lucataco/wan-2.2-first-last-frame 🔢📝🖼️ → 🖼️

▶️ 2.6K runs 📅 Sep 2025 ⚙️ Cog 0.16.2 🔗 GitHub

image-morphing image-to-video video-transition

About

Wan 2.2 First and Last Frame using 8-step inference w/ Lightning LoRA

Example Output

Prompt:

"the action figure turns into superman"

Output

Performance Metrics

20.03s Prediction Time

20.04s Total Time

All Input Parameters

{
  "seed": 0,
  "shift": 8,
  "prompt": "the action figure turns into superman",
  "end_image": "https://replicate.delivery/pbxt/NeiXSXNcPlJ9QKnI1g7cBbpDkaYubJL1ny0bZAlbS7e3hOK6/replicate-prediction-jhn2q5vqp9rmc0cs2r1r4s9nj4.webp",
  "start_image": "https://replicate.delivery/pbxt/NeiXRujYqqZvGLmdnBNCg3yi4DybjgDltDbqTeDek13yCMWA/replicate-prediction-qvv7v0yg11rmc0cs2r1tpmfvq8.jpeg",
  "guidance_scale": 1,
  "negative_prompt": "色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走,过曝，",
  "duration_seconds": 5,
  "guidance_scale_2": 1,
  "num_inference_steps": 8
}

Input Parameters

seed Type: integer: Random seed for reproducibility (None for random)
shift Type: numberDefault: 8Range: 1 - 10: Scheduler shift parameter
prompt Type: stringDefault: animate: Prompt describing the transition between images
end_image (required) Type: string: End frame image
start_image (required) Type: string: Start frame image
guidance_scale Type: numberDefault: 1Range: 0 - 10: Guidance scale for high noise
negative_prompt Type: stringDefault: 色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走,过曝，: Negative prompt
duration_seconds Type: numberDefault: 5Range: 0.5 - 10: Video duration in seconds
guidance_scale_2 Type: numberDefault: 1Range: 0 - 10: Guidance scale for low noise
num_inference_steps Type: integerDefault: 8Range: 1 - 30: Number of inference steps

Output Schema

Output

Type: string • Format: uri

Example Execution Logs

Running prediction...
Using seed: 0
The config attributes {'beta_start': 0.0001, 'beta_end': 0.02, 'beta_schedule': 'linear', 'trained_betas': None, 'solver_order': 2, 'prediction_type': 'flow_prediction', 'thresholding': False, 'dynamic_thresholding_ratio': 0.995, 'sample_max_value': 1.0, 'predict_x0': True, 'solver_type': 'bh2', 'lower_order_final': True, 'disable_corrector': [], 'solver_p': None, 'use_flow_sigmas': True, 'flow_shift': 3.0, 'timestep_spacing': 'linspace', 'steps_offset': 0, 'final_sigmas_type': 'zero', 'rescale_betas_zero_snr': False} were passed to FlowMatchEulerDiscreteScheduler, but are not expected and will be ignored. Please verify your scheduler_config.json configuration file.
`num_frames - 1` has to be divisible by 4. Rounding to the nearest number.
  0%|          | 0/8 [00:00<?, ?it/s]
 12%|█▎        | 1/8 [00:02<00:15,  2.16s/it]
 25%|██▌       | 2/8 [00:03<00:11,  1.84s/it]
 38%|███▊      | 3/8 [00:05<00:09,  1.98s/it]
 50%|█████     | 4/8 [00:08<00:08,  2.05s/it]
 62%|██████▎   | 5/8 [00:10<00:06,  2.09s/it]
 75%|███████▌  | 6/8 [00:12<00:04,  2.11s/it]
 88%|████████▊ | 7/8 [00:14<00:02,  2.12s/it]
100%|██████████| 8/8 [00:16<00:00,  2.13s/it]
100%|██████████| 8/8 [00:16<00:00,  2.08s/it]
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)

Version Details

Version ID: 003fd8a38ff17cb6022c3117bb90f7403cb632062ba2b098710738d116847d57
Version Created: September 4, 2025

Run on Replicate →