Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Make the README.md prettier #2

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Make the README.md prettier #2

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 6 additions & 6 deletions experiments/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,13 +8,13 @@ Requirements (the code was successfully ran using the following versions):

First, clone this repository:

```
```bash
git clone git@github.com:claravania/indo-nli.git
cd indo-nli
```

After that, install jiant under the current directory (better to use a virtual environment):
```
```bash
git clone -b indo-nli git@github.com:claravania/jiant.git
cd jiant
conda create -n jiant python=3.7
Expand All @@ -25,7 +25,7 @@ cd ..
```

The `experiments` folder contains some configuration files and datasets that are used for experiment. To add a new dataset, first make a similar directory under `experiments/tasks/data`, and then create a configuration file for that dataset and put it under `experiments/tasks/config`. If using dataset that has the same format as already implemented task (e.g., indo-nli), we can just write a config file for the new dataset but still using the same original task name. As an example, if we want to use an augmented training data for `indo_nli` (e.g., `indo_nli_augment_config.json`):
```
```json
{
"task": "indo_nli",
"paths": {
Expand All @@ -42,7 +42,7 @@ The `experiments` folder contains some configuration files and datasets that are
Notes: please check the script if you need to change some path to adjust your JIANT experimentation path.

To run an experiments, first we need to prepare (tokenize and cache the dataset). This is done using `run_preprocess.sh` script (change `indo_nli` to the new dataset name):
```
```bash
bash run_preprocess.sh indobenchmark/indobert-large-p2 indo_nli
bash run_preprocess.sh indobenchmark/indobert-lite-base-p2 indo_nli
bash run_preprocess.sh xlm-roberta-large indo_nli
Expand All @@ -51,7 +51,7 @@ bash run_preprocess.sh bert-base-multilingual-cased indo_nli

After that, we can run experiment using `runner.sh` script. To run a set of experiments (4 pretrained models, 3 random seeds) using the same best hyperparameter as `indo_nli`, you can run the following command (change `indo_nli` to the new dataset name):

```
```bash
bash runner.sh indobenchmark/indobert-lite-base-p2 indo_nli 1
bash runner.sh indobenchmark/indobert-lite-base-p2 indo_nli 11
bash runner.sh indobenchmark/indobert-lite-base-p2 indo_nli 12
Expand All @@ -70,7 +70,7 @@ bash runner.sh xlm-roberta-large indo_nli 32

Once all models are trained, you can use `call_predict.py` to generate predictions:

```
```bash
mkdir experiments\predict_files
python call_predict.py indobenchmark/indobert-lite-base-p2 indo_nli 1
python call_predict.py indobenchmark/indobert-lite-base-p2 indo_nli 11
Expand Down