Commits · 1f8d04b2ef2f5a0b8ff359c4b580e5ae159f8052 · novelai-storage / Basedformer

20 Apr, 2022 1 commit
- implement cached generation · 1f8d04b2
  novelailab authored Apr 20, 2022
  
  1f8d04b2
14 Apr, 2022 6 commits
- noemblm attention cleanup · 3a52a64a
  novelailab authored Apr 14, 2022
  
  3a52a64a
- fix · eaba913c
  novelailab authored Apr 14, 2022
  
  eaba913c
- clean gptj attention · 3c7bd057
  novelailab authored Apr 14, 2022
  
  3c7bd057
- if cache · 24e93cbf
  novelailab authored Apr 14, 2022
  
  24e93cbf
- fix attn more · fa2cfe52
  novelailab authored Apr 14, 2022
  
  fa2cfe52
- simplify attention on gpt2 · 39568281
  novelailab authored Apr 14, 2022
  
  39568281
12 Apr, 2022 1 commit
- upstream · 86e815ab
  novelailab authored Apr 12, 2022
  
  86e815ab
11 Apr, 2022 1 commit
- opt/model save/load works! · 1ab8bac6
  novelailab authored Apr 10, 2022
  
  1ab8bac6
10 Apr, 2022 2 commits
- optimizer save works · 2d0b32de
  novelailab authored Apr 09, 2022
  
  2d0b32de
- optimizer save load completed · 971ed5dc
  novelailab authored Apr 09, 2022
  
  971ed5dc
08 Apr, 2022 1 commit
- not too bad · 41b51369
  novelailab authored Apr 07, 2022
  
  41b51369
07 Apr, 2022 10 commits
- more fixing · 6cbab785
  novelailab authored Apr 06, 2022
  
  6cbab785
- fix formatting · 0879cf91
  novelailab authored Apr 06, 2022
  
  0879cf91
- a · 4141d527
  novelailab authored Apr 06, 2022
  
  4141d527
- create nn.py · 34739983
  novelailab authored Apr 06, 2022
  
  34739983
- change name · 67259a5b
  novelailab authored Apr 06, 2022
  
  67259a5b
- remove lm_train, merge utils, move optimizer · 02d93202
  novelailab authored Apr 06, 2022
  
  02d93202
- some more cleanup · 773e886f
  novelailab authored Apr 06, 2022
  
  773e886f
- refactor kinda done, yeet main · 09e4a19b
  novelailab authored Apr 06, 2022
  
  09e4a19b
- cleanup · 840dd7f4
  novelailab authored Apr 06, 2022
  
  840dd7f4
- ugh · 1c9d3a31
  novelailab authored Apr 06, 2022
  
  1c9d3a31
03 Apr, 2022 1 commit
- k · 3a2e4799
  novelailab authored Apr 03, 2022
  
  3a2e4799
31 Mar, 2022 1 commit
- bfloat16 model again · 3b246976
  novelailab authored Mar 31, 2022
  
  3b246976
30 Mar, 2022 1 commit
- might have broken stuff, will clean sshit later · 5ff36559
  novelailab authored Mar 30, 2022
  
  5ff36559
29 Mar, 2022 1 commit
- reproduced golden · 2a8307f2
  novelailab authored Mar 29, 2022
  
  2a8307f2
27 Mar, 2022 4 commits
- hypertrain GRU and pass layer_ids · 4596b61e
  novelailab authored Mar 27, 2022
  
  4596b61e
- fix checkpointing · c99ffa47
  novelailab authored Mar 26, 2022
  
  c99ffa47
- bighyper · 7e65fc56
  novelailab authored Mar 26, 2022
  
  7e65fc56
- hypernetwork training · b0155c91
  novelailab authored Mar 26, 2022
  
  b0155c91
26 Mar, 2022 3 commits
- things fixed · bc280afb
  novelailab authored Mar 26, 2022
  
  bc280afb
- set seed, everything works · fb25b47c
  novelailab authored Mar 26, 2022
  
  fb25b47c
- messy · 4f87dce5
  novelailab authored Mar 25, 2022
  
  4f87dce5
21 Mar, 2022 3 commits
- fp16 AMP maybe works · db8c9ae8
  novelailab authored Mar 21, 2022
  
  db8c9ae8
- pickle optimizer save/load · 347ef912
  novelailab authored Mar 21, 2022
  
  347ef912
- add todos and some stuff · b07251f0
  novelailab authored Mar 20, 2022
  
  b07251f0
18 Mar, 2022 3 commits
- ree. · d9df990b
  novelailab authored Mar 18, 2022
  
  d9df990b
- no dataset · 02e24910
  novelailab authored Mar 18, 2022
  
  02e24910
- it trains! · a9eca288
  novelailab authored Mar 18, 2022
  
  a9eca288
17 Mar, 2022 1 commit
- fast and good grad checkpoint, gelu torch script · 9a167649
  novelailab authored Mar 17, 2022
  
  9a167649