David Olaru
b5ea1cf228
Align dependencies between requirement files and setup.py ( #460 )
2022-04-27 07:14:49 -05:00
Benjamin Trent
8294224e34
[ML] Fix XGBoost model import for xgboost>=1.6
2022-04-20 09:20:50 -05:00
Seth Michael Larson
cb839a9ac9
Release 8.1.0
2022-03-31 17:12:26 -05:00
P. Sai Vinay
76a52b7947
Add support for eland.Series.unqiue()
2022-03-31 08:33:15 -05:00
Benjamin Trent
15a3007288
[ML] add roberta bart transformer upload support ( #443 )
...
Related to: https://github.com/elastic/elasticsearch/pull/84777
This allows BART and RoBERTa models to be uploaded to Elasticsearch for our currently defined NLP tasks.
2022-03-14 12:26:12 -04:00
David Kyle
5678525b15
Fix mypy type errors for elasticsearch-python v8.0.0
2022-03-08 17:50:39 -06:00
David Kyle
5c5e5af54d
Add --ca-certs and --insecure option for configuring TLS
2022-03-08 15:44:13 -06:00
Seth Michael Larson
abd05df50b
Release 8.0.0
2022-02-10 14:29:54 -06:00
Ashton Sidhu
e3bff8a623
Add option to disable schema enforcement for pandas_to_eland
2022-01-14 07:35:58 -06:00
István Zoltán Szabó
9206941659
[DOCS] Adds NLP with PyTorch section to ML-related page in Eland docs
2022-01-11 09:08:00 -06:00
Benjamin Trent
72856e2c3f
[ML] Add support for MPNet PyTorch models
2022-01-10 11:21:30 -06:00
Ashton Sidhu
64daa07a65
Using the 'date' field for datetime64+timezone columns
2022-01-04 22:03:49 -06:00
Florian Winkler
3db93cd789
Allow using datetime types in filters
2022-01-04 14:46:18 -06:00
Seth Michael Larson
c14bc24032
Release 8.0.0-beta1
2021-12-16 07:42:38 -06:00
Seth Michael Larson
ffe7c792dc
Update Notebook examples for 8.0
2021-12-15 16:01:32 -06:00
Seth Michael Larson
cd0897f5d7
Add a warning when connecting to incompatible Elasticsearch versions
2021-12-15 14:08:20 -06:00
Seth Michael Larson
109387184a
Support the v8.0 Elasticsearch client
2021-12-09 15:01:26 -06:00
Josh Devins
1ffbe002c4
Upgrade PyTorch dependencies to latest
...
In preparation for an 8.0 release, this updates PyTorch NLP dependencies
to more recent and latest minor versions. Amongst other things, this
introduces a fix from transformers that is helpful for text embedding
tasks with certain DPR models.
See: https://github.com/huggingface/transformers/issues/13670
Co-authored-by: Seth Michael Larson <seth.larson@elastic.co>
2021-12-06 09:05:54 -06:00
Seth Michael Larson
e6bb917d83
Add quotes to versions in test-matrix.yml
2021-12-03 09:37:37 -06:00
Seth Michael Larson
4e489de424
Bump version to 8.0.0
2021-12-02 08:41:11 -06:00
Seth Michael Larson
f98ebd4c29
Update Jenkins jobs for 8.x and 7.x
2021-12-01 14:01:48 -06:00
Josh Devins
5bc1a824a7
Add PyTorch modules to noxfile
...
We added the `pytorch` module which is type checked but was not in the
noxfile as such. This change also addresses type errors that arose after
adding type checking.
2021-11-29 08:03:25 -08:00
Josh Devins
7209f61773
Adds max_length padding to transformer tracing ( #411 )
...
The padding parameter needs to be set on the tokenization call and not
in the constructor. Furthermore, the True value will only pad to the
largest input in a batch, however we don't trace with batches so this
value had no effect. The proper place to pass this parameter is in the
tokenization call itself and the proper value to use is "max_length"
which will pad the input to the maximum input size specified by the
model. Although we measure no functional or performance impact of this
setting, it has been suggested that this is a best practice.
See: https://huggingface.co/transformers/serialization.html#dummy-inputs-and-standard-lengths
2021-11-11 13:18:55 +01:00
Benjamin Trent
a3b0907c5b
[ML] Add inference results tests for PyTorch transformer models
2021-11-10 06:50:10 -06:00
Seth Michael Larson
66e3e4eaad
Set 'script.max_compilations_rate: use-context'
2021-11-02 10:09:25 -04:00
Josh Devins
1e5b475bee
Adds NLP with PyTorch basic example to README
...
The Machine Learning section now has two sub-sections — one for
traditional regression/classification and the other for NLP with
PyTorch. The examples show two ways to upload models from the Hugging
Face model hub.
2021-11-02 08:00:33 -05:00
Josh Devins
df51f8af07
Document how to install transitive binary dependencies, add repo Dockerfile
...
Co-authored-by: Seth Michael Larson <seth.larson@elastic.co>
2021-10-28 12:05:39 -05:00
Seth Michael Larson
19014f1227
Avoid DeprecationWarnings when using the new Elasticsearch client (7.15+)
2021-10-28 09:24:36 -05:00
Benjamin Trent
79b66eb6b4
Updating node type to larger ubuntu node ( #404 )
...
* Updating node type to larger ubuntu node
* adding torch location
* formatting
* formatting
;
* removing torch location specification
2021-10-25 14:48:26 -05:00
Benjamin Trent
d39c1cd784
[ML] Make eland_import_hub_model an installable script
2021-10-19 11:29:58 -05:00
P. Sai Vinay
704c8982bc
Optimize to_pandas() internally to improve performance
...
Co-authored-by: Seth Michael Larson <seth.larson@elastic.co>
2021-10-13 13:23:04 -05:00
James Rodewig
6088f2e39d
[DOCS] Retitle Eland Python Client docs book
2021-10-12 20:22:17 -05:00
P. Sai Vinay
f9d2defb1b
Add number_samples to sklearn MLModel
2021-10-07 08:14:54 -05:00
Josh Devins
014943d3b8
Add initial implementation of PyTorch ML models
2021-10-06 08:44:40 -05:00
P. Sai Vinay
995f2432b6
Add number_samples to LightGBM MLModel and leaf_count to leaf nodes
...
* Add number_samples to lightgbm ML Model
* Add leaf_count for leaf nodes
2021-09-30 08:13:44 -05:00
P. Sai Vinay
dabb327b8b
Refactor df.info() for better readability
2021-09-28 15:12:29 -05:00
P. Sai Vinay
bc201e22dd
Improve coverage for eland.dataframe
2021-09-28 15:11:57 -05:00
Seth Michael Larson
b8e192b7d0
Rename Jenkins job to 'main'
2021-09-28 10:07:16 -05:00
P. Sai Vinay
f241ae971a
Add flynt and --cov-report=term-missing
2021-09-21 11:18:01 -05:00
Seth Michael Larson
7aabc88e4a
Rename 'master' branch to 'main'
2021-09-08 11:51:50 -05:00
Jabin Kong
77f9a455e9
Fix docstring formatting
2021-09-07 11:40:19 -05:00
P. Sai Vinay
315f94b201
Add excluded lines for coverage and improve coverage
2021-09-07 11:39:19 -05:00
Seth Michael Larson
a50c3657c4
Release v7.14.1b1
2021-08-30 13:42:55 -05:00
Seth Michael Larson
7a2e845a76
Speedup CI by only installing Nox in Dockerfile
2021-08-20 08:39:02 -05:00
Jabin Kong
1aa193da9e
Add iterrows() and itertuples() to DataFrame
...
Co-authored-by: Seth Michael Larson <seth.larson@elastic.co>
2021-08-20 08:34:52 -05:00
Seth Michael Larson
e4f88a34a6
Yield list of hits from _search_yield_hits() instead of individual hits
2021-08-17 12:16:10 -05:00
P. Sai Vinay
011bf29816
Simplify ES->pandas logic by removing Collectors
2021-08-16 12:22:02 -05:00
Seth Michael Larson
76d83ea47f
Bump version to 7.14.0b1
2021-08-09 09:21:49 -05:00
Seth Michael Larson
b0c8434c06
Release 7.14.0b1
2021-08-09 09:11:57 -05:00
Seth Michael Larson
15ba8d3e02
Fallback on using scroll searches for Elasticsearch <7.12
...
PIT+search_after became universally safe in Elasticsearch 7.12 by adding an automatic sort tiebreaker field when using PITs called `_shard_doc` but now we need to do feature detection to make sure we use the previous scroll method on Elasticsearch <7.12 clusters
2021-08-08 12:19:41 -05:00
