Tacotron2+Waveglow/PyT * AMP support * Data preprocessing for Tacotron 2 training * Fixed dropouts on LSTMCells SSD/PyT * script and notebook for inference * AMP support * README update * updates to examples/* BERT/PyT * initial release GNMT/PyT * Default container updated to NGC PyTorch 19.05-py3 * Mixed precision training implemented using APEX AMP * Added inference throughput and latency results on NVIDIA Tesla V100 16G * Added option to run inference on user-provided raw input text from command line NCF/PyT * Updated performance tables. * Default container changed to PyTorch 19.06-py3. * Caching validation negatives between runs Transformer/PyT * new README * jit support added UNet Medical/TF * inference example scripts added * inference benchmark measuring latency added * TRT/TF-TRT support added * README updated GNMT/TF * Performance improvements Small updates (mostly README) for other models.
64 lines
1.7 KiB
Bash
Executable file
64 lines
1.7 KiB
Bash
Executable file
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
|
|
|
|
#!/bin/bash
|
|
set -e
|
|
|
|
DATASET_NAME=${1:-'ml-20m'}
|
|
RAW_DATADIR=${2:-'/data'}
|
|
CACHED_DATADIR=${3:-"${RAW_DATADIR}/cache/${DATASET_NAME}"}
|
|
|
|
# you can add another option to this case in order to support other datasets
|
|
case ${DATASET_NAME} in
|
|
'ml-20m')
|
|
ZIP_PATH=${RAW_DATADIR}/'ml-20m.zip'
|
|
RATINGS_PATH=${RAW_DATADIR}'/ml-20m/ratings.csv'
|
|
;;
|
|
'ml-1m')
|
|
ZIP_PATH=${RAW_DATADIR}/'ml-1m.zip'
|
|
RATINGS_PATH=${RAW_DATADIR}'/ml-1m/ratings.dat'
|
|
;;
|
|
*)
|
|
echo "Unsupported dataset name: $DATASET_NAME"
|
|
exit 1
|
|
esac
|
|
|
|
if [ ! -d ${RAW_DATADIR} ]; then
|
|
mkdir -p ${RAW_DATADIR}
|
|
fi
|
|
|
|
if [ ! -d ${CACHED_DATADIR} ]; then
|
|
mkdir -p ${CACHED_DATADIR}
|
|
fi
|
|
|
|
rm -f log
|
|
|
|
if [ ! -f ${ZIP_PATH} ]; then
|
|
echo 'Dataset not found, downloading...'
|
|
./download_dataset.sh ${DATASET_NAME} ${RAW_DATADIR}
|
|
fi
|
|
|
|
if [ ! -f ${RATINGS_PATH} ]; then
|
|
unzip -u ${ZIP_PATH} -d ${RAW_DATADIR}
|
|
fi
|
|
|
|
if [ ! -f ${CACHED_DATADIR}/train_ratings.pickle ]; then
|
|
echo "preprocessing ${RATINGS_PATH} and save to disk"
|
|
t0=$(date +%s)
|
|
python convert.py --path ${RATINGS_PATH} --output ${CACHED_DATADIR}
|
|
t1=$(date +%s)
|
|
delta=$(( $t1 - $t0 ))
|
|
echo "Finish preprocessing in $delta seconds"
|
|
else
|
|
echo 'Using cached preprocessed data'
|
|
fi
|
|
|
|
echo "Dataset $DATASET_NAME successfully prepared at: $CACHED_DATADIR"
|