damdam775/portraits_dialogues 🔢📝❓🖼️✓ → 🖼️

▶️ 72 runs 📅 Oct 2025 ⚙️ Cog 0.16.0

game-asset-generation image-consistent-character-generation image-to-image

About

Create RPG like expressive character portraits for in game dialogs.

Example Output

Prompt:

"Generate a character expression sheet for the provided character. The sheet should feature multiple head-and-shoulders portraits, each displaying a distinct facial expression suitable for dialogue" . Many facial expressions : anger, amusement, bored, frightened. Keep details, accessories, upper body clothing and artistic style. Same character. Keep same clothing"

Output

Performance Metrics

18.53s Prediction Time

18.62s Total Time

All Input Parameters

{
  "prompt": "Generate a character expression sheet for the provided character. The sheet should feature multiple head-and-shoulders portraits, each displaying a distinct facial expression suitable for dialogue\\\" . Many facial expressions : anger, amusement, bored, frightened. Keep details, accessories, upper body clothing and artistic style. Same character. Keep same clothing",
  "guidance": 7.01,
  "megapixels": "1",
  "input_image": "https://replicate.delivery/pbxt/NuqrcCDEslFlZ042AorhCUOGvskqPhPQQjgx9t8WTFNIVCB3/2852058-hopey.png",
  "aspect_ratio": "1:1",
  "lora_strength": 1,
  "output_format": "png",
  "output_quality": 100,
  "num_inference_steps": 50
}

Input Parameters

seed Type: integer: Random seed for reproducible generation. Leave blank for random.
prompt (required) Type: string: Text description of what you want to generate, or the instruction on how to edit the given image.
guidance Type: numberDefault: 2.5Range: 0 - 10: Guidance scale for generation
megapixels Default: 1: Approximate number of megapixels for generated image
input_image (required) Type: string: Image to use as reference. Must be jpeg, png, gif, or webp.
aspect_ratio Default: match_input_image: Aspect ratio of the generated image. Use 'match_input_image' to match the aspect ratio of the input image.
lora_strength Type: numberDefault: 1: Strength of the lora
output_format Default: webp: Output image format
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
replicate_weights Type: string: Path to the lora weights
num_inference_steps Type: integerDefault: 30Range: 4 - 50: Number of inference steps
disable_safety_checker Type: booleanDefault: false: Disable NSFW safety checker

Output Schema

Output

Type: string • Format: uri

Example Execution Logs

Unloaded 304 layers
LoRAs unloaded in 0.0033s
Loading LoRA weights for /src/weights-cache/5348236c5d2b11fb
LoRA weights loaded
Extracting keys
Keys extracted
Applying LoRA:   0%|          | 0/304 [00:00<?, ?it/s]
Applying LoRA: 100%|██████████| 304/304 [00:00<00:00, 3458.32it/s]
Loading LoRA in bf16
LoRA applied in 0.16s
Using seed: 63519
Target dimensions: 1024x1024
Total safe images: 1 out of 1

Version Details

Version ID: a53f5fe4bc42f2e5d4eda9dc261024bd6af6208948ab170b428aada28622dea7
Version Created: October 20, 2025

Run on Replicate →