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

[WIP, DO NOT MERGE] edgelist adding #244

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

[WIP, DO NOT MERGE] edgelist adding #244

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
50b41ca
jank edgelist adding
rguo123 Dec 21, 2023
43b57e8
undirected edges
rguo123 Dec 22, 2023
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
34 changes: 24 additions & 10 deletions nomic/project.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1299,6 +1299,8 @@ def _add_data(
self,
data: pa.Table,
pbar=None,
is_edgelist: bool = False,
is_directed: bool = True
):
'''
Low level interface to upload an Arrow Table. Users should generally call 'add_text' or 'add_embeddings.'
Expand All @@ -1323,18 +1325,30 @@ def _add_data(
nrow = len(data)

shard_size = 5_000
n_chunks = int(np.ceil(nrow / shard_size))
# Chunk into 16MB pieces. These will probably compress down a bit.
if bytesize / n_chunks > 16_000_000:
shard_size = int(np.ceil(nrow / (bytesize / 16_000_000)))

data = self._validate_and_correct_arrow_upload(
data=data,
project=self,
)

upload_endpoint = "/v1/project/data/add/arrow"

if is_edgelist:
shard_size = 100_000
upload_endpoint = "/v1/project/data/add/edges"
if not is_directed:
source = data.column("source")
dest = data.column("dest")
flipped = pa.Table.from_arrays([dest, source], names=["source", "dest"])
data = pa.concat_tables([data, flipped])
sort_indices = pc.sort_indices(data, sort_keys=[('source', 'ascending')])
data = data.take(sort_indices)

else:
n_chunks = int(np.ceil(nrow / shard_size))
# Chunk into 16MB pieces. These will probably compress down a bit.
if bytesize / n_chunks > 16_000_000:
shard_size = int(np.ceil(nrow / (bytesize / 16_000_000)))

data = self._validate_and_correct_arrow_upload(
data=data,
project=self,
)

# Actually do the upload
def send_request(i):
data_shard = data.slice(i, shard_size)
Expand Down