prunaai/hunyuan3d-2 ❓🔢🖼️ → ❓
About
hunyuan3d-2 optimised with the pruna toolkit: https://github.com/PrunaAI/pruna
Example Output
Output
Performance Metrics
22.96s
Prediction Time
251.55s
Total Time
All Input Parameters
{
"file_type": "glb",
"face_count": 40000,
"image_path": "https://raw.githubusercontent.com/Tencent/Hunyuan3D-2/dfa8967c893898745587435160eb96d4d08a82c0/assets/demo.png",
"num_chunks": 20000,
"generator_seed": 12345,
"octree_resolution": 200,
"num_inference_steps": 50
}
Input Parameters
- file_type
- File type
- face_count
- Target number of faces for simplification
- image_path
- Input image for hunyuan3d control
- num_chunks
- Number of chunks
- speed_mode
- Speed optimization level
- generator_seed
- Seed for random generator
- octree_resolution
- Octree resolution
- num_inference_steps
- Number of inference steps
Output Schema
- mesh_paint
- Mesh Paint
Example Execution Logs
Diffusion Sampling:: 0%| | 0/50 [00:00<?, ?it/s] Diffusion Sampling:: 2%|▏ | 1/50 [00:00<00:10, 4.83it/s] Diffusion Sampling:: 6%|▌ | 3/50 [00:00<00:06, 7.33it/s] Diffusion Sampling:: 8%|▊ | 4/50 [00:00<00:06, 6.82it/s] Diffusion Sampling:: 10%|█ | 5/50 [00:00<00:06, 6.54it/s] Diffusion Sampling:: 12%|█▏ | 6/50 [00:00<00:06, 6.37it/s] Diffusion Sampling:: 14%|█▍ | 7/50 [00:01<00:06, 6.26it/s] Diffusion Sampling:: 18%|█▊ | 9/50 [00:01<00:05, 8.10it/s] Diffusion Sampling:: 22%|██▏ | 11/50 [00:01<00:04, 9.34it/s] Diffusion Sampling:: 26%|██▌ | 13/50 [00:01<00:03, 10.18it/s] Diffusion Sampling:: 34%|███▍ | 17/50 [00:01<00:02, 14.43it/s] Diffusion Sampling:: 42%|████▏ | 21/50 [00:01<00:01, 17.37it/s] Diffusion Sampling:: 48%|████▊ | 24/50 [00:02<00:01, 17.59it/s] Diffusion Sampling:: 56%|█████▌ | 28/50 [00:02<00:01, 19.56it/s] Diffusion Sampling:: 68%|██████▊ | 34/50 [00:02<00:00, 24.55it/s] Diffusion Sampling:: 76%|███████▌ | 38/50 [00:02<00:00, 24.42it/s] Diffusion Sampling:: 84%|████████▍ | 42/50 [00:02<00:00, 24.31it/s] Diffusion Sampling:: 92%|█████████▏| 46/50 [00:02<00:00, 24.25it/s] Diffusion Sampling:: 98%|█████████▊| 49/50 [00:03<00:00, 17.23it/s] Diffusion Sampling:: 100%|██████████| 50/50 [00:03<00:00, 15.37it/s] Volume Decoding: 0%| | 0/407 [00:00<?, ?it/s] Volume Decoding: 16%|█▌ | 64/407 [00:00<00:00, 623.02it/s] Volume Decoding: 31%|███ | 127/407 [00:00<00:00, 284.21it/s] Volume Decoding: 41%|████ | 165/407 [00:00<00:00, 251.41it/s] Volume Decoding: 48%|████▊ | 195/407 [00:00<00:00, 236.66it/s] Volume Decoding: 55%|█████▍ | 222/407 [00:00<00:00, 227.53it/s] Volume Decoding: 61%|██████ | 247/407 [00:00<00:00, 221.02it/s] Volume Decoding: 66%|██████▋ | 270/407 [00:01<00:00, 216.51it/s] Volume Decoding: 72%|███████▏ | 293/407 [00:01<00:00, 213.22it/s] Volume Decoding: 77%|███████▋ | 315/407 [00:01<00:00, 210.95it/s] Volume Decoding: 83%|████████▎ | 337/407 [00:01<00:00, 209.34it/s] Volume Decoding: 88%|████████▊ | 359/407 [00:01<00:00, 207.97it/s] Volume Decoding: 93%|█████████▎| 380/407 [00:01<00:00, 207.03it/s] Volume Decoding: 99%|█████████▊| 401/407 [00:01<00:00, 206.32it/s] Volume Decoding: 100%|██████████| 407/407 [00:01<00:00, 204.24it/s] The first timestep on the custom timestep schedule is 989, not `self.config.num_train_timesteps - 1`: 999. You may get unexpected results when using this timestep schedule. The custom timestep schedule contains the following timesteps which are not on the original training/distillation timestep schedule: [tensor(890), tensor(791), tensor(692), tensor(593), tensor(494), tensor(395), tensor(296), tensor(197), tensor(98)]. You may get unexpected results when using this timestep schedule.
Version Details
- Version ID
6dd3e3e1f8a29a38807e8f23aaf8953a0051996ccc8c1861f709a5b1ee6826b5- Version Created
- April 24, 2025