Commits

Branch selector

 master

User selector

carpedm20

Datepicker

All time

Commit History

Commits on Mar 10, 2017

• Merge pull request devsisters#26 from rsnk96/master

  Show description for 3fbfc59
  carpedm20

  authored
  3fbfc59

  Copy full SHA for 3fbfc59

Commits on Jan 24, 2017

• Merge pull request devsisters#23 from WillBrennan/bug/issue-10

  Show description for a2b1143
  carpedm20

  authored
  a2b1143

  Copy full SHA for a2b1143

Commits on Jan 6, 2017

• increase TF version compatibility from 0.8.0 to 0.12.0

  carpedm20

  committed
  460d882

  Copy full SHA for 460d882

Commits on Dec 19, 2016

• specify the verion of tf which is used for this code

  carpedm20

  committed
  aceeade

  Copy full SHA for aceeade

• fix error on q error clippling with devsisters#16

  carpedm20

  committed
  6d11105

  Copy full SHA for 6d11105

Commits on Oct 3, 2016

• Merge pull request devsisters#11 from pemami4911/HW-fix

  Show description for 220a7c2
  carpedm20

  authored
  220a7c2

  Copy full SHA for 220a7c2

Commits on Jul 1, 2016

• updates for asynchronous q-learning

  carpedm20

  committed
  2f37d23

  Copy full SHA for 2f37d23

Commits on Jun 29, 2016

• increase test step by 5

  carpedm20

  committed
  c672b9c

  Copy full SHA for c672b9c

• Merge pull request devsisters#8 from david4096/master

  Show description for fcab5e8
  carpedm20

  authored
  fcab5e8

  Copy full SHA for fcab5e8

• Merge branch 'master' of github.com:devsisters/DQN-tensorflow

  carpedm20

  committed
  7a862f6

  Copy full SHA for 7a862f6

• add double_Q and duel but hyperparameter and gradient clipping should be added

  carpedm20

  committed
  9f54532

  Copy full SHA for 9f54532

• fix README.md

  carpedm20

  committed
  0df29da

  Copy full SHA for 0df29da

• update README.md

  carpedm20

  committed
  64197f9

  Copy full SHA for 64197f9

Commits on Jun 28, 2016

• fix README.md

  carpedm20

  committed
  474b757

  Copy full SHA for 474b757

Commits on Jun 27, 2016

• update fix

  carpedm20

  committed
  28e973f

  Copy full SHA for 28e973f

• Revert to 557a57f

  carpedm20

  committed
  da73c97

  Copy full SHA for da73c97

Commits on Jun 23, 2016

• should reduce starting learning_rate by 10

  carpedm20

  committed
  9db8d97

  Copy full SHA for 9db8d97

Commits on Jun 22, 2016

• np.maximum efficiently

  carpedm20

  committed
  f1011b2

  Copy full SHA for f1011b2

• fix gpu configure

  carpedm20

  committed
  d7aa213

  Copy full SHA for d7aa213

Commits on Jun 21, 2016

• add option for config.minus_one_if_dead

  carpedm20

  committed
  eea7ee5

  Copy full SHA for eea7ee5

• glitching fix should be done in single small step, not repeated action level

  carpedm20

  committed
  45ffe6b

  Copy full SHA for 45ffe6b

• remove -1 reward when terminated. this spoils the training

  carpedm20

  committed
  a7d769c

  Copy full SHA for a7d769c

• no manual -1 reward

  carpedm20

  committed
  42dabb3

  Copy full SHA for 42dabb3

Commits on Jun 20, 2016

• fix environment error

  carpedm20

  committed
  d582311

  Copy full SHA for d582311

• add environment screen maxing for avoid glitch

  carpedm20

  committed
  15dbba9

  Copy full SHA for 15dbba9

• fix env error

  carpedm20

  committed
  e2ff644

  Copy full SHA for e2ff644

• fix error for double q-learning

  carpedm20

  committed
  8a0b943

  Copy full SHA for 8a0b943

• add double_q, dueling option to main.py

  carpedm20

  committed
  8b410d7

  Copy full SHA for 8b410d7

• add double_q and dueling network

  carpedm20

  committed
  25ec07a

  Copy full SHA for 25ec07a

• update graph, fix devsisters#4

  carpedm20

  committed
  12a13df

  Copy full SHA for 12a13df

• Merge branch 'master' of github.com:devsisters/DQN-tensorflow

  carpedm20

  committed
  d3a5803

  Copy full SHA for d3a5803

• update main.py add newest checkpoints

  carpedm20

  committed
  0e3f756

  Copy full SHA for 0e3f756

• change board tag

  carpedm20

  committed
  557a57f

  Copy full SHA for 557a57f

Commits on Jun 17, 2016

• change board tag

  carpedm20

  committed
  b162860

  Copy full SHA for b162860

Commits on Jun 5, 2016

• padding: SAME to VALID, reduce learning_decay_step

  carpedm20

  committed
  2d23a08

  Copy full SHA for 2d23a08

Pagination

Previous

Next