atomicarchitects/equiformer_v2
[ICLR 2024] EquiformerV2: Improved Equivariant Transformer for Scaling to Higher-Degree Representations
PythonMIT
Issues
- 1
Performance on OC20 S2EF 2M
#20 opened by TommyDzh - 4
Large performance gap in MD17/22 dataset
#12 opened by TommyDzh - 1
Question about Output Head
#19 opened by feyhong1112 - 1
Incorporating vector node features
#10 opened by patriksimurka - 1
Inquiry About rescale_factor in SO3.py
#18 opened by QuantumLab-ZY - 2
Question about the edge_rot_mat
#17 opened by HeegerGao - 5
Question on eSCN
#15 opened by liangzhixin-202169 - 1
- 1
Using Equiformer for Encoding Protein 3D Structure
#13 opened by mahdip72 - 7
Small equivariant example
#5 opened by DeNeutoy - 4
Training logs of QM9 dataset for EquiformerV2
#11 opened by fmocking - 3
e3nn tensors compatibility issue
#9 opened by liyy2 - 1
How to use the v2 to train IS2RE datasets?
#7 opened by mumu-1010 - 1
- 1
Speed compared to `TorchMD-Net`
#6 opened by shenoynikhil - 1
Multi-node multi-gpu training
#4 opened by gxglxy - 2
some questions
#3 opened by mlfffinder - 1
How to train on a customied dataset?
#2 opened by xiehuanyi - 4
Upload models to the Hugging Face Hub
#1 opened by osanseviero