Pretrained Graph Transformers

Models:

• old-model_max500 > Small datasets, unshuffled
• old-model_max1000-0 > Big datasets, unshuffled
• old-model_max1000 > Big datasets, shuffled
• model_max500-task > Big datasets, shuffled and with new homogeneous tokenizers
• with-nodes-list > Big datasets, shuffled, with self edges for all nodes