Commits · fface86d417ae7f59bfd1e89fd75fbea1482f770 · Lei Wang / dTRG

17 Feb, 2020 1 commit
- add newton in dimer · fface86d
  Lei Wang authored Feb 17, 2020
  
  fface86d
22 Sep, 2019 1 commit
- add h in the ising tensor · b4485128
  Lei Wang authored Sep 22, 2019
  
  b4485128
12 Sep, 2019 1 commit
- we SHOULD differentiate into the 2d contraction · c734f9b7
  Lei Wang authored Sep 12, 2019
  
  c734f9b7
23 Aug, 2019 1 commit
- added adlib (before I was hacking pytorch source) · 2b604e71
  Lei Wang authored Aug 23, 2019
  
  2b604e71
14 Dec, 2018 8 commits
- update · 2b3c0035
  Lei Wang authored Dec 14, 2018
  
  2b3c0035
- start ctmrg from tensor · dc5ed25b
  Lei Wang authored Dec 14, 2018
  
  dc5ed25b
- looks like ctmrg with iteration is the fastest · 4162baf1
  Lei Wang authored Dec 14, 2018
  
  4162baf1
- fix · 4ff4842c
  Lei Wang authored Dec 14, 2018
  
  4ff4842c
- make ising work · f2593182
  Lei Wang authored Dec 14, 2018
  
  f2593182
- fix · 306844bb
  Lei Wang authored Dec 14, 2018
  
  306844bb
- provide lanczos_steps · 5868058b
  Lei Wang authored Dec 14, 2018
  
  5868058b
- start lanczos from randn · a255e94e
  Lei Wang authored Dec 14, 2018
  
  a255e94e
13 Dec, 2018 11 commits
- lanczos works for 2d Ising, need to test for 3d case · 35a00557
  Lei Wang authored Dec 13, 2018
  
  35a00557
- lanczos can have matsize=1 · c21098b1
  Lei Wang authored Dec 13, 2018
  
  c21098b1
- clean up · 2bf5025b
  Lei Wang authored Dec 13, 2018
  
  2bf5025b
- is able to use lanczos to diagonalize transfer matrix · 143e0e03
  Lei Wang authored Dec 13, 2018
  
  143e0e03
- backprop through lanczos is slower than forward; lanczos is faster than symeig when Dcut>50 · e0459b6a
  Lei Wang authored Dec 13, 2018
  
  e0459b6a
- cache boundary MPS · f1e0e64e
  Lei Wang authored Dec 13, 2018
  
  f1e0e64e
- symmetrize T in vmps · 1afdb8f6
  Lei Wang authored Dec 13, 2018
  
  1afdb8f6
- even better encapsulation; inner loop optimize A (boundary MPS) , outer loop... · a144be5b
  Lei Wang authored Dec 13, 2018
```
even better encapsulation; inner loop optimize A (boundary MPS) , outer loop optimize T (boundary PEPS)
```
  a144be5b
- better control of gradient flow · fc796aba
  Lei Wang authored Dec 13, 2018
  
  fc796aba
- compute physical obs in vmps · d9aea05f
  Lei Wang authored Dec 13, 2018
  
  d9aea05f
- eigensolver instead of matrix product · 94759d68
  Lei Wang authored Dec 13, 2018
  
  94759d68
12 Dec, 2018 1 commit
- Merge branch 'master' of gitlab.theory.iphy.ac.cn:wanglei/dtrg · 10f2d290
  Lei Wang authored Dec 12, 2018
  
  10f2d290
11 Dec, 2018 5 commits
- iterative variational calculation · e98b40e7
  Lei Wang authored Dec 11, 2018
  
  e98b40e7
- added vmps · 246dca5e
  Lei Wang authored Dec 11, 2018
  
  246dca5e
- change to more symmetric ising tensor · 1df446a5
  Lei Wang authored Dec 11, 2018
  
  1df446a5
- correct symmetry · 400ab95a
  Lei Wang authored Dec 11, 2018
  
  400ab95a
- simplify symmetrize;ctmrg for ising · 5d8e8921
  Lei Wang authored Dec 11, 2018
  
  5d8e8921
10 Dec, 2018 5 commits
- it is also possible to do torch.symeig · 4aefaec8
  Lei Wang authored Dec 10, 2018
  
  4aefaec8
- trying to speedup einsum in ctmrg; but it is minor, so the bottneck is svd · 22dd9226
  Lei Wang authored Dec 10, 2018
  
  22dd9226
- symmetrize tensors · e28bad7f
  Lei Wang authored Dec 10, 2018
  
  e28bad7f
- added ctmrg, compute lnZ using Z1*Z3/Z2^2 · 43862c5f
  Lei Wang authored Dec 10, 2018
  
  43862c5f
- added some tests for performance of einsum versus torchdot · e1ba1120
  Lei Wang authored Dec 10, 2018
  
  e1ba1120
07 Dec, 2018 1 commit
- added TODO.md · e67218ae
  Lei Wang authored Dec 07, 2018
  
  e67218ae
06 Dec, 2018 2 commits
- optimize hotrg2 · 25b4f3dc
  Lei Wang authored Dec 06, 2018
  
  25b4f3dc
- use same ludr convention for hotrg2 · e8cab639
  Lei Wang authored Dec 06, 2018
  
  e8cab639
05 Dec, 2018 3 commits
- use svd instead of symeig in hotrg2 · ef1c3fcb
  Lei Wang authored Dec 05, 2018
  
  ef1c3fcb
- correct trace in hotrg · 22fc0b77
  Lei Wang authored Dec 05, 2018
  
  22fc0b77
- correct truncation · afd0593a
  Lei Wang authored Dec 05, 2018
  
  afd0593a