Access the output of local-embedding network #52
Replies: 3 comments
-
Hi Luigi Please take a look at the code at the devel branch. When freeze the model, please add "o_descriptor" to the default node list by the command line argument "-n", e.g. Then you may want to modify the class source/train/DeepPot.py to output the descriptors by using the python interface of model inference Please let us know if it works Han |
Beta Was this translation helpful? Give feedback.
-
Dear Han, I cloned the devel branch and tried to install the DeePMD-kit's python interface. I am able to install it via pip, but when I try to check the installation with
Please note that this error was first occurring related to libop_api.so, then I changed the cmake part of setup.py to have -DTF_GOOGLE_BIN=true and now the error is related to libop_grads.so. The configuration I am using is the following:
Do you have any suggestions? Thank you, |
Beta Was this translation helpful? Give feedback.
-
Hi, |
Beta Was this translation helpful? Give feedback.
-
Hi all,
I wanted to ask you whether it is possible to print the output of the local-embedding network, in order to check the inputs of the fitting network.
If this is not implemented yet, could you please give some directions about the part of the code that needs to be modified in order to get it?
Thank you,
Luigi
Beta Was this translation helpful? Give feedback.
All reactions