Awesome

GraphGPT

Implementation of GraphGPT: Graph enhanced GPT for conditioned molecular generation

Note:

The training code and the full trained model are coming soon!

generate:

One can specify different translate configurations in gen_guacamol.sh as the sample code below.

To replicate our results, download the pre-trained checkpoints from here.

python generate/generate.py
  --model_weight guacamol_sas.pt \
  --props sas \
  --data_name guacamol2 \

Node: Our code was developed with reference to the code of GPT-1 and MolGPT, and we would like to express our gratitude to them. If you have any questions, feel free to contact Hao Lu, luhao@stu.ouc.edu.cn