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!

232 次程式碼提交 (5cf1e3d1b594cdbeff88c9e97268449e05089cac)
 

作者 SHA1 備註 提交日期
  Stanislaw Adaszewski 5cf1e3d1b5 torch_stablesort CUDA version works and is fast. 4 年之前
  Stanislaw Adaszewski 897a6f0722 torch_stablesort_cuda.cu getting there... 4 年之前
  Stanislaw Adaszewski c1353168f3 CUDA performance needs improvement. 4 年之前
  Stanislaw Adaszewski 62af8fc5f1 Progress with CUDA version of torch_stablesort. 4 年之前
  Stanislaw Adaszewski f8f2901eec Start working on CUDA implementation of torch_stablesort. 4 年之前
  Stanislaw Adaszewski b439a46fc3 Use if constexpr to make stable_sort_impl() more compact. 4 年之前
  Stanislaw Adaszewski 283d9387a8 Add torch_stablesort.md. 4 年之前
  Stanislaw Adaszewski 87953842a6 Try to implement gradient. 4 年之前
  Stanislaw Adaszewski 747156a971 torch_stablesort_cpp seems to work. 4 年之前
  Stanislaw Adaszewski 14321daf62 Improved dispatch. 4 年之前
  Stanislaw Adaszewski 37c5db1330 Start torch_stablesort.0 4 年之前
  Stanislaw Adaszewski 356f3af3c3 With this weird little trick the new fixed_unigram_candidate_sampler() finally seems to be fully robust and to actually work. 4 年之前
  Stanislaw Adaszewski 79b0890a70 Mini fix. 4 年之前
  Stanislaw Adaszewski d7d442c5e3 New fixed_unigram_candidate_sampler() still requires work. 4 年之前
  Stanislaw Adaszewski 490b4f9281 New robust implementation of fixed_unigram_candidate_sampler() seems to be working. 4 年之前
  Stanislaw Adaszewski 689dbeed20 Add mergesort switch. 4 年之前
  Stanislaw Adaszewski 670237c3f8 Work on fixed_unigram_candidate_sampler_new(). 4 年之前
  Stanislaw Adaszewski 346f99fa83 Introduce fixed_unigram_candidate_sampler_new(). 4 年之前
  Stanislaw Adaszewski 6cf8251539 fixed_unigram_candidate_sampler() still requires some work to be perfect. 4 年之前
  Stanislaw Adaszewski 58b0aac18b Fixes in sampling. 4 年之前
  Stanislaw Adaszewski 2ff358f7ef Protect against endless loop. 4 年之前
  Stanislaw Adaszewski 8fbad74dfa Debug negative sampling. 4 年之前
  Stanislaw Adaszewski 9defd6c9e4 Add test_dropout. 4 年之前
  Stanislaw Adaszewski 9ff1038728 Add extra checks in Data. 4 年之前
  Stanislaw Adaszewski 7e83fed37b Add test_normalize. 4 年之前
  Stanislaw Adaszewski e184699365 Add test_split_data_01/02(). 4 年之前
  Stanislaw Adaszewski 6d27e26fa7 Add test_split_edge_type_01/02/03/04(). 4 年之前
  Stanislaw Adaszewski 604b81675e Add test_split_adj_mat_(01|02|03)(). 4 年之前
  Stanislaw Adaszewski 7094ae87db Fix negative_sample_adj_mat(). 4 年之前
  Stanislaw Adaszewski 7fa7b7372c Work on loop, split and sampling. 4 年之前
  Stanislaw Adaszewski 2260f55399 Add test_merge_pos_neg_batch_01/02(). 4 年之前
  Stanislaw Adaszewski cee0c23bd7 Add test_cumcount. 4 年之前
  Stanislaw Adaszewski ebcd38c910 Add first triacontagon experiment. 4 年之前
  Stanislaw Adaszewski 4c8c06c63c Add triacontagon.TrainLoop. 4 年之前
  Stanislaw Adaszewski c206df638a Add tests for _same_data_org() and DualBatcher. 4 年之前
  Stanislaw Adaszewski 6b15ec8c10 Finish DualBatcher. 4 年之前
  Stanislaw Adaszewski 0a061af526 Attempt DualBatcher. 4 年之前
  Stanislaw Adaszewski 8f86fc0833 More batcher tests. 4 年之前
  Stanislaw Adaszewski 14e3d14b36 Add Batcher. 4 年之前
  Stanislaw Adaszewski 56fdd9ffeb Make negative sampling more correct and more efficient at the same time. 4 年之前
  Stanislaw Adaszewski bf39e7d71f Small fix. 4 年之前
  Stanislaw Adaszewski 4ed3715b83 Add cumcount(). 4 年之前
  Stanislaw Adaszewski e557b18762 Add trainprep and negative_sample_(adj_mat|data)(). 4 年之前
  Stanislaw Adaszewski 7ea82a6267 Add test_model_decode_01(). 4 年之前
  Stanislaw Adaszewski b1a5f0d0ee Add common_one_hot_encoding(). 4 年之前
  Stanislaw Adaszewski 82d5c06eee Fix aggregation of representations from different edge types. 4 年之前
  Stanislaw Adaszewski bd8143469c Make work convolve() in new model. 4 年之前
  Stanislaw Adaszewski b607a727ae Rename _per_layer_required_rows to _vertices. 4 年之前
  Stanislaw Adaszewski 4ed6626e02 Work on required vertices per layer. 4 年之前
  Stanislaw Adaszewski f8a02191cb _clear_adjacency_matrix_except_rows() might yet be fast enough on the GPU. 4 年之前