Models BOTnet and scaleshiftBOTnet fails to train about mace HOT 5 CLOSED

The key is that BOTNet can be reproduced as a special case of MACE, but you need to set the options Ilyes said above. In your input specify correlation and num_interactions. By default for MACE they are 3 and 2, but to get BOTNet you need to set them to 1 and 5.

from mace.

Dear @utkarshp1161,

This way of creating BOTnet is no longer supported from the arg parser. However you can create a BOTNet model very easily by just selecting:

--correlation=1 --num_interactions=5

from mace.

Thanks for getting back.

--model flag should be scaleshiftmace or scaleshiftbotnet ?

I am using following command to train(as mentioned on GitHub page as well)

python ./mace/scripts/run_train.py
--name="MACE_model"
--train_file="train.xyz"
--valid_fraction=0.05
--test_file="test.xyz"
--config_type_weights='{"Default":1.0}'
--E0s='{1:-13.663181292231226, 6:-1029.2809654211628, 7:-1484.1187695035828, 8:-2042.0330099956639}'
--model="MACE"
--hidden_irreps='128x0e + 128x1o'
--r_max=5.0
--batch_size=10
--max_num_epochs=1500
--swa
--start_swa=1200
--ema
--ema_decay=0.99
--amsgrad
--restart_latest
--device=cuda \

from mace.

scaleshiftmace for this one

from mace.

Thanks ilyes.

from mace.