8000 docs: crud vector kv indexer by florian-hoenicke · Pull Request #1814 · jina-ai/serve · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

docs: crud vector kv indexer #1814

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.

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

Merged
merged 8 commits into from
Feb 3, 2021
Merged

docs: crud vector kv indexer #1814

merged 8 commits into from
Feb 3, 2021

Conversation

florian-hoenicke
Copy link
Member

No description provided.

@codecov
Copy link
codecov bot commented Jan 29, 2021

Codecov Report

Merging #1814 (dbed042) into master (39116ee) will not change coverage.
The diff coverage is 66.66%.

Impacted file tree graph

@@           Coverage Diff           @@
##           master    #1814   +/-   ##
=======================================
  Coverage   85.85%   85.85%           
=======================================
  Files         145      145           
  Lines        6944     6944           
=======================================
  Hits         5962     5962           
  Misses        982      982           
Impacted Files Coverage Δ
jina/drivers/index.py 100.00% <ø> (ø)
jina/drivers/search.py 96.49% <ø> (ø)
jina/executors/indexers/cache.py 97.18% <ø> (ø)
jina/executors/indexers/keyvalue.py 100.00% <ø> (ø)
jina/executors/indexers/vector.py 93.78% <ø> (ø)
jina/drivers/cache.py 93.75% <66.66%> (ø)

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 39116ee...2c4cfc1. Read the comment docs.

@github-actions
Copy link
github-actions bot commented Jan 29, 2021

Latency summary

Current PR yields:

  • 🐢🐢 index QPS at 1256, delta to last 3 avg.: -6%
  • 🐢🐢 query QPS at 25, delta to last 3 avg.: -7%

Breakdown

Version Index QPS Query QPS
current 1256 25
0.9.27 1348 26
0.9.26 1349 27

Backed by latency-tracking. Further commits will update this comment.

@jina-bot jina-bot added size/M and removed size/S labels Jan 29, 2021
- ``level=all``: D x C x K
- ``granularity=0``: D x K
- ``granularity=1``: D x C x K
- ``granularity=1``: D x C x C x K
Copy link
Contributor

Choose a reason for hiding this comment

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

this is granularity 0 and 1.

- ``level=doc``: D x K
- ``level=all``: D x C x K
- ``granularity=0``: D x K
- ``granularity=1``: D x C x K
Copy link
Contributor

Choose a reason for hiding this comment

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

I would reflect it with traversal_paths arguments and not granularity.

If

traversal_paths = ['m'] => D x K
traversal_paths = ['r'] => D
traversal_paths = ['cm'] => D x C x K
traversal_paths = ['m', 'cm'] => D x K + D x C x K

Is this what we want to express?

Copy link
Member Author

Choose a reason for hiding this comment

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

really awesome illustration. Thanks I will take this. This is much more understandable for the user

@@ -12,7 +12,7 @@


class BaseIndexer(BaseExecutor):
""" base class for storing and searching any kind of data structure
"""base class for storing and searching any kind of data structure
Copy link
Contributor

Choose a reason for hiding this comment

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

Capital letter at the beginning?

Copy link
Member Author

Choose a reason for hiding this comment

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

check

@@ -42,7 +41,7 @@ def _apply_all(self, docs: 'DocumentSet', *args, **kwargs) -> None:
self.on_hit(d, result)

def on_miss(self, doc: 'Document', data) -> None:
"""Function to call when doc is missing, the default behavior is add to cache when miss
"""Function to call when document is missing, the default behavior is add to cache when miss.
Copy link
Contributor

Choose a reason for hiding this comment

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

'the default behavior is add to cache when miss. '
does it mean 'the default behavior is to add the document to cache when miss. '?

@@ -31,24 +31,26 @@ def __init__(


class KVSearchDriver(BaseSearchDriver):
"""Fill in the doc/chunk-level top-k results using the :class:`jina.executors.indexers.meta.BinaryPbIndexer`
"""Fill in the results using the :class:`jina.executors.indexers.meta.BinaryPbIndexer`
Copy link
Contributor

Choose a reason for hiding this comment

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

Probably we could all use first-person expressions(like 'fill') or third person(like 'fills')? Now some docstrings use first-person but some use third-person

Copy link
Member

Choose a reason for hiding this comment

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

Let's stick to the first-person expression

@nan-wang nan-wang merged commit ba4cea3 into master Feb 3, 2021
@nan-wang nan-wang deleted the docs-crud-vector-kv-indexer branch February 3, 2021 05:32
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
4882
Development

Successfully merging this pull request may close these issues.

6 participants
0