GitLab

Co-authored-by: Tobias Morville <tmo@Tobiass-MacBook-Pro.local>

It might seem redundant, but I followed this whole tutorial not knowing what <path-to-tensorflow> meant, and tried to install into the actual "tensorflow" package inside of my python installation.

The URL of the script for downloading and converting ImageNet data to TFRecord format has changed. I update the URL.

PiperOrigin-RevId: 308369201

* Update directory path 

- Update directory path with Tensorflow directoy folder: tensorflow/models/research/
- Add option to refresh source ~/.bashrc

* Update installation.md

Updated owners

Updated owners

PiperOrigin-RevId: 308302588

Add axis=1 for to reconstruction error calculation. This will avoid taking the sum of cost so that the training objective will no longer be  batch-size dependent.

Fixes #2243

Updated owner

No Maintenance Intended

No Maintenance Intended

No Maintenance Intended

No Maintenance Intended

No Maintenance Intended

No Maintenance Intended

Added "No maintenance intended"

Added TensorFlow requirement badges

Added TensorFlow requirement badges

Added TensorFlow requirement badges

PiperOrigin-RevId: 308278540

* Fix maintainers list into a table for easy read.

* Taken duplicate name out of the maintainers table.

Fixed bug

Updated owners

Plus minor formatting proposals

typo fixed

* Guard the unsafe tf.exp to prevent Inf cost

* Update run_lfads.py

Add the hyperparameter `_clip_value` to resolve the issue in the prototype example

Fixed typo

* add scripts to generate training data

* add scripts to generate training data

* add missing POS tag mapping

* tweak flags to not rely on directory structure, fix issues with output.

* add blurbs for new programs

* oops, need to initialize the embeddings!

* Misspelled 'info'

* editing
Co-authored-by: Chris Waterson <waterson@google.com>