 rene
|
826357a41f
|
refactor network models; remove depths
|
2017-11-07 20:32:08 +01:00 |
|
|
rene
|
e12bbda8c5
|
move model creation back into models package
|
2017-11-07 20:09:20 +01:00 |
|
|
rene
|
b1f48c1895
|
add soft parameter sharing network
|
2017-11-06 21:51:49 +01:00 |
|
|
rene
|
7b8dfcebbe
|
add long final implementation
|
2017-11-05 22:52:50 +01:00 |
|
|
rene
|
14fef66a55
|
train multiple models at once
|
2017-11-04 17:58:21 +01:00 |
|
|
rene
|
88e3eda595
|
refactor hyperband; fix domain generation
integrate hyperband option in training procedure - start refactoring - remove the index erro in generation and add helper functions
|
2017-11-04 12:47:08 +01:00 |
|
|
rene
|
8b17bd0701
|
add TSNE embedding; server evaluation visualization
|
2017-10-19 17:39:37 +02:00 |
|
|
rene
|
a860f0da34
|
refactor server training into separate file; add additional info to hyperband log
|
2017-10-19 17:37:29 +02:00 |
|
|
rene
|
d1da3d6ca3
|
fix model selection
|
2017-10-09 15:10:15 +02:00 |
|
|
rene
|
a686f147f0
|
add bulk embedding visualization and deep1 network
|
2017-10-09 14:19:01 +02:00 |
|
|
rene
|
33063f3081
|
add sample weight metrics to fit function
|
2017-10-08 22:09:09 +02:00 |
|
|
rene
|
e8473048cb
|
refactor visualization; start plotting server results
|
2017-10-08 11:52:10 +02:00 |
|
|
rene
|
0b26c6125c
|
fix flat 2 model bug caused by changing input layer
|
2017-10-06 16:36:58 +02:00 |
|
|
rene
|
5741f8ee0e
|
fix staggered training
|
2017-10-06 10:38:00 +02:00 |
|
|
rene
|
508667d1d0
|
add server classification model
|
2017-10-05 15:26:53 +02:00 |
|
|
rene
|
345afbaef5
|
add error bar vis, comment unused parameters from parser
|
2017-10-05 14:50:59 +02:00 |
|
|
rene
|
b24fa770f9
|
change hyperband to count minimal val_loss over all losses
|
2017-10-05 12:55:46 +02:00 |
|
|
rene
|
371a1dad05
|
add hyperband savefile config, minor change of parameter name
|
2017-10-03 18:58:54 +02:00 |
|
|
rene
|
68254d6629
|
add load function for hyper band results
|
2017-10-02 07:34:04 +02:00 |
|
|
rene
|
f02e0b7f99
|
fix hyperband wrong variable names
|
2017-09-29 23:34:39 +02:00 |
|
|
rene
|
a1e553f8f1
|
fix rerun wrong path
|
2017-09-29 23:03:07 +02:00 |
|
|
rene
|
605447440f
|
refactor hyperband implementation
|
2017-09-29 22:59:57 +02:00 |
|
|
rene
|
090c89a127
|
add retrain mode
|
2017-09-28 12:23:22 +02:00 |
|
|
rene
|
b157ca6a19
|
add first version of model averaging visualization
|
2017-09-26 19:25:37 +02:00 |
|
|
rene
|
49ad506a96
|
remove file name in run script
|
2017-09-22 10:03:01 +02:00 |
|
|
rene
|
607d74998c
|
add tsne (does not work with big data)
fix model loading with custom selu function
|
2017-09-22 10:01:12 +02:00 |
|
|
rene
|
e2bf2dc90f
|
fix missing parameters, add flat network structure, make larger graphics
|
2017-09-20 14:43:28 +02:00 |
|
|
rene
|
fbe6d6a584
|
remove input shape of first conv layer in networks because unnecessary
add selu activation to deeper network designs
|
2017-09-17 17:26:09 +02:00 |
|
|
rene
|
6a47b5f245
|
add sample weights function to utils for later usage
|
2017-09-17 10:23:23 +02:00 |
|
|
rene
|
f2845e635e
|
fix validation split removal by loading h5data into memory
|
2017-09-17 09:56:18 +02:00 |
|
|
rene
|
ec5a1101be
|
remove model selection based on validation loss
|
2017-09-16 15:25:34 +02:00 |
|
|
rene
|
b0e0cd904e
|
add start parameter for run script
|
2017-09-15 11:12:42 +02:00 |
|
|
rene
|
7f49021a63
|
refactor training - separate staggered training; make differences as small as possible
|
2017-09-12 08:36:23 +02:00 |
|
|
rene
|
6ce8fb464f
|
remove reguliarizer on dense too
|
2017-09-11 16:48:49 +02:00 |
|
|
rene
|
3a44efa775
|
minor changes for pauls test
|
2017-09-11 12:59:13 +02:00 |
|
|
rene
|
6d8d7b19f3
|
fix covariance normalization; add run_model script for multi times training
|
2017-09-11 12:42:44 +02:00 |
|
|
rene
|
6121eac448
|
remove regularizer for conv and domain
|
2017-09-10 23:40:14 +02:00 |
|
|
rene
|
1cf62423e1
|
add regularization to small networks, fix model name in args, fix visualizations
|
2017-09-10 18:06:40 +02:00 |
|
|
rene
|
6fef2b8b84
|
refactor all visualization for pauls changes - evaluate on max windows per users
|
2017-09-08 22:59:55 +02:00 |
|
|
rene
|
9a51b6ea34
|
refactor test function working on full unfiltered data
|
2017-09-08 19:10:23 +02:00 |
|
|
rene
|
edc75f4f44
|
refactor dataset creation, split up functions
|
2017-09-08 17:11:13 +02:00 |
|
|
rene
|
528829bb33
|
fix file name in run script
|
2017-09-08 13:57:18 +02:00 |
|
|
rene
|
70d00efb01
|
refactor using joblib for test results, make h5py store/load more flexible
|
2017-09-08 13:55:13 +02:00 |
|
|
rene
|
1ab0108c78
|
add window to file names for visualization
|
2017-09-07 17:38:21 +02:00 |
|
|
rene
|
595c2ea894
|
change argument interface
- add more properties for network specification
- change names for consistency
|
2017-09-07 15:53:58 +02:00 |
|
|
rene
|
71f218888d
|
set server to be not trainable too; refactor visualization script
|
2017-09-07 15:31:04 +02:00 |
|
|
rene
|
5bd8e41711
|
add staggered model training for intermediate sever prediction; refactor model return values
|
2017-09-07 14:24:55 +02:00 |
|
|
rene
|
2080444fb7
|
add parser argument for naming in multi model modes, minor fixes, re-set fix vals for network - need to make them flexible
|
2017-09-05 17:40:57 +02:00 |
|
|
rene
|
ed4f478bad
|
add parser argument for naming in multi model modes, minor fixes,
|
2017-09-05 12:40:37 +02:00 |
|
|
rene
|
1da31cc97c
|
visualize per user stats
|
2017-09-04 13:37:26 +02:00 |
|
|
rene
|
3f6779fa3d
|
load names with data for per-user evaluation
|
2017-09-02 16:02:48 +02:00 |
|
|
rene
|
0db8427457
|
update test script with bash loop
|
2017-09-02 12:58:48 +02:00 |
|
|
rene
|
dc9180da10
|
refactor visualization, change arguments for model type and its depth
|
2017-09-01 10:42:26 +02:00 |
|
|
rene
|
933eaae04a
|
change exception type in get_flow_per_user function and replace index to new range index
|
2017-08-31 13:49:33 +02:00 |
|
|
rene
|
dceaf47211
|
fix typo
|
2017-08-31 11:25:49 +02:00 |
|
|
rene
|
954dfcf9f9
|
fix run scripts
|
2017-08-05 09:40:40 +02:00 |
|
|
rene
|
5a02f582cd
|
change model - add dense before server output in new model
add some new run scripts
|
2017-08-05 09:33:07 +02:00 |
|
|
rene
|
6e7dc1297c
|
fix lazy domain loading and generation process
|
2017-08-03 12:27:17 +02:00 |
|
|
rene
|
7f1d13658f
|
store domain embeddings while test main
|
2017-08-03 09:08:24 +02:00 |
|
|
rene
|
452f9e0456
|
fix array() to zeros(0)
|
2017-08-03 08:00:20 +02:00 |
|
|
rene
|
787f43b328
|
fix test predictions depending on model output specification
|
2017-08-03 07:51:58 +02:00 |
|
|
rene
|
8ac195ba6f
|
add some static configs to run script
|
2017-08-02 15:13:09 +02:00 |
|
|
rene
|
1e781d5491
|
add argument for model outputs
BUG: need to check --new_model --model_output server
|
2017-08-02 12:58:09 +02:00 |
|
|
rene
|
f4da147688
|
refactor cmd argument to have single value for mode
|
2017-07-30 15:49:37 +02:00 |
|
|
rene
|
e24f596f40
|
add argument for using the new model architecture
|
2017-07-30 14:07:39 +02:00 |
|
|
rene
|
ebaeb6b96e
|
move vocab_size into implementation (not user dependent)
|
2017-07-30 13:47:11 +02:00 |
|
|
rene
|
d97785f646
|
replace softmax by sigmoid in final layer, also adjust dataset for that
|
2017-07-30 12:50:26 +02:00 |
|
|
rene
|
b0da2de0ea
|
move utils functions to new file
|
2017-07-29 19:47:02 +02:00 |
|
|
rene
|
820a5d1a4d
|
add new network architecture - server label moves to the middle
|
2017-07-29 19:42:36 +02:00 |
|
|
rene
|
8cd1023165
|
replace pca reduction by sklearn's truncated svd
|
2017-07-29 19:41:14 +02:00 |
|
|
rene
|
2593131e9e
|
add embedding visualization and domain encoding generator
|
2017-07-29 10:43:59 +02:00 |
|
|
rene
|
c1535b941b
|
remove max pooling from models for better infromation flow
|
2017-07-28 17:25:08 +02:00 |
|
|
rene
|
18b60e1754
|
add extended test mode for embeddings
|
2017-07-17 19:30:56 +02:00 |
|
|
rene
|
79fc441fe1
|
wip
|
2017-07-17 08:44:58 +02:00 |
|
|
rene
|
d33c9f44ec
|
fix chunks per user function bug caused by numpy version of array_split
|
2017-07-16 18:49:14 +02:00 |
|
|
rene
|
844494eca9
|
add multi-threading for pre-processing
|
2017-07-16 09:42:52 +02:00 |
|
|
rene
|
336be37032
|
fix network props, add PCA to visualize main
|
2017-07-14 21:01:08 +02:00 |
|
|
rene
|
6b787792db
|
add custom class weights based on sklearn balance
|
2017-07-14 15:57:52 +02:00 |
|
|
rene
|
b35f23e518
|
add visualization for training curves, pr, roc
|
2017-07-14 14:58:17 +02:00 |
|
|
rene
|
d0418b9efa
|
fix hyperband adding h5data and remove param sample
|
2017-07-12 15:18:45 +02:00 |
|
|
rene
|
2afaccc84b
|
refactor argparser into separate file, add logger
|
2017-07-12 10:25:55 +02:00 |
|
|
rene
|
9f0bae33d5
|
refactor dataset generation, add callbacks
|
2017-07-11 21:06:58 +02:00 |
|
|
rene
|
a196daa895
|
add simple flow feature extraction function
|
2017-07-11 13:46:25 +02:00 |
|
|
rene
|
522854ee0d
|
add h5 support for pauls best config main
|
2017-07-11 11:12:03 +02:00 |
|
|
rene
|
41b38de1ab
|
add feature: generate and use h5 data
|
2017-07-09 23:58:08 +02:00 |
|
|
rene
|
fdc03c9922
|
add h5py example
|
2017-07-08 17:46:07 +02:00 |
|
|
rene
|
4a9f94a029
|
add output for main_test
|
2017-07-08 15:04:58 +02:00 |
|
|
rene
|
21b9d7be73
|
merged
|
2017-07-08 13:04:07 +02:00 |
|
|
rene
|
36cdba3fdf
|
add pauls config test (TMP)
|
2017-07-08 13:01:58 +02:00 |
|
|
rene
|
a3324b5e04
|
add pauls config test (TMP)
|
2017-07-08 11:53:51 +02:00 |
|
|
rene
|
be56112b33
|
added params
|
2017-07-07 16:59:08 +02:00 |
|
|
rene
|
3c4be52bb6
|
refactor main functions - separate things into different functions
|
2017-07-07 08:43:16 +02:00 |
|
|
rene
|
933f6bf1d7
|
add feature to use both hits information from dataset
|
2017-07-06 16:27:47 +02:00 |
|
|
rene
|
b2f5c56019
|
refactor dataset generation
|
2017-07-05 21:19:19 +02:00 |
|
|
rene
|
772b07847f
|
WPI
|
2017-07-05 19:16:03 +02:00 |
|
|
rene
|
a70d1cb03a
|
fix: replace X_tr by its elements; choose selected samples for training data too
|
2017-07-05 18:37:29 +02:00 |
|
|
rene
|
7c05ef6a12
|
refactor models package: create separate modules for pauls and renes networks
|
2017-07-05 18:10:22 +02:00 |
|
|
rene
|
3862dce975
|
added new networks for domain embedding and classification task
|
2017-07-05 17:37:08 +02:00 |
|
|
rene
|
59c1176e85
|
refactor network to use new input format
|
2017-07-04 20:42:48 +02:00 |
|
|
rene
|
5743127b7f
|
new dataset format: multi-lists -> two arrays
|
2017-07-04 09:18:50 +02:00 |
|