| def binop(fun: int, x: int, y: int) -> int: | |
| val = (x << 1) + y | |
| return (fun >> val) & 1 | |
| possible = set() | |
| # h(x, g(y, z)) | |
| for h in range(16): | |
| for g in range(16): | |
| outputs = 0 | |
| for n in range(8): |
| # define hashes | |
| GIT_HASH=0a7a9d10 | |
| QUILT_HASH=3722a498 | |
| DOCKER_HASH=sha256:8a4f4123c92a7fe2e8ca4c404094ab95dc1fb868ad077d2e084ba4082a5a29c1 | |
| # pull image | |
| DOCKER_IMAGE=quiltdata/pytorch-detectron2-demo@${DOCKER_HASH} | |
| docker pull ${DOCKER_IMAGE} | |
| # run image (interactively for illustration) | |
| nvidia-docker run -it \ | |
| --ulimit memlock=-1 \ |
| # define hashes | |
| GIT_HASH=0a7a9d10 | |
| QUILT_HASH=3722a498 | |
| DOCKER_HASH=sha256:8a4f4123c92a7fe2e8ca4c404094ab95dc1fb868ad077d2e084ba4082a5a29c1 | |
| # pull image | |
| DOCKER_IMAGE=quiltdata/pytorch-detectron2-demo@${DOCKER_HASH} | |
| docker pull ${DOCKER_IMAGE} | |
| # run image (interactively for illustration) | |
| nvidia-docker run -it \ | |
| --ulimit memlock=-1 \ |
| pip install quilt3 | |
| # Note: installing COCO requires at least 21GB of disk | |
| quilt3 install cv/coco2017 \ | |
| --top-hash 3722a4 \ | |
| --dest ./detectron2/datasets/coco/ \ | |
| --registry s3://quilt-ml |
This document outlines several key use cases and proposes code samples to satisfy those usecases. Both usecases and code samples are open to discussion.
The document is about UX, not implementation. It does not go much into performance needs, but the expectation is that the implementation will be written such that any API we expose is as high-performance as possible.
API names are rough and will need to be improved.
| def post_to_es(event_type, size, text, key, meta, version_id=''): | |
| ES_HOST = os.environ['ES_HOST'] | |
| ES_INDEX = 'drive' | |
| data = { | |
| 'type': event_type, | |
| 'size': size, | |
| 'text': text, | |
| 'key': key, |
| NB_VERSION = 4 | |
| def extract_text(notebook_str): | |
| formatted = nbformat.reads(notebook_str, as_version=NB_VERSION) | |
| text = [] | |
| for cell in formatted.get('cells', []): | |
| if 'source' in cell and 'cell_type' in cell: | |
| if cell['cell_type'] == 'code' or cell['cell_type'] == 'markdown': | |
| text.append(cell['source']) |
| def handler(event, context): | |
| for record in event['Records']: | |
| try: | |
| response = S3_CLIENT.get_object(Bucket=bucket, Key=key) | |
| eventname = record['eventName'] | |
| bucket = unquote(record['s3']['bucket']['name']) | |
| if eventname == 'ObjectRemoved:Delete': | |
| event_type = 'Delete' | |
| post_to_es(event_type, 0, '', key, {}) | |
| continue |
| from quilt.data.robmarkcole import birds_vs_not_birds as birds | |
| from quilt.asa.img import plot | |
| %matplotlib inline | |
| birds.birds(asa=plot()) |
| Adapted from [scikit learn cheat sheet](http://scikit-learn.org/stable/tutorial/machine_learning_map/index.html). | |
| * More than 50 samples? | |
| * No | |
| * Get more data | |
| * Yes | |
| * Predicting a category? | |
| * No | |
| * Predicting a quantity? | |
| * No |