Add Hunyuan 3D 2.1 Support

[yousef-rafat]

I have created the reimplementation of the VAE. The main contributions are in rewriting/avoiding the use of unnecessary libraries and modularizing the code by removing code redundancy and better structuring of the code.

The implemented VAE is faster by approximately 5 seconds in the encoding process and 2 seconds in the decoding process.

The fps.py file contains the implementation of Farthest Point Sampling, avoiding using torch_cluster. The preprocess.py contains the Volume Decoder, Surface Extractor, and Trimesh exporting. The point_attention.py contains the code for the encoder. The transformer.py contains the code for the Transformer model that is used in the geo decoder of the VAE and the Self-Attention in the encoder. The vae.py contains the VAE code along with a test example.

[zapan-669]

please do it

[comfyui-wiki]

@yousef-rafat can you share the hunyuan3d 2.1 workflow with me? I tried hunyuan3d 2.0 workflow ,but sometime the output doesen't seem to right.

[yousef-rafat]

@comfyui-wiki

[guillaumeparra]

hi @yousef-rafat ,

Not sure I'm doing the right thing here, it doesn't work for me.

I get this error when executing the graph. I downloaded the model from here and renamed it hunyuan3d-dit-v2-1.ckpt :

https://huggingface.co/tencent/Hunyuan3D-2.1/tree/main/hunyuan3d-dit-v2-1

Here is the error message :

`got prompt
Warning, This is not a checkpoint file, trying to load it as a diffusion model only.
!!! Exception during processing !!! 'collections.OrderedDict' object has no attribute 'nelement'
Traceback (most recent call last):
File "/apps/comfyui/execution.py", line 496, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/apps/comfyui/execution.py", line 315, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/apps/comfyui/execution.py", line 289, in _async_map_node_over_list
await process_inputs(input_dict, i)
File "/apps/comfyui/execution.py", line 277, in process_inputs
result = f(**inputs)
^^^^^^^^^^^
File "/apps/comfyui/comfy_extras/nodes_video_model.py", line 22, in load_checkpoint
out = comfy.sd.load_checkpoint_guess_config(ckpt_path, output_vae=True, output_clip=False, output_clipvision=True, embedding_directory=folder_paths.get_folder_paths("embeddings"))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/apps/comfyui/comfy/sd.py", line 1076, in load_checkpoint_guess_config
out = load_state_dict_guess_config(sd, output_vae, output_clip, output_clipvision, embedding_directory, output_model, model_options, te_model_options=te_model_options, metadata=metadata)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/apps/comfyui/comfy/sd.py", line 1096, in load_state_dict_guess_config
diffusion_model = load_diffusion_model_state_dict(sd, model_options={})
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/apps/comfyui/comfy/sd.py", line 1192, in load_diffusion_model_state_dict
parameters = comfy.utils.calculate_parameters(sd)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/apps/comfyui/comfy/utils.py", line 109, in calculate_parameters
params += w.nelement()
^^^^^^^^^^
AttributeError: 'collections.OrderedDict' object has no attribute 'nelement'

Prompt executed in 2.58 seconds`

Thanks for your help !

hunyuan3d-v2.1-mesh-generation.json

[comfyui-wiki]

@guillaumeparra Use this workflow, I will add it to the template soon
hunyuan3d-v2.1.json