ggfm.models¶

`ggfm.models.HGT`	The Heterogeneous Graph Transformer (HGT) operator from the "Heterogeneous Graph Transformer" paper.
`ggfm.models.MPNN`	Message Passing Neural Network (MPNN) layer :param in_channels: Number of input features :type in_channels: int :param hidden_channels: Number of hidden features :type hidden_channels: int :param out_channels: Number of output features :type out_channels: int
`ggfm.models.GPT_GNN`	"GPT-GNN: Generative Pre-Training of Graph Neural Networks" paper.
`ggfm.models.LLAGA`	"LLaGA: Large Language and Graph Assistant" paper.
`ggfm.models.Matcher`	Matching between a pair of nodes to conduct link prediction.
`ggfm.models.RNNModel`	Container module with an encoder, a recurrent module, and a decoder.
`ggfm.models.Classifier`	Classifier for graph node classification task.
`ggfm.models.LinkPredictor`	LinkPredictor for graph link prediction task.
`ggfm.models.CLIP`	CLIP model class
`ggfm.models.GraphLlamaModel`	Graph Llama model
`ggfm.models.GraphLlamaForCausalLM`	Graph Llama model for causal language modeling
`ggfm.models.GraphGPT_pl`	Graph GPT model
`ggfm.models.PT_HGNN`	PT-HGNN the model for heterogeneous graph neural network with pre-training.
`ggfm.models.SGFormer`	SGFormer model from the `"SGFormer: Spatial Graph Transformer for Molecular Property Prediction" :param in_channels: Number of input features. :type in_channels: int :param hidden_channels: Number of hidden features. :type hidden_channels: int :param out_channels: Number of output :type out_channels: int.
`ggfm.models.HeteroLlamaForCausalLM`	HeteroLLaMA model for causal language modeling with heterogeneous graphs.