-
Notifications
You must be signed in to change notification settings - Fork 33
/
run.sh
executable file
·41 lines (28 loc) · 1.21 KB
/
run.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
#!/bin/bash
# Fail on any error.
set -e
# Display commands being run.
set -x
TMP_DIR="../"
virtualenv --python=python3.6 "${TMP_DIR}/gns/venv"
source "${TMP_DIR}/gns/venv/bin/activate"
# Install dependencies.
pip install --upgrade -r requirements.txt
# Run some training and evaluation in one of the dataset samples.
# Download a sample of a dataset.
DATASET_NAME="WaterDropSample"
# bash ./download_dataset.sh ${DATASET_NAME} "${TMP_DIR}/datasets"
# Train for a few steps.
DATA_PATH="${TMP_DIR}/datasets/${DATASET_NAME}/"
MODEL_PATH="${TMP_DIR}/models/${DATASET_NAME}/"
python -m gns.train --data_path=${DATA_PATH} --model_path=${MODEL_PATH} --ntraining_steps=10 --mode='train'
# Evaluate on validation split.
python -m gns.train --data_path=${DATA_PATH} --model_path=${MODEL_PATH} --model_file="model.pt" --mode="valid"
# Generate test rollouts.
ROLLOUT_PATH="${TMP_DIR}/rollouts/${DATASET_NAME}/"
mkdir -p ${ROLLOUT_PATH}
python -m gns.train --data_path=${DATA_PATH} --model_path=${MODEL_PATH} --model_file="model.pt" --output_path=${ROLLOUT_PATH} --mode='rollout'
# Plot the first rollout.
python -m gns.render_rollout --rollout_path="${ROLLOUT_PATH}/rollout_0.pkl" --block_on_show=False
# Clean up.
rm -r ${TMP_DIR}