Commits · 8d55b31d7a0dd23995ee1060e511835998725417 · Kia / Hammer FPGA backend

Aug 19, 2020
- remove print debugging · 8d55b31d
  Kia authored 4 years ago
  
  8d55b31d
- remove debug lines · 73b6feb9
  Kia authored 4 years ago
  
  73b6feb9
- start explanation of actual Boltzmann sampling · 989b49cd
  Kia authored 4 years ago
  
  989b49cd
Aug 18, 2020
- CFGBoltzmann.py · 4b067ff4
  Kia authored 4 years ago
  
  4b067ff4
- it works. · ad01478f
  Kia authored 4 years ago
  
  ad01478f
- it was an off-by-one. and the code works now. · faab1453
  Kia authored 4 years ago
  
  faab1453
- prepared instrumentation for tracing the mutual recursions to investigate this... · cedbf7f8
  Kia authored 4 years ago
```
prepared instrumentation for tracing the mutual recursions to investigate this bug (which looks suspiciously like a highly elaborate off-by-one)
```
  cedbf7f8
- use arrays not strings · a0e28c97
  Kia authored 4 years ago
  
  a0e28c97
- Gprim written but buggy · 2730320c
  Kia authored 4 years ago
  
  2730320c
Aug 17, 2020
- correct explanation of Case D (explains why we use Fprim) · 782bae6c
  Kia authored 4 years ago
  
  782bae6c
- case analysis explained for Gprim · 86a04902
  Kia authored 4 years ago
  
  86a04902
- Gzero and normalized_choice · 92ea3f3b
  Kia authored 4 years ago
  
  92ea3f3b
- bigger memoization table size · 5312c7a6
  Kia authored 4 years ago
  
  5312c7a6
- memoization! · 955ebb4e
  Kia authored 4 years ago
  
  955ebb4e
- there was an off-by-one that somehow managed to morph into an off-by-two · d53e2456
  Kia authored 4 years ago
  
  d53e2456
- think i'm done with the case analysis but it doesn't work as it, gonna debug that · dbae3014
  Kia authored 4 years ago
  
  dbae3014
- use sum() properly · a28706fc
  Kia authored 4 years ago
  
  a28706fc
- off-by-ones added and removed here · 77ff379d
  Kia authored 4 years ago
  
  77ff379d
Aug 16, 2020
- start case D · 44653a06
  Kia authored 4 years ago
  
  44653a06
- all cases done except case D · 422e11ea
  Kia authored 4 years ago
  
  422e11ea
- list of nonterminals must be indexed identically to that of the rules · ea72acc0
  Kia authored 4 years ago
  
  ea72acc0
- cases A and B · a39b8fc2
  Kia authored 4 years ago
  
  a39b8fc2
- start case analyse · 741fbe0c
  Kia authored 4 years ago
  
  741fbe0c
- work on Fprim some more · 2420da64
  Kia authored 4 years ago
  
  2420da64
- unsure why we made the nonterminals and terminals live in the same enum but that was a bad idea. · b7156514
  Kia authored 4 years ago
  
  b7156514
- start work on Fone proper · c5129a11
  Kia authored 4 years ago
  
  c5129a11
- make it live inside a class if only so we can do mutual recursion without NameError · f98375cd
  Kia authored 4 years ago
  
  f98375cd
- grammar preprocessor working · cd49ffdb
  Kia authored 4 years ago
  
  cd49ffdb
- start implementation proper · 698b5507
  Kia authored 4 years ago
  
  698b5507
- finished pseudocode · f678494b
  Kia authored 4 years ago
  
  f678494b
- continue description/documentation of proper CFG-enumeration algorithm · 7a0bc336
  Kia authored 4 years ago
```
in preparation for implementation
```
  7a0bc336
- continue work · e2bfd275
  Kia authored 4 years ago
  
  e2bfd275
- start implementing the McKenzie paper, because we actually do need osmething like it · e8101f74
  Kia authored 4 years ago
  
  e8101f74
Aug 12, 2020
- more work on CFG generation/manipulation stuff · 2910a77d
  Kia authored 4 years ago
  
  2910a77d
- cfg stuff separate out · de69da15
  Kia authored 4 years ago
  
  de69da15
- nmigen change · 0ac8e5a3
  Kia authored 4 years ago
  
  0ac8e5a3
Aug 11, 2020
- start work on actual Boltzmann sampling · 6c03de25
  Kia authored 4 years ago
  
  6c03de25
Aug 07, 2020
- sorta works; working now on making it properly Boltzmann-sampling · ab002474
  Kia authored 4 years ago
  
  ab002474
Aug 06, 2020
- sorta working · 199b3fe6
  Kia authored 4 years ago
  
  199b3fe6
- working, need tree finalization still · 08deb974
  Kia authored 4 years ago
  
  08deb974