Graph Machine Learning
frostedoyster commited on
Commit
606a8e6
·
verified ·
1 Parent(s): 21a010b

Update with PET-OMATPES model details

Browse files
Files changed (1) hide show
  1. README.md +12 -10
README.md CHANGED
@@ -9,21 +9,23 @@ This repository contains the FlashMD universal models. These are models designed
9
 
10
  ## Model Details
11
 
12
- These models offer acceleration with respect to traditional molecular dynamics by allowing the use of larger steps. We offer models that predict molecular dynamics
13
- using the following time steps:
 
 
 
 
 
 
 
 
14
 
15
- - 1 fs
16
- - 4 fs
17
- - 8 fs
18
- - 16 fs
19
- - 32 fs
20
- - 64 fs
21
 
22
  ### Model Sources
23
 
24
  - **Repository:** https://github.com/lab-cosmo/flashmd
25
- - **Paper [optional]:** Not yet available
26
- - **Demo [optional]:** Not yet available
27
 
28
  ## How to use
29
 
 
9
 
10
  ## Model Details
11
 
12
+ These models offer acceleration with respect to traditional molecular dynamics by allowing the use of larger steps. Each FlashMD model follows the following naming
13
+ convention: `flashmd_{mlip}_{timestep}fs.ckpt`, where "mlip" is the name of the MLIP (machine-learned interatomic potential) that was used to generate the training
14
+ trajectories for the FlashMD model, and {timestep} is the FlashMD timestep in fs. The corresponding MLIP is also made available as `mlip_{mlip}.ckpt`.
15
+
16
+ At the moment, we have:
17
+
18
+ - mlip = "pet-omatpes" (with timestep={1,2,4,8,16,32,64,128}): these are FlashMD models trained on the PET-OMATPES MLIP, and therefore allowing to run FlashMD at the
19
+ r2SCAN level of theory
20
+ - earlier models used in the reference paper ("flashmd_{timestep}fs.pt", with timestep={1,4,8,16,32,64}), only supported by `flashmd<=0.1.2` and useful to reproduce
21
+ older results. These are trained on the PET-MAD MLIP and therefore reproduce molecular dynamics at the PBEsol level of theory
22
 
 
 
 
 
 
 
23
 
24
  ### Model Sources
25
 
26
  - **Repository:** https://github.com/lab-cosmo/flashmd
27
+ - **Paper:** https://arxiv.org/abs/2505.19350
28
+ - **Demo:** https://atomistic-cookbook.org/examples/flashmd/flashmd-demo.html
29
 
30
  ## How to use
31