Commits · d33a0b42c33470913e2a1c4655a70aa209116401 · ModelZoo / ResNet50_tensorflow

"third_party/kaldi/README.md" did not exist on "9e58e75cfd8eae48b36165da1fd9812f3f595197"

06 Feb, 2019 1 commit
- [LFADS] Fixes typo in distributions.py (#6161) · e42c769c
  Niru Maheswaranathan authored Feb 06, 2019
  
  e42c769c
18 Jun, 2018 1 commit
- Cleaning comments for posterior_push_mean · 20da056d
  Dan O'Shea authored Jun 18, 2018
  
  20da056d
24 May, 2018 1 commit
- Removing defaultless case statement · 8c5c60ca
  Dan O'Shea authored May 24, 2018
  
  8c5c60ca
22 Mar, 2018 1 commit
- Support for --kind=posterior_push_mean altnerative to sample and average · 32daecc9
  Dan O'Shea authored Mar 22, 2018
  
  32daecc9
03 Feb, 2018 5 commits
- Adjusting init_linear comment · 4070fa63
  Dan O'Shea authored Feb 02, 2018
  
  4070fa63
- Restoring old do_bias=False logic, keeping trainable option · 2c9a1100
  Dan O'Shea authored Feb 02, 2018
  
  2c9a1100
- Fixing init_linear with do_bias false, formatting · 0c1d2961
  Dan O'Shea authored Feb 02, 2018
  
  0c1d2961
- Formatting fixes · b1354256
  Dan O'Shea authored Feb 02, 2018
  
  b1354256
- lfads: Adding do_train_readin option defaulting to true, allowing fixed readin matrices · 32afad9c
  Dan O'Shea authored Feb 02, 2018
```
For stitched models, the readin matrices and bias vectors are initialized to the
"alignment" matrix and bias specified in each dataset's .h5 file. If do_train_readin is
True, these will be trainable, and if not, they will be fixed.
```
  32afad9c
20 Jan, 2018 1 commit
- from six.moves import xrange (en masse) · 848cc592
  cclauss authored Jan 20, 2018
  
  848cc592
26 Nov, 2017 1 commit
- change to lfads to allow training of encoder weights only · 9b3a7754
  Chethan Pandarinath authored Nov 25, 2017
  
  9b3a7754
21 Sep, 2017 1 commit
- Move the research models into a research subfolder (#2430) · f87a58cd
  Neal Wu authored Sep 21, 2017
  
  f87a58cd