Skip to content

Conversation

@MarcelWilnicki
Copy link
Collaborator

No description provided.

help="path to directory with ImageNet validation images")
parser.add_argument("--fixed_input", action='store_true',
help="truncate input to fixed shape")
parser.add_argument("--input_size", type=int, default=384,
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Input_size will only take effect when fixed_input is True. Therefore I suggest making a single arg out of this that both enabled fixed_input and specifies the size. Say --fixed_input_size.

wget https://ampereaimodelzoo.s3.eu-central-1.amazonaws.com/aio_objdet_dataset.tar.gz > /dev/null 2>&1
tar -xf aio_objdet_dataset.tar.gz > /dev/null
apt-get update && apt-get install -y ffmpeg
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should be handled by setup_deb.sh . One of the goals of this test to make sure all deps are installed by it, so if something is missing it should be added there.

wget $S3_URL_RESNET_50_V15_TF_FP32 > /dev/null 2>&1
IGNORE_DATASET_LIMITS=1 python3 computer_vision/classification/resnet_50_v15/run.py -m resnet_50_v15_tf_fp32.pb -p fp32 -f tf --timeout=60
OMP_NUM_THREADS=32 AIO_NUM_THREADS=32 IGNORE_DATASET_LIMITS=1 python3 computer_vision/classification/resnet_50_v15/run.py -m resnet_50_v15_tf_fp32.pb -p fp32 -f tf --timeout=60
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Use envs starting in line 115 for this. No need to specify for each process separately.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants