globus-labs mof-generation-at-scale issues

globus-labs / mof-generation-at-scale

Create new MOFs by combining generative AI and simulation on HPC

MIT License

18 stars 5 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Make it possible to turn off DFT

#140 WardLT opened 2 weeks ago
0
Update environment due to Polaris upgrades

#139 WardLT closed 1 month ago
0
Introduces tests for CP2K

#138 WardLT closed 2 months ago
0
Test every monday

#137 WardLT opened 2 months ago
0
Do not hard code generation batch size

#136 WardLT opened 5 months ago
0
Add more testing scripts

#135 WardLT closed 5 months ago
1
Enforce single XPU training

#134 WardLT closed 6 months ago
0
Update Sunspot Configurations and Tests

#133 WardLT closed 6 months ago
1
Update or remove serial example

#132 WardLT opened 7 months ago
0
Add data and latency analysis notebooks

#131 gpauloski closed 7 months ago
0
Add files via upload

#130 williamyxl closed 7 months ago
1
High capacity mof repo

#129 williamyxl closed 7 months ago
1
Increase the number of MPI ranks per GPU for CP2K

#128 WardLT opened 7 months ago
0
Increase the Number of LAMMPS per GPU

#127 WardLT opened 7 months ago
0
Diagnose why multi-gpu training fails above 3k entires

#126 WardLT opened 7 months ago
0
Update plot-node-usage-polaris.ipynb

#125 williamyxl closed 7 months ago
0
Switch to training on RASPA later in run

#124 WardLT closed 7 months ago
1
Ignore conv. failure for CP2K during optimization

#123 WardLT closed 7 months ago
1
Start training on best gas capacity after enough complete

#122 WardLT closed 7 months ago
0
Integrate RASPA into workflow

#121 WardLT closed 7 months ago
1
Graphical Scoring Implementation

#120 nathaniel-hudson opened 7 months ago
0
Raspa gcmc

#119 williamyxl closed 7 months ago
1
Pull `.to` statements out of `__init__` functions in DiffLinker

#118 WardLT closed 7 months ago
0
Dedicate a whole node to training

#117 WardLT closed 7 months ago
1
Increase training threshold to 10% strain

#116 WardLT closed 7 months ago
0
Dedicate nodes to training DiffLinker

#115 WardLT closed 7 months ago
0
Consider ignoring convergence failures in CP2K

#114 WardLT closed 7 months ago
0
Always start from initial, filter using stability

#113 WardLT closed 7 months ago
1
Allow >1 LAMMPS per GPU

#112 WardLT closed 7 months ago
1
Catching up

#111 nathaniel-hudson closed 7 months ago
1
Don't Retrain on Unstable MOFs

#110 WardLT closed 7 months ago
0
Restart DiffLinker from Initial Weights

#109 WardLT closed 7 months ago
0
Additional updates to workflow with DFT

#108 WardLT closed 7 months ago
1
Save Result from ligand post processing after storage

#107 WardLT closed 7 months ago
0
Add optimization as an option for CP2K

#106 WardLT closed 7 months ago
1
Reduce LAMMPS I/O

#105 WardLT closed 7 months ago
1
Move saturation from inference worker to post processing

#104 WardLT opened 7 months ago
0
major change to ligand generation and postprocess to boost MOF stability

#103 williamyxl closed 7 months ago
2
Plot how distribution of molecules changes over time

#102 WardLT opened 7 months ago
0
Confirm what structure to use for partial charge calculation

#101 WardLT closed 7 months ago
1
Add a setup function to HPCConfig

#100 WardLT opened 8 months ago
0
Add DFT into Workflow

#99 WardLT closed 7 months ago
1
Training fixes from test of workflow on Polaris

#98 WardLT closed 8 months ago
1
Offload ligand post-processing to idle CPU cores

#97 WardLT closed 8 months ago
1
Measure the Performance of Each Round

#96 WardLT opened 8 months ago
0
Use OT for CP2K Computations

#95 WardLT closed 7 months ago
0
Use MPS for LAMMPS Tasks on Polaris

#94 WardLT closed 7 months ago
1
Switch to Writing LAMMPS Outputs to RAM Disk

#93 WardLT closed 7 months ago
0
Adjust How We Pick which Ligand to Generate On-the-Fly

#92 WardLT opened 8 months ago
0
Diagnose Why Sunspot Doesn't Run PyTorch Lightning

#91 WardLT opened 8 months ago
0