Commits · bbe87a1e9ac2ac1227c35a0e8dfcb3873a6a16fa · Kia / Hammer FPGA backend · GitLab

Snippets Groups Projects

Aug 24, 2020
- i think this tree equality check is less catastrophically broken. · bbe87a1e
  Kia authored 4 years ago
  
  bbe87a1e
- added stochastic test to ensure that the tree-equality check is working... · bec893c0
  Kia authored 4 years ago
  bec893c0
- tree equality check! · 02403ee6
  Kia authored 4 years ago
  
  02403ee6
- add a quick-and-dirty special-casing for handling the start symbol in the... · 939a32aa
  Kia authored 4 years ago
  939a32aa
- limit how far we look into the serialized parse tree memory (in preparation... · dc427a91
  Kia authored 4 years ago
  dc427a91
- removed off-by-one that led to anomalous-looking tree prettyprinting · b1007c31
  Kia authored 4 years ago
  
  b1007c31
- tree reversing is done; now to handle the spurious double-start-symbol · de6c1f32
  Kia authored 4 years ago
  
  de6c1f32
Aug 23, 2020
- Gprim has effects and we must respect this by scheduling the addition of X_ijk... · 1af7a564
  Kia authored 4 years ago
  1af7a564
- we are generating almost-correct trees. two issues remain: how to handle the... · fdeb387c
  Kia authored 4 years ago
  fdeb387c
Aug 22, 2020
- continue derivation tree work; it's not quite yet working · b86c1dc7
  Kia authored 4 years ago
  
  b86c1dc7
- return a tuple so split out here · a9685bf9
  Kia authored 4 years ago
  
  a9685bf9
- bug found · 55fcd335
  Kia authored 4 years ago
  
  55fcd335
- introduced a bug somehow, committing this to be able to bisect · 50d1b6e6
  Kia authored 4 years ago
  
  50d1b6e6
- renames and add utility function for tree index dereferencing · cce90d5b
  Kia authored 4 years ago
  
  cce90d5b
- .gitignore · ecc6a8e0
  Kia authored 4 years ago
  
  ecc6a8e0
- continue work on creating the derivation tree · 3d105a52
  Kia authored 4 years ago
  
  3d105a52
Aug 20, 2020
- start augmenting the function signatures to do the tree-building on the fly · 13d968ac
  Kia authored 4 years ago
  
  13d968ac
Aug 19, 2020
- removed duplicate Boltzmann code from cfg_utils · e6c7c9d7
  Kia authored 4 years ago
  
  e6c7c9d7
- add code for working with trees · a1a651e7
  Kia authored 4 years ago
  
  a1a651e7
- working on generating the derivation/parse tree during the generation of the random string · 0c4468ac
  Kia authored 4 years ago
  
  0c4468ac
- convert tokens to name (with an awful ad-hoc scheme that we should revise... · 9821434c
  Kia authored 4 years ago
  9821434c
- we can successfully generate a string and parse it · e5696dd8
  Kia authored 4 years ago
  
  e5696dd8
- remove spurious import · 3a146cfc
  Kia authored 4 years ago
  
  3a146cfc
- 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
  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