IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an email to s dot adaszewski at gmail dot com. User accounts are meant only to report issues and/or generate pull requests. This is a purpose-specific Git hosting for ADARED projects. Thank you for your understanding!
sadaszewski1
/
decagon-pytorch
atdalīts no sadaszewski/decagon-pytorch
1
0
Atdalīts 0
Kods Problēmas 0 Izmaiņu pieprasījumi 0 Laidieni 0 Vikivietne Aktivitāte
104 Revīzijas
1 Atzars
866KB
Koks: 4ab4bd6eb2
Atzari Tagi
${ item.name }
Izveidot atzaru ${ searchTerm }
no '4ab4bd6eb2'
${ noResults }
Revīziju grafs

75 Revīzijas (4ab4bd6eb21553ceb4af5234fd38714074ffb08a)

Autors SHA1 Ziņojums Datums
  Stanislaw Adaszewski ff4fb96bed Add CrossEntropyLoss and test_cross_entropy_loss_01(). pirms 4 gadiem
  Stanislaw Adaszewski 25e05cf1c2 Use torch.nn.Parameter(List) in decode. pirms 4 gadiem
  Stanislaw Adaszewski 55a5f3a2d2 Use torch.nn.ModuleList in convlayer. pirms 4 gadiem
  Stanislaw Adaszewski 8767db601d Remove node_type_row/node_type_column as redundant in add_relation_type(). pirms 4 gadiem
  Stanislaw Adaszewski e8122c3321 Get test_decode_layer_01 to pass. pirms 4 gadiem
  Stanislaw Adaszewski 832f620a78 Store relation types in a list instead of a dictionary and related fixes. pirms 4 gadiem
  Stanislaw Adaszewski 7eda6bdfb9 Fix regressions in trainprep. pirms 4 gadiem
  Stanislaw Adaszewski 915c568490 Small fix. pirms 4 gadiem
  Stanislaw Adaszewski 6c52bc1f71 Fix regressions in test_convlayer. pirms 4 gadiem
  Stanislaw Adaszewski b5d2f8fcda Start implementing RelationFamily. pirms 4 gadiem
  Stanislaw Adaszewski 56ce7aa60b Add type checks to DecodeLayer. pirms 4 gadiem
  Stanislaw Adaszewski 15e95bdc72 Use hints instead of is_autogenerated. pirms 4 gadiem
  Stanislaw Adaszewski 53cd182e6b Modify decoders to handle one relation at a time. pirms 4 gadiem
  Stanislaw Adaszewski eda27bee0e Increase test coverage. pirms 4 gadiem
  Stanislaw Adaszewski bd894a0257 Move back to using single-level dictionary for Data.relation_types. pirms 4 gadiem
  Stanislaw Adaszewski f51855e5b3 Add first test for declayer. pirms 4 gadiem
  Stanislaw Adaszewski 1868014ca3 Work on declayer. pirms 4 gadiem
  Stanislaw Adaszewski dd0eb81251 Add tests for decode. pirms 4 gadiem
  Stanislaw Adaszewski a305dc70aa Add tests for convlayer. pirms 4 gadiem
  Stanislaw Adaszewski d4dd1f2923 Add input to icosagon. pirms 4 gadiem
  Stanislaw Adaszewski 07a82e0da9 Work on icosagon. pirms 4 gadiem
  Stanislaw Adaszewski d503fb5738 Bring more elements to icosagon. pirms 4 gadiem
  Stanislaw Adaszewski 706ff0b009 New implementation of normalize for one/two node types and sparse/dense - good. pirms 4 gadiem
  Stanislaw Adaszewski 6f80b65a46 New implementation for one node type normalization seems to work. pirms 4 gadiem
  Stanislaw Adaszewski 05b1ecf47a Start rework normalize. pirms 4 gadiem
  Stanislaw Adaszewski 584bf19c3f Work on icosagon.trainprep. pirms 4 gadiem
  Stanislaw Adaszewski 7ed4bc373a Work on icosagon.trainprep. pirms 4 gadiem
  Stanislaw Adaszewski 3795545674 Start icosagon. pirms 4 gadiem
  Stanislaw Adaszewski 456a037e58 Fix typo. pirms 4 gadiem
  Stanislaw Adaszewski a272a17f23 Work on trainprep. pirms 4 gadiem
  Stanislaw Adaszewski e201608caf Add trainprep. pirms 4 gadiem
  Stanislaw Adaszewski 28272e3d29 Add AdjListRelationType, AdjListData. pirms 4 gadiem
  Stanislaw Adaszewski fab210448d Correct support for assymetric relations in Data. pirms 4 gadiem
  Stanislaw Adaszewski a74884dadd Add train_val_test_split_adj_mat(). pirms 4 gadiem
  Stanislaw Adaszewski 7ab97e2fb6 Add test_unigram_03(). pirms 4 gadiem
  Stanislaw Adaszewski f2121bef53 Add fixed_unigram_candidate_sampler(). pirms 4 gadiem
  Stanislaw Adaszewski 39aa1b4211 Implement and test pairwise decoders. pirms 4 gadiem
  Stanislaw Adaszewski c1689b4985 Split decoders into cartesian and pairwise. pirms 4 gadiem
  Stanislaw Adaszewski b08ae2e160 Test for an approach to pairwise (rather than cartesian product) link predition. pirms 4 gadiem
  Stanislaw Adaszewski 4f243fbc04 Copyright notes. pirms 4 gadiem
  Stanislaw Adaszewski c37e4dc01e Ok the very first decoding seems to work end-to-end. pirms 4 gadiem
  Stanislaw Adaszewski 1d2fac0919 Split layer into submodules. pirms 4 gadiem
  Stanislaw Adaszewski 52be4061b3 Adjust DecagonLayer to normalize per edge type. pirms 4 gadiem
  Stanislaw Adaszewski 241a2f92b9 Fix tests. pirms 4 gadiem
  Stanislaw Adaszewski c74555fef5 Refactor convolve into 3 separate modules. pirms 4 gadiem
  Stanislaw Adaszewski 02bbfc4958 Add test_decagon_layer_04(). pirms 4 gadiem
  Stanislaw Adaszewski 69e8ce5081 Add universal convolution units that can handle both sparse and dense inputs. pirms 4 gadiem
  Stanislaw Adaszewski 99dbcdeb91 Prefix dense versions of convolution classes with Dense. pirms 4 gadiem
  Stanislaw Adaszewski 29e81f4eba Add test for DecagonLayer. pirms 4 gadiem
  Stanislaw Adaszewski 69dd9a49e2 Dummy run of DecagonLayer seems to work, that's something. pirms 4 gadiem