diff options
| author | Navan Chauhan <navanchauhan@gmail.com> | 2020-09-24 15:53:54 +0530 |
|---|---|---|
| committer | Navan Chauhan <navanchauhan@gmail.com> | 2020-09-24 15:53:54 +0530 |
| commit | 6255d82fa7f35d4087e870b5df8f58ecdae1b17e (patch) | |
| tree | b933d29eb1603b958485f268642017f47c7936fe | |
| parent | f15aeecedf794c26752a38ff0a142ffa29868db3 (diff) | |
added instruction about tensorflow
| -rw-r--r-- | README.md | 16 |
1 files changed, 15 insertions, 1 deletions
@@ -10,7 +10,7 @@ Tested on: # Quick Start (Docker-Compose) -You can quickly get started and test Curie-Web without needing to manage dependencies by using the `docker-compose` image. This has all features except AR/3D Model support and should be production ready. +You can quickly get started and test Curie-Web without needing to manage dependencies by using the `docker-compose` image. This has all features except AR/3D Model support and LSTM Generator. Otherwise, it is production ready. ``` @@ -55,6 +55,12 @@ macOS users can use Homebrew to install it via `brew install open-babel` Users using apt can install it via `sudo apt install openbabel python3-openbabel` +### 1.5 Tensorflow (Optional) + +Please ensure you install Tensorflow 2.x. + +If tensorflow is not properly installed, it will automatically disable the LSTM Module. + ## 2. Changing the Configuration Replace the values in `config.ini` @@ -176,3 +182,11 @@ python3 removeSample.py * To enable the server to start on boot run `sudo systemctl enable curie` + +## Misc. + +### Fine-Tuning the LSTM Module + +Currently the base model has been trained on 1.5 million compounds and took ~4 hours to train on Google Colab's GPU runtime. + +*Information to be added soon*
\ No newline at end of file |