laion-ai/puck 🖼️🔢❓📝✓ → 🖼️

▶️ 4.9K runs 📅 Jun 2022 ⚙️ Cog 0.3.13 🔗 GitHub ⚖️ License
image-inpainting pixel-art text-to-image

About

Generate retro videogame art using text.

Example Output

Prompt:

"super mario world level 1-1"

Output

https://replicate.delivery/mgxm/6358008e-3e8c-4bfa-a94c-7831d8c26ac5/current_0.png,https://replicate.delivery/mgxm/6998d47e-b6e0-4aa3-a71e-6bd26c0d60ba/current_1.png,https://replicate.delivery/mgxm/ad346d22-df84-4dd6-af88-759b68697aed/current_2.png,https://replicate.delivery/mgxm/c5b8b53e-c312-42b0-86a8-7c942e4b849b/current_3.png,https://replicate.delivery/mgxm/bd9dbf2b-3246-4cfb-8ca5-0edbcbde6a2f/current_4.png,https://replicate.delivery/mgxm/5ef2ef46-2b65-4c38-bc68-79d620de1b82/current_5.png

Performance Metrics

37.78s Prediction Time
37.92s Total Time
All Input Parameters 
{
  "seed": -1,
  "steps": "100",
  "width": 256,
  "height": 256,
  "prompt": "super mario world level 1-1",
  "batch_size": "6",
  "guidance_scale": "8.19",
  "aesthetic_rating": 9,
  "aesthetic_weight": 0
}
Input Parameters
mask Type: string
a mask image for inpainting an init_image. white pixels = keep, black pixels = discard. resized to width = image width/8, height = image height/8
seed Type: integerDefault: -1Range: -1 - 4294967295
Seed for random number generator. If -1, a random seed will be chosen.
steps Type: integerDefault: 100Range: 15 - 250
Number of diffusion steps to run. Due to PLMS sampling, using more than 100 steps is unnecessary and may simply produce the exact same output.
width Default: 256
Target width
height Default: 256
Target height
prompt Type: stringDefault:
Your text prompt.
negative Type: stringDefault:
(optional) Negate the model's prediction for this text from the model's prediction for the target text.
batch_size Type: integerDefault: 1Range: 1 - 16
Batch size. (higher = slower)
init_image Type: string
(optional) Initial image to use for the model's prediction. If provided alongside a mask, the image will be inpainted instead.
guidance_scale Type: numberDefault: 5Range: -20 - 100
Classifier-free guidance scale. Higher values will result in more guidance toward caption, with diminishing returns. Try values between 1.0 and 40.0. In general, going above 5.0 will introduce some artifacting.
aesthetic_rating Type: integerDefault: 9
Aesthetic rating (1-9) - embed to use.
aesthetic_weight Type: numberDefault: 0.5
Aesthetic weight (0-1). How much to guide towards the aesthetic embed vs the prompt embed.
init_skip_fraction Type: numberDefault: 0Range: 0 - 1
Fraction of sampling steps to skip when using an init image. Defaults to 0.0 if init_image is not specified and 0.5 if init_image is specified.
intermediate_outputs Type: booleanDefault: false
Whether to return intermediate outputs. Enable to visualize the diffusion process and/or debug the model. May slow down inference.
Output Schema

Output

Type: array • Items Type: array

Example Execution Logs 
Using seed 1415277919
Running simulation for super mario world level 1-1
Encoding text embeddings with super mario world level 1-1 dimensions
Using aesthetic embedding 9 with weight 0.0
Using inpaint model but no image is provided. Initializing with zeros.
Running diffusion...

  0%|          | 0/100 [00:00<?, ?it/s]
  1%|          | 1/100 [00:01<02:00,  1.21s/it]
  2%|▏         | 2/100 [00:02<01:54,  1.17s/it]
  3%|▎         | 3/100 [00:03<01:52,  1.16s/it]
  4%|▍         | 4/100 [00:03<01:18,  1.23it/s]
  5%|▌         | 5/100 [00:04<00:59,  1.60it/s]
  6%|▌         | 6/100 [00:04<00:47,  1.96it/s]
  7%|▋         | 7/100 [00:04<00:40,  2.29it/s]
  8%|▊         | 8/100 [00:04<00:35,  2.57it/s]
  9%|▉         | 9/100 [00:05<00:32,  2.80it/s]
 10%|█         | 10/100 [00:05<00:30,  2.98it/s]
 11%|█         | 11/100 [00:05<00:28,  3.12it/s]
 12%|█▏        | 12/100 [00:06<00:27,  3.22it/s]
 13%|█▎        | 13/100 [00:06<00:26,  3.29it/s]
 14%|█▍        | 14/100 [00:06<00:25,  3.34it/s]
 15%|█▌        | 15/100 [00:06<00:25,  3.39it/s]
 16%|█▌        | 16/100 [00:07<00:24,  3.41it/s]
 17%|█▋        | 17/100 [00:07<00:24,  3.42it/s]
 18%|█▊        | 18/100 [00:07<00:23,  3.43it/s]
 19%|█▉        | 19/100 [00:08<00:23,  3.45it/s]
 20%|██        | 20/100 [00:08<00:23,  3.45it/s]
 21%|██        | 21/100 [00:08<00:22,  3.46it/s]
 22%|██▏       | 22/100 [00:08<00:22,  3.46it/s]
 23%|██▎       | 23/100 [00:09<00:22,  3.46it/s]
 24%|██▍       | 24/100 [00:09<00:21,  3.46it/s]
 25%|██▌       | 25/100 [00:09<00:21,  3.46it/s]
 26%|██▌       | 26/100 [00:10<00:21,  3.46it/s]
 27%|██▋       | 27/100 [00:10<00:21,  3.46it/s]
 28%|██▊       | 28/100 [00:10<00:20,  3.46it/s]
 29%|██▉       | 29/100 [00:10<00:20,  3.46it/s]
 30%|███       | 30/100 [00:11<00:20,  3.45it/s]
 31%|███       | 31/100 [00:11<00:19,  3.46it/s]
 32%|███▏      | 32/100 [00:11<00:19,  3.46it/s]
 33%|███▎      | 33/100 [00:12<00:19,  3.45it/s]
 34%|███▍      | 34/100 [00:12<00:19,  3.45it/s]
 35%|███▌      | 35/100 [00:12<00:18,  3.46it/s]
 36%|███▌      | 36/100 [00:13<00:18,  3.46it/s]
 37%|███▋      | 37/100 [00:13<00:18,  3.45it/s]
 38%|███▊      | 38/100 [00:13<00:17,  3.45it/s]
 39%|███▉      | 39/100 [00:13<00:17,  3.45it/s]
 40%|████      | 40/100 [00:14<00:17,  3.45it/s]
 41%|████      | 41/100 [00:14<00:17,  3.45it/s]
 42%|████▏     | 42/100 [00:14<00:16,  3.44it/s]
 43%|████▎     | 43/100 [00:15<00:16,  3.44it/s]
 44%|████▍     | 44/100 [00:15<00:16,  3.44it/s]
 45%|████▌     | 45/100 [00:15<00:15,  3.44it/s]
 46%|████▌     | 46/100 [00:15<00:15,  3.44it/s]
 47%|████▋     | 47/100 [00:16<00:15,  3.43it/s]
 48%|████▊     | 48/100 [00:16<00:15,  3.44it/s]
 49%|████▉     | 49/100 [00:16<00:14,  3.44it/s]
 50%|█████     | 50/100 [00:17<00:14,  3.44it/s]
 51%|█████     | 51/100 [00:17<00:14,  3.44it/s]
 52%|█████▏    | 52/100 [00:17<00:13,  3.44it/s]
 53%|█████▎    | 53/100 [00:17<00:13,  3.44it/s]
 54%|█████▍    | 54/100 [00:18<00:13,  3.43it/s]
 55%|█████▌    | 55/100 [00:18<00:13,  3.43it/s]
 56%|█████▌    | 56/100 [00:18<00:12,  3.43it/s]
 57%|█████▋    | 57/100 [00:19<00:12,  3.44it/s]
 58%|█████▊    | 58/100 [00:19<00:12,  3.44it/s]
 59%|█████▉    | 59/100 [00:19<00:11,  3.43it/s]
 60%|██████    | 60/100 [00:20<00:11,  3.43it/s]
 61%|██████    | 61/100 [00:20<00:11,  3.43it/s]
 62%|██████▏   | 62/100 [00:20<00:11,  3.44it/s]
 63%|██████▎   | 63/100 [00:20<00:10,  3.43it/s]
 64%|██████▍   | 64/100 [00:21<00:10,  3.43it/s]
 65%|██████▌   | 65/100 [00:21<00:10,  3.43it/s]
 66%|██████▌   | 66/100 [00:21<00:09,  3.43it/s]
 67%|██████▋   | 67/100 [00:22<00:09,  3.42it/s]
 68%|██████▊   | 68/100 [00:22<00:09,  3.43it/s]
 69%|██████▉   | 69/100 [00:22<00:09,  3.43it/s]
 70%|███████   | 70/100 [00:22<00:08,  3.42it/s]
 71%|███████   | 71/100 [00:23<00:08,  3.42it/s]
 72%|███████▏  | 72/100 [00:23<00:08,  3.42it/s]
 73%|███████▎  | 73/100 [00:23<00:07,  3.42it/s]
 74%|███████▍  | 74/100 [00:24<00:07,  3.42it/s]
 75%|███████▌  | 75/100 [00:24<00:07,  3.42it/s]
 76%|███████▌  | 76/100 [00:24<00:07,  3.42it/s]
 77%|███████▋  | 77/100 [00:24<00:06,  3.42it/s]
 78%|███████▊  | 78/100 [00:25<00:06,  3.43it/s]
 79%|███████▉  | 79/100 [00:25<00:06,  3.43it/s]
 80%|████████  | 80/100 [00:25<00:05,  3.43it/s]
 81%|████████  | 81/100 [00:26<00:05,  3.43it/s]
 82%|████████▏ | 82/100 [00:26<00:05,  3.42it/s]
 83%|████████▎ | 83/100 [00:26<00:04,  3.42it/s]
 84%|████████▍ | 84/100 [00:27<00:04,  3.41it/s]
 85%|████████▌ | 85/100 [00:27<00:04,  3.41it/s]
 86%|████████▌ | 86/100 [00:27<00:04,  3.41it/s]
 87%|████████▋ | 87/100 [00:27<00:03,  3.41it/s]
 88%|████████▊ | 88/100 [00:28<00:03,  3.41it/s]
 89%|████████▉ | 89/100 [00:28<00:03,  3.41it/s]
 90%|█████████ | 90/100 [00:28<00:02,  3.41it/s]
 91%|█████████ | 91/100 [00:29<00:02,  3.41it/s]
 92%|█████████▏| 92/100 [00:29<00:02,  3.41it/s]
 93%|█████████▎| 93/100 [00:29<00:02,  3.41it/s]
 94%|█████████▍| 94/100 [00:29<00:01,  3.41it/s]
 95%|█████████▌| 95/100 [00:30<00:01,  3.41it/s]
 96%|█████████▌| 96/100 [00:30<00:01,  3.41it/s]
 97%|█████████▋| 97/100 [00:30<00:00,  3.40it/s]
 98%|█████████▊| 98/100 [00:31<00:00,  3.41it/s]
Saving final sample/s
 99%|█████████▉| 99/100 [00:31<00:00,  3.40it/s]
100%|██████████| 100/100 [00:32<00:00,  1.59it/s]
100%|██████████| 100/100 [00:32<00:00,  3.05it/s]
Version Details
Version ID
ba29165ba5d8c023e32347885e30a21d44509530a3c6aa53724c79a0e1dd3926
Version Created
June 30, 2022
Run on Replicate →