@@ -29,7 +29,11 @@ Identifying dynamics states, and learning their equation-of-motion, can be perfo
<br></br>
**Datasets:**
We uploaded two batches of datasets: Batch-a) Equilibrium dynamics at a constant temperature at zero electric-field, with one trajectory each for 4-different defect structures (i.e. SETs 1 to 4). Each trajectory has a time-step of 0.25 fs, and is run for 7775000 time-steps, with snapshots written out every 4 time-steps (i.e. every 1fs); and Batch-b) Non-Equilibrium dynamics at a constant temperature with the same 0.25 fs time-step, but data dumped every 500 times-steps (i.e. every 125fs) for each of the SETs. The total trajectory defect structure (i.e. each SET) is 2800000 time-steps (5600 snapshots), with stepping of electric-field by 0.01 $`V/Å`$, after every 100,000 time-steps (i.e. every 200 snapshots), from E=0 to E=0.05 $`V/Å`$ to E = -0.05 $`V/Å`$ to E=0.05 $`V/Å`$.
DATASET LINK : https://drive.google.com/drive/folders/1W62AFl-_q-xC_aTkJeh1RxGt3-MHBXEN?usp=sharing
We uploaded two batches of datasets: Batch-a) Equilibrium dynamics at a constant temperature at zero electric-field, with one trajectory each for 4-different defect structures (i.e. SETs 1 to 4). Each trajectory has a time-step of 0.25 fs, and is run for 7775000 time-steps, with snapshots written out every 4 time-steps (i.e. every 1fs); and Batch-b) Non-Equilibrium dynamics at a constant temperature with the same 0.25 fs time-step, but data dumped every 500 times-steps (i.e. every 125fs) for each of the SETs. The total trajectory of each defect structure (i.e. each SETs 1 to 4) is 2800000 time-steps (5600 snapshots), with stepping of electric-field by 0.01 $`V/Å`$, after every 100,000 time-steps (i.e. every 200 snapshots), from E=0 to E=0.05 $`V/Å`$ to E = -0.05 $`V/Å`$ to E=0.05 $`V/Å`$.
