To receive notifications about scheduled maintenance, please subscribe to the mailing-list gitlab-operations@sympa.ethz.ch. You can subscribe to the mailing-list at https://sympa.ethz.ch

Commit 52c56545 authored by zpgeng's avatar zpgeng
Browse files

Add final experimental results ver.2

parent 907806f7
......@@ -9,6 +9,21 @@ Predicting eye gaze with DL
Please use the command
`module load gcc/6.3.0 python_gpu/3.8.5`
`module load gcc/6.3.0 python_gpu/3.8.5 hdf5/1.10.1`
before training the model. Please make sure that the Tensorflow version should be 2.x.
## deepEye3 tuning
nb_filter: [32, 64]
depth: [9, 12, 20]
kernel_size:[40, 20]
residual_jump: [3, 4]
Large depth causes overfitting, same for the number of filters. Kernel size seems to have tiny affect on validation. Residual jump for 4 (i.e. `depth % (res_jump) == (res_jump - 1)`) is not so good in our task, but I think it would be useful for future tasks.
The best setting is **nb_filter == 32, depth == 9, kernel_size == 40, res_jump == 3**
before training the model. Please make sure that the Tensorflow version should be 2.x.
\ No newline at end of file
epoch;accuracy;loss;val_accuracy;val_loss
0;0.6818621158599854;0.5965843796730042;0.7106970548629761;0.561839759349823
1;0.7250673174858093;0.5425289869308472;0.7148377895355225;0.5786265134811401
2;0.7409759163856506;0.5219770669937134;0.7218771576881409;0.5637990236282349
3;0.7518462538719177;0.5057228803634644;0.7069703340530396;0.5961728096008301
4;0.7570570707321167;0.49716559052467346;0.7352656722068787;0.5476036667823792
5;0.7671682238578796;0.4800756573677063;0.7227053046226501;0.5658014416694641
6;0.7777624130249023;0.4625657796859741;0.7209109663963318;0.5598932504653931
7;0.7862516641616821;0.45215412974357605;0.7269841432571411;0.5753936171531677
8;0.7933949828147888;0.4397102892398834;0.7269841432571411;0.5786483883857727
9;0.8020567297935486;0.42472347617149353;0.7235334515571594;0.5906829833984375
10;0.8107184767723083;0.40943020582199097;0.7181504368782043;0.6123196482658386
11;0.8197253346443176;0.39387670159339905;0.7124913930892944;0.6097022891044617
12;0.8264890313148499;0.3807852268218994;0.7142857313156128;0.6146067976951599
13;0.8409137725830078;0.3588753342628479;0.7040717601776123;0.6386616230010986
14;0.8474360108375549;0.34541934728622437;0.7037957310676575;0.6494848132133484
15;0.8540962338447571;0.33202990889549255;0.7020013928413391;0.6702932715415955
16;0.861101508140564;0.3159162700176239;0.7135955691337585;0.6722756028175354
17;0.8691076040267944;0.29970842599868774;0.7033816576004028;0.7020342946052551
18;0.8756642937660217;0.2901645004749298;0.7039337754249573;0.6805644035339355
19;0.8815998435020447;0.27583715319633484;0.7044858336448669;0.7085856795310974
20;0.8892263174057007;0.2613828778266907;0.7044858336448669;0.7444139122962952
21;0.8955069184303284;0.24831454455852509;0.709868848323822;0.7950683832168579
22;0.8984401822090149;0.2393287569284439;0.7086266279220581;0.7413702011108398
23;0.9063427448272705;0.22508296370506287;0.7036576867103577;0.8459625244140625
24;0.9119676947593689;0.21429939568042755;0.7025535106658936;0.8548511266708374
25;0.9140382409095764;0.20880228281021118;0.7048999071121216;0.8876222968101501
26;0.9182138442993164;0.19922763109207153;0.6974465250968933;0.8612427115440369
27;0.917799711227417;0.19967049360275269;0.7035196423530579;0.8836870193481445
28;0.9221823215484619;0.18760919570922852;0.6945479512214661;0.8663473725318909
29;0.9256677627563477;0.18301357328891754;0.7004830837249756;0.9127931594848633
30;0.9268065690994263;0.17902716994285583;0.6930297017097473;0.9827239513397217
31;0.9294292330741882;0.1726013720035553;0.6897170543670654;0.9457045793533325
32;0.9324660301208496;0.1667884886264801;0.7008971571922302;0.8864448666572571
accuracy,loss,val_accuracy,val_loss
0.7570570707321167,0.49716559052467346,0.7352656722068787,0.5476036667823792
\ No newline at end of file
INFO:root:Started the Logging
INFO:root:X training loaded.
INFO:root:(129, 500, 36223)
INFO:root:y training loaded.
INFO:root:(1, 36223)
INFO:root:Setting the shapes
INFO:root:(36223, 500, 129)
INFO:root:(36223, 1)
INFO:root:Started running deepeye-lstm. If you want to run other methods please choose another model in the config.py file.
This diff is collapsed.
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment