Commit Graph

Select branches

Hide Pull Requests

master

Mono Color

• 73d2572fbe
  
  Fix typo in README master Mattéo Delabre 2019-12-18 11:26:57 -0500
• 6921be6ca6
  
  Augment illustration Mattéo Delabre 2019-12-18 11:21:23 -0500
• 54813c4c46
  
  Fixed name clash in plot.py Mattéo Delabre 2019-12-18 11:21:14 -0500
• 376c7c0bc4
  
  Add illustration Mattéo Delabre 2019-12-18 11:04:58 -0500
• cacf4408e5
  
  Add README Mattéo Delabre 2019-12-18 11:02:03 -0500
• af8862d360
  
  Reflect parameter changes in `experiment.py` Mattéo Delabre 2019-12-18 11:01:45 -0500
• d8a140d793
  
  Simplify arguments for ConstellationNet Mattéo Delabre 2019-12-18 10:26:35 -0500
• a7e9dd2230
  
  Save whole model to avoid definition duplicate Mattéo Delabre 2019-12-18 10:17:59 -0500
• 0769a61fcf
  
  Use best parameters as found by experimentation Mattéo Delabre 2019-12-18 09:41:50 -0500
• 5cb087d971
  
  Add experiment results Mattéo Delabre 2019-12-17 01:47:54 -0500
• fe23fda86d
  
  Shuffle configurations in experiment Mattéo Delabre 2019-12-16 13:32:32 -0500
• 8d64f91661
  
  Add partitioning to experiment Mattéo Delabre 2019-12-16 13:27:43 -0500
• 0dfbed1bb7
  
  Add experiment code Mattéo Delabre 2019-12-16 12:59:39 -0500
• 7efcbd3948
  
  Add product_dict util function Mattéo Delabre 2019-12-16 11:32:25 -0500
• e8e4dcec83
  
  Set learning rate to 10^-3 Mattéo Delabre 2019-12-16 10:31:29 -0500
• af59421dc0
  
  Fix variable naming clash Mattéo Delabre 2019-12-16 10:31:16 -0500
• fb2518b321
  
  Show final loss after training Mattéo Delabre 2019-12-16 10:20:01 -0500
• 1d39184036
  
  Use euclidean distance for stop criterion Mattéo Delabre 2019-12-16 09:55:06 -0500
• 4f1514accf
  
  Add noise to plot Mattéo Delabre 2019-12-16 09:22:24 -0500
• 2c4786bdba
  
  Remove reconstruction examples after training Mattéo Delabre 2019-12-16 09:17:15 -0500
• 4bf0c0f363
  
  Add batch normalization and ReLU activation Mattéo Delabre 2019-12-16 02:35:13 -0500
• 8fa6b46ca8
  
  Change training strategy to adaptive learning rate Mattéo Delabre 2019-12-16 02:30:05 -0500
• 3f2c6d18a3
  
  Manually seed training for reproducibility Mattéo Delabre 2019-12-16 01:50:03 -0500
• 4d8d419d31
  
  Only plot noise if noise_samples > 0 Mattéo Delabre 2019-12-16 01:49:35 -0500
• 2b93a5f1bc
  
  Fix wrong input dimension for decision region plot Mattéo Delabre 2019-12-15 23:53:14 -0500
• e4457400a6
  
  Add larger set of colors for plot Mattéo Delabre 2019-12-15 23:52:57 -0500
• 722c02ef24
  
  Cleanup channel model Mattéo Delabre 2019-12-15 23:12:00 -0500
• 3638749998
  
  Share plotting code between plot.py and train.py Mattéo Delabre 2019-12-15 20:05:20 -0500
• 34cb3b863b
  
  Simplify channel model to account for normalized power Mattéo Delabre 2019-12-15 19:48:49 -0500
• 59d7adf6bd
  
  Add convergence criterion plot during train Mattéo Delabre 2019-12-15 19:42:50 -0500
• 3c199dfc41
  
  Remove activation function from ConstellationNet Mattéo Delabre 2019-12-15 19:22:35 -0500
• 989a51b72e
  
  Increase batch size halfway through training Mattéo Delabre 2019-12-15 09:42:48 -0500
• 3365603614
  
  Add normalization layer Mattéo Delabre 2019-12-15 09:42:33 -0500
• b97ba61f42
  
  Improve plot legibility Mattéo Delabre 2019-12-15 02:07:52 -0500
• 8a31b22b83
  
  Increase noise stddev Mattéo Delabre 2019-12-15 01:21:11 -0500
• 04c2938b9e
  
  Decrease alpha for noise plot Mattéo Delabre 2019-12-15 01:21:03 -0500
• 9a87b322d4
  
  Plot noise Mattéo Delabre 2019-12-15 01:07:05 -0500
• 6ea0e653c1
  
  Balance training examples Mattéo Delabre 2019-12-15 00:59:10 -0500
• 99c96162c0
  
  Use SELU instead of ReLU Mattéo Delabre 2019-12-15 00:57:31 -0500
• 6d31ab3a13
  
  Use ReLU with normalization Mattéo Delabre 2019-12-15 00:03:20 -0500
• 197a01e993
  
  Fix plot scaling Mattéo Delabre 2019-12-15 00:03:02 -0500
• 8f6363ee21
  
  Plot decision regions from the decoder Mattéo Delabre 2019-12-14 23:31:47 -0500
• 3b40e27070
  
  Add Gaussian channel model Mattéo Delabre 2019-12-14 23:04:35 -0500
• d0af8fc3da
  
  Use Tanh activation to preserve negative values Mattéo Delabre 2019-12-14 23:03:02 -0500
• a01d83f339
  
  Improve plot Mattéo Delabre 2019-12-13 17:10:40 -0500
• 49e63775dd
  
  Save trained models and plot encoding Mattéo Delabre 2019-12-13 15:17:57 -0500
• 2af8354a07
  
  Add initial implementation Mattéo Delabre 2019-12-13 12:11:09 -0500
• 7463e9fe5b
  
  Initial commit Mattéo Delabre 2019-12-11 19:13:40 -0500