aboutsummaryrefslogtreecommitdiff
path: root/config
Commit message (Collapse)AuthorAgeFilesLines
* Add memory network config files (not sure how usefull they are)Alex Auvolat2015-07-142-0/+114
|
* Batch shufflingAlex Auvolat2015-07-101-2/+2
|
* Add batch shuffle preprocessing stepAlex Auvolat2015-07-061-0/+40
|
* Add small modelsAlex Auvolat2015-07-023-2/+79
|
* Remove _simpleÉtienne Simon2015-07-0218-17/+17
|
* Step rule & dropout params cleanupAlex Auvolat2015-07-021-2/+5
|
* Unify parameters for joint_simple_tgtcls_111_cswdtx_bigger{,_dropout}Alex Auvolat2015-07-023-9/+8
|
* Load model parameters only when specified in the configAdeB2015-06-241-0/+2
|
* New configs. training step rule out of train.pyAdeB2015-06-242-11/+57
|
* Update memory networkÉtienne Simon2015-06-221-2/+3
|
* Fix candidate_encoder dim_inputÉtienne Simon2015-06-211-1/+1
|
* Add bugged memory networksÉtienne Simon2015-06-211-0/+43
|
* Model with only embeddings.AdeB2015-06-211-0/+31
|
* Add basic RNNÉtienne Simon2015-06-111-0/+33
|
* Remove useless n_valid in configs.Étienne Simon2015-05-2717-34/+0
|
* Add parametrizability for how the training data is presentedAlex Auvolat2015-05-2217-1/+21
|
* Fix embeddings sizes.Étienne Simon2015-05-228-16/+16
|
* Remove useless n_end_pts from configs.Étienne Simon2015-05-2117-17/+0
|
* Fix new models.Alex Auvolat2015-05-213-9/+12
|
* Merge branch 'alex' into newAlex Auvolat2015-05-213-0/+170
|\ | | | | | | | | | | Conflicts: model/joint_simple_mlp_tgtcls.py train.py
| * Model changesAlex Auvolat2015-05-213-0/+170
| |
* | Use os.path.join and close filesÉtienne Simon2015-05-188-8/+16
| |
* | Refactor models, clean the code and separate training from testing.Étienne Simon2015-05-1814-21/+83
|/
* Add model with hidden layers specific to time/dest prediction.Alex Auvolat2015-05-081-0/+55
|
* Add scaling factor between two costs in joint modelAlex Auvolat2015-05-081-0/+3
|
* Add model for a network that predicts both time and destination.Alex Auvolat2015-05-081-0/+52
|
* Fusion AddFirstK and AddLastK transformers, and add 'input_time' field.Alex Auvolat2015-05-081-2/+0
|
* Add target class based model for time prediction (seems to work)Alex Auvolat2015-05-071-0/+41
|
* Move weights init to config files ; fix s/time/travel_timeAlex Auvolat2015-05-0610-1/+69
|
* Adapt Alexandre's model to the new interfaceÉtienne Simon2015-05-051-3/+5
|
* Merge branch 'master' of github.com:adbrebs/taxiAdeB2015-05-057-16/+30
|\
| * Clean data module and generalize use of hdf5.Étienne Simon2015-05-057-16/+30
| |
* | New hyperparameters. Training error is monitored.AdeB2015-05-051-0/+30
|/
* Add models for time predictioAdd models for time predictionAlex Auvolat2015-05-059-7/+52
|
* Add day type and taxi idAlex Auvolat2015-05-053-2/+58
|
* Add date/time transformer and new model that uses itAlex Auvolat2015-05-055-16/+28
|
* Refactor architecture so that embedding sizes can be easily changed.Alex Auvolat2015-05-054-6/+42
|
* New config (added a hidden layer), small changes to train.pyAlex Auvolat2015-05-051-0/+25
|
* Fix itAlex Auvolat2015-05-041-1/+1
|
* Add model with predefined target classesAlex Auvolat2015-05-041-0/+25
|
* Restructure model & configAlex Auvolat2015-05-041-0/+2
|
* Mew data analysis tool: clustering of arrival points.Alex Auvolat2015-05-041-2/+2
|
* FixesAlex Auvolat2015-05-041-1/+1
|
* FixAlex Auvolat2015-04-291-1/+1
|
* Fix CSV import (partially)Alex Auvolat2015-04-291-5/+1
|
* Adapt model to hdf5 dataset. WIPÉtienne Simon2015-04-292-0/+21