davisbrown/flux-half-illustration 🖼️🔢❓📝✓ → 🖼️

▶️ 42.1K runs 📅 Aug 2024 ⚙️ Cog 0.9.14 📄 Paper
image-inpainting image-to-image lora text-to-image

About

Flux lora, use "in the style of TOK" to trigger generation, creates half photo half illustrated elements

Example Output

Prompt:

"In the style of TOK, a photo editorial avant-garde dramatic action pose of a person wearing 90s round wacky sunglasses pulling glasses down looking forward, in Tokyo with large marble structures and bonsai trees at sunset with a vibrant illustrated jacket surrounded by illustrations of flowers, smoke, flames, ice cream, sparkles, rock and roll"

Output

Example output

Performance Metrics

20.17s Prediction Time
35.12s Total Time
All Input Parameters 
{
  "model": "dev",
  "prompt": "In the style of TOK, a photo editorial avant-garde dramatic action pose of a person wearing 90s round wacky sunglasses pulling glasses down looking forward, in Tokyo with large marble structures and bonsai trees at sunset with a vibrant illustrated jacket surrounded by illustrations of flowers, smoke, flames, ice cream, sparkles, rock and roll",
  "lora_scale": 1,
  "num_outputs": 1,
  "aspect_ratio": "3:2",
  "output_format": "webp",
  "guidance_scale": 3.5,
  "output_quality": 80,
  "num_inference_steps": 28
}
Input Parameters
mask Type: string
Image mask for image inpainting mode. If provided, aspect_ratio, width, and height inputs are ignored.
seed Type: integer
Random seed. Set for reproducible generation
image Type: string
Input image for image to image or inpainting mode. If provided, aspect_ratio, width, and height inputs are ignored.
model Default: dev
Which model to run inference with. The dev model performs best with around 28 inference steps but the schnell model only needs 4 steps.
width Type: integerRange: 256 - 1440
Width of generated image. Only works if `aspect_ratio` is set to custom. Will be rounded to nearest multiple of 16. Incompatible with fast generation
height Type: integerRange: 256 - 1440
Height of generated image. Only works if `aspect_ratio` is set to custom. Will be rounded to nearest multiple of 16. Incompatible with fast generation
prompt (required) Type: string
Prompt for generated image. If you include the `trigger_word` used in the training process you are more likely to activate the trained object, style, or concept in the resulting image.
go_fast Type: booleanDefault: false
Run faster predictions with model optimized for speed (currently fp8 quantized); disable to run in original bf16
extra_lora Type: string
Load LoRA weights. Supports Replicate models in the format <owner>/<username> or <owner>/<username>/<version>, HuggingFace URLs in the format huggingface.co/<owner>/<model-name>, CivitAI URLs in the format civitai.com/models/<id>[/<model-name>], or arbitrary .safetensors URLs from the Internet. For example, 'fofr/flux-pixar-cars'
lora_scale Type: numberDefault: 1Range: -1 - 3
Determines how strongly the main LoRA should be applied. Sane results between 0 and 1 for base inference. For go_fast we apply a 1.5x multiplier to this value; we've generally seen good performance when scaling the base value by that amount. You may still need to experiment to find the best value for your particular lora.
megapixels Default: 1
Approximate number of megapixels for generated image
num_outputs Type: integerDefault: 1Range: 1 - 4
Number of outputs to generate
aspect_ratio Default: 1:1
Aspect ratio for the generated image. If custom is selected, uses height and width below & will run in bf16 mode
output_format Default: webp
Format of the output images
guidance_scale Type: numberDefault: 3Range: 0 - 10
Guidance scale for the diffusion process. Lower values can give more realistic images. Good values to try are 2, 2.5, 3 and 3.5
output_quality Type: integerDefault: 80Range: 0 - 100
Quality when saving the output images, from 0 to 100. 100 is best quality, 0 is lowest quality. Not relevant for .png outputs
prompt_strength Type: numberDefault: 0.8Range: 0 - 1
Prompt strength when using img2img. 1.0 corresponds to full destruction of information in image
extra_lora_scale Type: numberDefault: 1Range: -1 - 3
Determines how strongly the extra LoRA should be applied. Sane results between 0 and 1 for base inference. For go_fast we apply a 1.5x multiplier to this value; we've generally seen good performance when scaling the base value by that amount. You may still need to experiment to find the best value for your particular lora.
replicate_weights Type: string
Load LoRA weights. Supports Replicate models in the format <owner>/<username> or <owner>/<username>/<version>, HuggingFace URLs in the format huggingface.co/<owner>/<model-name>, CivitAI URLs in the format civitai.com/models/<id>[/<model-name>], or arbitrary .safetensors URLs from the Internet. For example, 'fofr/flux-pixar-cars'
num_inference_steps Type: integerDefault: 28Range: 1 - 50
Number of denoising steps. More steps can give more detailed images, but take longer.
disable_safety_checker Type: booleanDefault: false
Disable safety checker for generated images.
Output Schema

Output

Type: array • Items Type: string • Items Format: uri

Example Execution Logs 
Using seed: 52561
Prompt: In the style of TOK, a photo editorial avant-garde dramatic action pose of a person wearing 90s round wacky sunglasses pulling glasses down looking forward, in Tokyo with large marble structures and bonsai trees at sunset with a vibrant illustrated jacket surrounded by illustrations of flowers, smoke, flames, ice cream, sparkles, rock and roll
txt2img mode
Using dev model
Loading LoRA weights
Ensuring enough disk space...
Free disk space: 9650639761408
Downloading weights
2024-08-17T13:43:34Z | INFO  | [ Initiating ] chunk_size=150M dest=/src/weights-cache/13e46afe547b42cc url=https://replicate.delivery/yhqm/73FW5i0TrPrZO9vTmmqWvJXVUfiP7YzCXn4TjGbUJDBRdypJA/trained_model.tar
2024-08-17T13:43:37Z | INFO  | [ Complete ] dest=/src/weights-cache/13e46afe547b42cc size="172 MB" total_elapsed=3.678s url=https://replicate.delivery/yhqm/73FW5i0TrPrZO9vTmmqWvJXVUfiP7YzCXn4TjGbUJDBRdypJA/trained_model.tar
b''
Downloaded weights in 3.7078640460968018 seconds
LoRA weights loaded successfully
  0%|          | 0/28 [00:00<?, ?it/s]
  4%|▎         | 1/28 [00:00<00:07,  3.76it/s]
  7%|▋         | 2/28 [00:00<00:05,  4.34it/s]
 11%|█         | 3/28 [00:00<00:06,  4.06it/s]
 14%|█▍        | 4/28 [00:01<00:06,  3.94it/s]
 18%|█▊        | 5/28 [00:01<00:05,  3.88it/s]
 21%|██▏       | 6/28 [00:01<00:05,  3.83it/s]
 25%|██▌       | 7/28 [00:01<00:05,  3.81it/s]
 29%|██▊       | 8/28 [00:02<00:05,  3.80it/s]
 32%|███▏      | 9/28 [00:02<00:05,  3.79it/s]
 36%|███▌      | 10/28 [00:02<00:04,  3.78it/s]
 39%|███▉      | 11/28 [00:02<00:04,  3.78it/s]
 43%|████▎     | 12/28 [00:03<00:04,  3.78it/s]
 46%|████▋     | 13/28 [00:03<00:03,  3.77it/s]
 50%|█████     | 14/28 [00:03<00:03,  3.76it/s]
 54%|█████▎    | 15/28 [00:03<00:03,  3.76it/s]
 57%|█████▋    | 16/28 [00:04<00:03,  3.77it/s]
 61%|██████    | 17/28 [00:04<00:02,  3.76it/s]
 64%|██████▍   | 18/28 [00:04<00:02,  3.76it/s]
 68%|██████▊   | 19/28 [00:04<00:02,  3.76it/s]
 71%|███████▏  | 20/28 [00:05<00:02,  3.76it/s]
 75%|███████▌  | 21/28 [00:05<00:01,  3.76it/s]
 79%|███████▊  | 22/28 [00:05<00:01,  3.76it/s]
 82%|████████▏ | 23/28 [00:06<00:01,  3.76it/s]
 86%|████████▌ | 24/28 [00:06<00:01,  3.77it/s]
 89%|████████▉ | 25/28 [00:06<00:00,  3.76it/s]
 93%|█████████▎| 26/28 [00:06<00:00,  3.76it/s]
 96%|█████████▋| 27/28 [00:07<00:00,  3.76it/s]
100%|██████████| 28/28 [00:07<00:00,  3.77it/s]
100%|██████████| 28/28 [00:07<00:00,  3.79it/s]
Version Details
Version ID
687458266007b196a490e79a77bae4b123c1792900e1cb730a51344887ad9832
Version Created
August 17, 2024
Run on Replicate →