jens/pyRBM
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Commit Graph

  • 186fd423b9 updated notebooks jens 2026-01-06 20:20:43 +01:00
  • 7d3ab02e26 fixed dividing l2-norm by length of vector jens 2026-01-06 18:17:45 +01:00
  • 542b8ae9b0 added L2-Regulation jens 2026-01-06 18:00:26 +01:00
  • ff6254a15d - Status has entity paarmeter - train provides entity parameter to Status jens 2026-01-06 11:29:47 +01:00
  • 42520e5761 improved tests jens 2026-01-06 09:44:45 +01:00
  • ec5d259ff7 Entity: added property enable training jens 2026-01-06 09:44:29 +01:00
  • b1fe7229fa updated cifar_test jens 2026-01-05 21:46:33 +01:00
  • b6a3779bf9 updated cifar_test and mnist_test jens 2026-01-05 21:39:22 +01:00
  • 74138b5514 train: swapped epochs and mini-batch loop jens 2026-01-05 20:53:31 +01:00
  • 562e9e9bda added mnist_test jens 2026-01-05 19:42:01 +01:00
  • ae6ab9b9b5 updated cifar_test jens 2026-01-05 19:41:43 +01:00
  • d238d2954e grad_compute(): pre-multiply weight decay with learning rate jens 2026-01-05 19:40:05 +01:00
  • 10999ec731 updated notebook jens 2026-01-05 18:08:22 +01:00
  • 2fd55ee69c updated notebook jens 2026-01-05 14:49:03 +01:00
  • fdf4186d6b significantly improved GB-RBM training jens 2026-01-05 14:16:34 +01:00
  • 9f72c1f253 - added cifar_test as jupiter lab - started torch version of RBM - enable CUDA in matrix jens 2026-01-04 18:30:01 +01:00
  • 73d4486fb7 test_norbs: added mode hidden gaussian - hidden binary jens 2026-01-04 16:06:41 +01:00
  • 6546c37657 train: fixed gaussian sampling jens 2026-01-03 19:30:47 +01:00
  • e6c959d890 - fixed test_xor - added test_linear jens 2026-01-03 13:07:08 +01:00
  • e56f99535a improved tests jens 2026-01-03 12:51:43 +01:00
  • 529252166a train: sample gaussian visible units jens 2026-01-03 12:40:08 +01:00
  • 63bfd3ed80 updated jens 2026-01-03 12:11:04 +01:00
  • f828c26f36 refactored jens 2026-01-03 12:10:53 +01:00
  • 94595ddd8a train: added cd_binary_gaussian jens 2026-01-03 12:10:17 +01:00
  • fa9f595be6 use gray color map jens 2026-01-02 22:50:03 +01:00
  • 082bb3c566 Label: added OneHote encoding jens 2026-01-02 22:49:00 +01:00
  • 2e714685d1 model: forward() is not abstract method jens 2026-01-02 19:42:22 +01:00
  • c87b8f8ed0 fixed jens 2026-01-02 19:41:26 +01:00
  • 48f3642e44 fixed TrainingParams jens 2026-01-02 18:36:17 +01:00
  • c7f820ec63 - adapted test to TrainingParameter as parat of Entity - learn_norbs_labes start working jens 2026-01-02 14:53:45 +01:00
  • 10691894d1 - training params are (again) attrubute of Entity - ditched layout concept jens 2026-01-02 14:51:19 +01:00
  • 8b97350564 cd_binary_binary: added gibbs sampling jens 2026-01-01 20:02:22 +01:00
  • e67207504b fixed cd_binary_binary jens 2026-01-01 15:55:35 +01:00
  • dc998d30a2 - initroduced types: BB-RBM GB-RBM, GG-RBM as property of entity - training simplification and for clarity: added specialized cd_funcs for each entity type jens 2026-01-01 12:05:39 +01:00
  • 38b834c640 - added training params as list for model.train() - added gaussian sample - introduced layout concept - updated README jens 2025-12-31 16:13:59 +01:00
  • 9b41dd6c02 added LabelLearner jens 2025-12-22 11:14:44 +01:00
  • 35e045a701 added label jens 2025-12-22 10:02:36 +01:00
  • 781870b971 add label encoder/decoder (not finished) jens 2025-12-21 20:20:24 +01:00
  • 362970574f refactored jens 2025-12-21 19:34:23 +01:00
  • ff05c729e8 - refactored read_armadillo into matrix - added test_norbs jens 2025-12-21 18:16:48 +01:00
  • e11f775aff Test model: removed unecessary import jens 2025-12-21 17:46:04 +01:00
  • b29937c149 Test model: improved output log jens 2025-12-21 17:45:10 +01:00
  • 4bd5cffd6b - state load/save revised - num gibbs samples is also part of entity - working 3-layer deep test model jens 2025-12-21 17:34:11 +01:00
  • 22d189cf2f added model jens 2025-12-21 16:39:53 +01:00
  • 7668c73ea4 refactored jens 2025-12-21 10:33:04 +01:00
  • 8ed6488b45 refactored train jens 2025-12-21 10:16:08 +01:00
  • 7f4505c1f8 Raise exception if doSampleBatch == True jens 2025-12-20 20:21:42 +01:00
  • 61c762150c - removed Optimizer - refactored training - use static seed for random (for better comparison) - simplified StackDeep.train jens 2025-12-20 20:12:19 +01:00
  • 5f5c7c6d77 Optimizer.train() shall ignore param.do_batch_sample jens 2025-12-19 17:29:32 +01:00
  • d3c9fe4681 - numGibbs no longer part of EntityParameter - refactored forward and reconstruct - conditionally use optimizer for training jens 2025-12-19 17:15:10 +01:00
  • a47922cb1c refactored jens 2025-12-19 15:20:04 +01:00
  • a293fc31a0 added layertest notebook jens 2025-12-19 12:25:17 +01:00
  • a9e67da43a refactored jens 2025-12-19 11:55:29 +01:00
  • 13a7cff094 layer: refactored xor() into test_xor.py jens 2025-12-19 11:39:55 +01:00
  • ed8274cc8b updated gitignore jens 2025-12-19 11:11:00 +01:00
  • cadbb34d1a fixed imports jens 2025-12-19 11:10:16 +01:00
  • a60f802443 refactored jens 2025-12-18 21:32:01 +01:00
  • b206ab5788 Entity: refactored taining algo into train module jens 2025-12-18 21:23:16 +01:00
  • 6664ebdb9b added shape to entity jens 2025-12-18 19:47:27 +01:00
  • c1c6c610ad - wrap numpy and cupy in matrix as np - added type alias Mat for np.ndarray jens 2025-12-18 18:14:03 +01:00
  • 9000e70607 - fixed training_remain - added total rms error after training jens 2025-12-18 16:46:33 +01:00
  • 3936a5be27 fixed progress indication jens 2025-12-18 16:36:47 +01:00
  • 1b51861daf fixed contamination of function parameter jens 2025-12-18 16:06:27 +01:00
  • 3b68b9ad79 - fixed v_to_ph parameter consideration - matrix: use different random implementations jens 2025-12-18 15:49:18 +01:00
  • 855c392970 - changed status interval - show number of remaining epochs at beginning of training jens 2025-12-18 12:49:17 +01:00
  • f6bccd3737 fixed deep stack status update jens 2025-12-18 12:31:32 +01:00
  • 5cbef6b09f refactored jens 2025-12-18 12:24:17 +01:00
  • 470d2bfa9a rbm: use update interval of 1% jens 2025-12-18 11:43:34 +01:00
  • 905ae8da81 - layer: only calculate rms_error_accu() if status report is necessary - Status: added wants_report jens 2025-12-18 11:39:45 +01:00
  • 28fee91ec7 fixed layer construction jens 2025-12-18 10:06:35 +01:00
  • 29839c082e added matplotlib jens 2025-12-18 07:56:45 +01:00
  • eb975da018 refactored jens 2025-12-18 07:56:08 +01:00
  • c06fc7f485 use argparser for project selection jens 2025-12-17 20:12:32 +01:00
  • b7f9bf66b0 use dot product for weights jens 2025-12-17 20:05:36 +01:00
  • 6c02e3c229 - use training data if test data not found jens 2025-12-17 19:33:14 +01:00
  • b4cd05d803 - fixed max index - call status once inside epoch loop jens 2025-12-17 19:29:46 +01:00
  • 8dcf0872b9 [Rbm] - abort by user [q] - cycle through training patetrn by [+],[-] keys [Status] - improved report interval - added on_report(): called every report iterval jens 2025-12-17 19:14:57 +01:00
  • 866b61c77b - refactored Status - added update interval jens 2025-12-17 18:52:18 +01:00
  • 4aba7585cc - refactored Rbm - added myStatus with realtime update of reconstruction - DeepStack: added up/down passes for multilayer Stacks jens 2025-12-17 18:43:08 +01:00
  • 6ef577c769 - fixed armadillo data read - added shape to layer - changed construction of layer from Stack factory - fixed rms_error_accu scaling - RBM: added image display using opencv - improved Status print jens 2025-12-17 17:05:40 +01:00
  • a1e7a803d1 fixed armadillo read jens 2025-12-17 14:04:26 +01:00
  • 4d05a59d0b added workdir parameter jens 2025-12-17 13:46:46 +01:00
  • 2c2625c778 Stack: added state load jens 2025-12-17 13:35:18 +01:00
  • 575ceb4ef1 revised status jens 2025-12-17 13:15:09 +01:00
  • 99657265af - added ready of armadillo dat-file (Quick and dirty) - added stack training for DeepStack jens 2025-12-17 12:43:10 +01:00
  • 3ec170234b revised stack jens 2025-12-17 11:31:13 +01:00
  • 80ba543323 - StackFactory removed default constructor - StackFactory added type hints jens 2025-12-17 10:53:19 +01:00
  • 437ffbdaad - add RbmParams.from_dict - added stack construction - added empty implementation of Deep Stack and Rnn Stack jens 2025-12-17 10:48:43 +01:00
  • 0bebf20bdd added Test passed message jens 2025-12-16 21:46:00 +01:00
  • 71b723e4e9 - added stack - refactored jens 2025-12-16 21:36:55 +01:00
  • f33df4b0e0 - fixed crash if weight not exists - optimized train loop - choose reasonable default params jens 2025-12-16 19:08:20 +01:00
  • 4a64279276 xor: add better documentation jens 2025-12-16 18:43:54 +01:00
  • 6a3c59999e xor: load and save weights jens 2025-12-16 18:40:51 +01:00
  • 5404f2bce0 fixed several training problems jens 2025-12-16 18:28:40 +01:00
  • b5c0c0e0f1 - added gibbs sampling for forward pass jens 2025-12-16 17:26:49 +01:00
  • e10004051c - fixed matrix operations - fixed rms error calculation jens 2025-12-16 17:03:23 +01:00
  • a3894f7c34 - added train - refactored CdTrain jens 2025-12-16 15:58:06 +01:00
  • 944b96448c refactored jens 2025-12-16 14:59:07 +01:00
  • b6e94e75ac Initial commit jens 2025-12-16 14:52:36 +01:00