |
+
|
| 465 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 466 |
+
|
| 467 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/diffusion_models/flux2/src/model.py", line 184, in forward
|
| 468 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/diffusion_models/flux2/src/model.py", line 184, in forward
|
| 469 |
+
img, txt = block(img, txt = block(
|
| 470 |
+
|
| 471 |
+
^^^^^^^^^^^^
|
| 472 |
+
|
| 473 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
|
| 474 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
|
| 475 |
+
return self._call_impl(*args, **kwargs)return self._call_impl(*args, **kwargs)
|
| 476 |
+
|
| 477 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 478 |
+
|
| 479 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
|
| 480 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
|
| 481 |
+
return forward_call(*args, **kwargs)return forward_call(*args, **kwargs)
|
| 482 |
+
|
| 483 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 484 |
+
|
| 485 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/diffusion_models/flux2/src/model.py", line 434, in forward
|
| 486 |
+
File "/home/segmindml/karthik/ai-toolkit/extensions_built_in/diffusion_models/flux2/src/model.py", line 434, in forward
|
| 487 |
+
img = img + img_mod2_gate * self.img_mlp(img = img + img_mod2_gate * self.img_mlp(
|
| 488 |
+
|
| 489 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 490 |
+
|
| 491 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
|
| 492 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
|
| 493 |
+
return self._call_impl(*args, **kwargs)return self._call_impl(*args, **kwargs)
|
| 494 |
+
|
| 495 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 496 |
+
|
| 497 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
|
| 498 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
|
| 499 |
+
return forward_call(*args, **kwargs)return forward_call(*args, **kwargs)
|
| 500 |
+
|
| 501 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 502 |
+
|
| 503 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/container.py", line 244, in forward
|
| 504 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/container.py", line 244, in forward
|
| 505 |
+
input = module(input)input = module(input)
|
| 506 |
+
|
| 507 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 508 |
+
|
| 509 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
|
| 510 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
|
| 511 |
+
return self._call_impl(*args, **kwargs)return self._call_impl(*args, **kwargs)
|
| 512 |
+
|
| 513 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 514 |
+
|
| 515 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
|
| 516 |
+
File "/home/segmindml/miniconda3/envs/at/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
|
| 517 |
+
return forward_call(*args, **kwargs)return forward_call(*args, **kwargs)
|
| 518 |
+
|
| 519 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 520 |
+
|
| 521 |
+
File "/home/segmindml/karthik/ai-toolkit/toolkit/network_mixins.py", line 305, in forward
|
| 522 |
+
File "/home/segmindml/karthik/ai-toolkit/toolkit/network_mixins.py", line 305, in forward
|
| 523 |
+
scaled_lora_output = scaled_lora_output.to(org_forwarded.dtype)scaled_lora_output = scaled_lora_output.to(org_forwarded.dtype)
|
| 524 |
+
|
| 525 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 526 |
+
|
| 527 |
+
torchtorch..OutOfMemoryErrorOutOfMemoryError: : CUDA out of memory. Tried to allocate 416.00 MiB. GPU 0 has a total capacity of 23.54 GiB of which 218.88 MiB is free. Including non-PyTorch memory, this process has 22.96 GiB memory in use. Of the allocated memory 21.27 GiB is allocated by PyTorch, and 1.23 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)CUDA out of memory. Tried to allocate 416.00 MiB. GPU 0 has a total capacity of 23.54 GiB of which 218.88 MiB is free. Including non-PyTorch memory, this process has 22.96 GiB memory in use. Of the allocated memory 21.27 GiB is allocated by PyTorch, and 1.23 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)
|
| 528 |
+
|
loss_log.db
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca30cb12e2567574ed7aaf2cb716c5c0af69f927f7977b1e6d1490e3ee3cf990
|
| 3 |
+
size 655360
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0c522e1283daf5f42a9c2422a4f0a459805962f38c968150c32ff63ec0ecab0
|
| 3 |
+
size 336791947
|
pid.txt
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
13069
|
samples/1771527536417__000000000_0.jpg
ADDED
|
Git LFS Details
|
samples/1771527614382__000000000_1.jpg
ADDED
|
Git LFS Details
|
samples/1771527699447__000000000_2.jpg
ADDED
|
Git LFS Details
|
samples/1771527810299__000000000_3.jpg
ADDED
|
Git LFS Details
|
samples/1771530259306__000000500_0.jpg
ADDED
|
Git LFS Details
|
samples/1771530341964__000000500_1.jpg
ADDED
|
Git LFS Details
|
samples/1771530431658__000000500_2.jpg
ADDED
|
Git LFS Details
|
samples/1771530549515__000000500_3.jpg
ADDED
|
Git LFS Details
|
samples/1771533082550__000001000_0.jpg
ADDED
|
Git LFS Details
|
samples/1771533164969__000001000_1.jpg
ADDED
|
Git LFS Details
|
samples/1771533254621__000001000_2.jpg
ADDED
|
Git LFS Details
|
samples/1771533372296__000001000_3.jpg
ADDED
|
Git LFS Details
|
samples/1771535855187__000001500_0.jpg
ADDED
|
Git LFS Details
|
samples/1771535937534__000001500_1.jpg
ADDED
|
Git LFS Details
|
samples/1771536026844__000001500_2.jpg
ADDED
|
Git LFS Details
|
samples/1771536144451__000001500_3.jpg
ADDED
|
Git LFS Details
|
samples/1771538638788__000002000_0.jpg
ADDED
|
samples/1771538721118__000002000_1.jpg
ADDED
|
Git LFS Details
|
samples/1771538810529__000002000_2.jpg
ADDED
|
Git LFS Details
|
samples/1771538927765__000002000_3.jpg
ADDED
|
Git LFS Details
|
samples/1771541399222__000002500_0.jpg
ADDED
|
samples/1771541481365__000002500_1.jpg
ADDED
|
Git LFS Details
|
samples/1771541570525__000002500_2.jpg
ADDED
|
Git LFS Details
|
samples/1771541687531__000002500_3.jpg
ADDED
|
Git LFS Details
|
samples/1771544172322__000003000_0.jpg
ADDED
|
samples/1771544254495__000003000_1.jpg
ADDED
|
Git LFS Details
|
samples/1771544343617__000003000_2.jpg
ADDED
|
Git LFS Details
|
samples/1771544460288__000003000_3.jpg
ADDED
|
Git LFS Details
|
samples/1771546958694__000003500_0.jpg
ADDED
|
samples/1771547040776__000003500_1.jpg
ADDED
|
Git LFS Details
|
samples/1771547129643__000003500_2.jpg
ADDED
|
Git LFS Details
|
samples/1771547246344__000003500_3.jpg
ADDED
|
Git LFS Details
|