zjukg / KGTransformer

[Paper][WWW2023] Structure Pre-training and Prompt Tuning for Knowledge Graph Transfer
https://arxiv.org/pdf/2303.03922.pdf
46 stars 5 forks source link

Could you release your pretrain model parameters #5

Closed WEIYanbin1999 closed 1 year ago

WEIYanbin1999 commented 1 year ago

Dear authors, Thanks for you good work, I am reproducing but the memory for pretrain is large to follow. So could you please release the pretrain_model file for us? pretrain_model = 'pretrain_models/BIG/model_layer-4_hidden-768_heads-12_seq-126_textE-cls_t0-1.0_t1-1.0_t2-1.0.ep4'#

Thanks a lot

WEIYanbin1999 commented 1 year ago

I have run all the path with improved machine memory.