Skip to content

Commit

Permalink
chore: merging main to v0.3 (#117)
Browse files Browse the repository at this point in the history
* fix_readme: mention namespace name when applying the pipeline (#88)
* fix: allow only patch updates in torch version due to cuda build errors on mac (#90)
* chore: bump to v0.2.7 for new release (#91)
* docs: updated CHANGELOG.md (#93)
* fix: change example pipeline mlflow port (#96)
* chore: use separate docker image for mlflow server (#101)
* Use mlflow-skinny instead of mlflow
* Update example dockerfile to have multi-stage; one for udf and one for mlflow
* Update docs to use quay.io images, instead of requiring build
* docs: updated CHANGELOG.md (#107)
* fix: sklearn base import for scikit learn v1.2 (#112)
* docs: updated CHANGELOG.md (#113)
* docs: updated CHANGELOG.md (#114)

Signed-off-by: Avik Basu <avikbasu93@gmail.com>
Co-authored-by: s0nicboOm <i.kushalbatra@gmail.com>
Co-authored-by: GitHub <noreply@github.com>
  • Loading branch information
3 people committed Jan 5, 2023
1 parent 2735d72 commit 612a101
Show file tree
Hide file tree
Showing 8 changed files with 89 additions and 22 deletions.
54 changes: 54 additions & 0 deletions CHANGELOG.md
@@ -1,5 +1,59 @@
# Changelog

## v0.3.0a1 (2022-12-22)

* [88d26ec](https://github.com/numaproj/numalogic/commit/88d26ec000ee719f967dc90f4f78f57784b1db7e) feat!: convert AE variants to lightning modules (#110)
* [ed94615](https://github.com/numaproj/numalogic/commit/ed9461517f68192d275d81ef38727bcbe62e887f) fix: fix and clean mlflow test cases (#109)

### Contributors

* Avik Basu
* Kushal Batra

## v0.3.0a0 (2022-12-08)

* [78cf5b4](https://github.com/numaproj/numalogic/commit/78cf5b4e26ffd6dffa5e0dc364cbd51b220bd928) fix: fix pipeline for 0.3 release (#106)
* [709553f](https://github.com/numaproj/numalogic/commit/709553f4ae879f7e786634b02867966756706174) fix: fix mlflow test cases (#98)
* [701812e](https://github.com/numaproj/numalogic/commit/701812efed6742915d9cf57aef985127b3cc449d) feat!: disentangle threshold selection from the main model (#89)

### Contributors

* Avik Basu
* Kushal Batra

## v0.2.9 (2022-12-21)

* [df20591](https://github.com/numaproj/numalogic/commit/df20591f5a3b45117525c4331e5668c4289bd118) fix: sklearn base import for scikit learn v1.2 (#112)

### Contributors

* Avik Basu

## v0.2.8 (2022-11-29)

* [7d5075d](https://github.com/numaproj/numalogic/commit/7d5075ddda5950fef772cf07ffdb8f949bed589f) remove mlflow full
* [05d6071](https://github.com/numaproj/numalogic/commit/05d6071a1de4ca85ed8095bab7654c2e77f366f5) fix: have mlflow-server as an optional extra
* [3c5bc83](https://github.com/numaproj/numalogic/commit/3c5bc8371a50a98a7ffc634b034692cc53ba597c) fix: lock file
* [d2c6293](https://github.com/numaproj/numalogic/commit/d2c6293f594ed67cc8a6ddabce1f8f0bdd84249f) fix: change example pipeline mlflow port (#96)

### Contributors

* Avik Basu
* Kushal Batra

## v0.2.7 (2022-11-14)

* [f9c74c9](https://github.com/numaproj/numalogic/commit/f9c74c90c37fca24db3465c24326a639e93bc802) fix: allow only patch updates in torch version due to cuda build errors on mac (#90)
* [0a5fcdf](https://github.com/numaproj/numalogic/commit/0a5fcdf6c1ea4732b2c6c97f57d983751cfe2956) fix_readme: mention namespace name when applying the pipeline (#88)
* [537fae5](https://github.com/numaproj/numalogic/commit/537fae56577954391d10b134cb318d388c1a7dd7) fix: AutoencoderPipeline logged loss mean (#55)
* [22f8e5d](https://github.com/numaproj/numalogic/commit/22f8e5d79183d848f68c7cbeb2d88126fb326d03) chore!: make mlflow as an optional dependency (#47)

### Contributors

* Avik Basu
* Kushal Batra
* diego-ponce

## v0.2.6 (2022-10-17)

* [5703d1b](https://github.com/numaproj/numalogic/commit/5703d1b75d642242dc5c521a2e85e9cbb2527923) fix: update readme with optional mlflow dependency
Expand Down
2 changes: 1 addition & 1 deletion Makefile
Expand Up @@ -23,7 +23,7 @@ lint: format

# install all dependencies
setup:
poetry install --all-extras
poetry install --with dev --all-extras

# test your application (tests in the tests/ directory)
test:
Expand Down
15 changes: 7 additions & 8 deletions docs/quick-start.md
Expand Up @@ -83,18 +83,17 @@ Once Numaflow is installed, create a simple Numalogic pipeline, which takes in t

For building this pipeline, navigate to [numalogic-simple-pipeline](https://github.com/numaproj/numalogic/tree/main/examples/numalogic-simple-pipeline) under the examples folder and execute the following commands.

1. Build the docker image, import it to k3d, and apply the pipeline.
1. Apply the pipeline. *Note Make sure the pipeline and, numaflow controllers and isbsvc pods are running in the same namespace (`default` in this case).*
```shell
docker build -t numalogic-simple-pipeline:v1 . && k3d image import docker.io/library/numalogic-simple-pipeline:v1

kubectl apply -f numa-pl.yaml
```
2. To verify if the pipeline has been deployed successfully, check the status of each pod.
```shell
> kubectl get pods
kubectl get pods
```
Output will be something like this:
```
NAME READY STATUS RESTARTS AGE
numaflow-server-d64bf6f7c-2czd7 1/1 Running 0 72s
numaflow-controller-c84948cbb-994fn 1/1 Running 0 72s
isbsvc-default-js-0 3/3 Running 0 68s
isbsvc-default-js-1 3/3 Running 0 68s
isbsvc-default-js-2 3/3 Running 0 68s
Expand All @@ -111,9 +110,9 @@ numalogic-simple-pipeline-in-0-tmd0v 1/1 Running 0

Once the pipeline has been created, the data can be sent to the pipeline by port-forwarding the input vertex.

1. Port-forward to the http-source vertex
1. Port-forward to the http-source vertex. From the above pod output, this would be:
```shell
kubectl port-forward simple-numalogic-pipeline-in-0-xxxxx 8443
kubectl port-forward numalogic-simple-pipeline-in-0-tmd0v 8443
```

2. Send the data to the pod via curl
Expand Down
6 changes: 6 additions & 0 deletions examples/numalogic-simple-pipeline/.dockerignore
@@ -0,0 +1,6 @@
.git/
__pycache__/
**/__pycache__/
*.py[cod]
*$py.class
.idea/
14 changes: 11 additions & 3 deletions examples/numalogic-simple-pipeline/Dockerfile
@@ -1,3 +1,7 @@
####################################################################################################
# builder: install needed dependencies
####################################################################################################

FROM python:3.10-slim-bullseye AS builder

ENV PYTHONFAULTHANDLER=1 \
Expand Down Expand Up @@ -27,9 +31,11 @@ RUN apt-get update \
# install dumb-init
&& wget -O /dumb-init https://github.com/Yelp/dumb-init/releases/download/v1.2.5/dumb-init_1.2.5_x86_64 \
&& chmod +x /dumb-init # \
# && curl -sSL https://install.python-poetry.org | python3 -
&& pip3 install --upgrade pip3

####################################################################################################
# mlflow: used for running the mlflow server
####################################################################################################
FROM builder AS mlflow

WORKDIR $PYSETUP_PATH
Expand All @@ -45,7 +51,9 @@ CMD ["/app/entry.sh"]

EXPOSE 5000


####################################################################################################
# udf: used for running the udf vertices
####################################################################################################
FROM builder AS udf

WORKDIR $PYSETUP_PATH
Expand All @@ -60,4 +68,4 @@ RUN chmod +x entry.sh
ENTRYPOINT ["/dumb-init", "--"]
CMD ["/app/entry.sh"]

EXPOSE 5000
EXPOSE 5000
12 changes: 6 additions & 6 deletions examples/numalogic-simple-pipeline/numa-pl.yaml
Expand Up @@ -18,7 +18,7 @@ spec:
min: 1
udf:
container:
image: docker.io/library/numalogic-simple-pipeline:v0.3.0a1
image: quay.io/numaio/numalogic/example-udf
env:
- name: WIN_SIZE
value: "12"
Expand All @@ -31,7 +31,7 @@ spec:
min: 1
udf:
container:
image: docker.io/library/numalogic-simple-pipeline:v0.3.0a1
image: quay.io/numaio/numalogic/example-udf
env:
- name: WIN_SIZE
value: "12"
Expand All @@ -44,7 +44,7 @@ spec:
min: 1
udf:
container:
image: docker.io/library/numalogic-simple-pipeline:v0.3.0a1
image: quay.io/numaio/numalogic/example-udf
env:
- name: WIN_SIZE
value: "12"
Expand All @@ -57,7 +57,7 @@ spec:
min: 1
udf:
container:
image: docker.io/library/numalogic-simple-pipeline:v0.3.0a1
image: quay.io/numaio/numalogic/example-udf
env:
- name: WIN_SIZE
value: "12"
Expand All @@ -78,7 +78,7 @@ spec:
env:
- name: WIN_SIZE
value: "12"
image: docker.io/library/numalogic-simple-pipeline:v0.3.0a1
image: quay.io/numaio/numalogic/example-udf
args:
- python
- starter.py
Expand Down Expand Up @@ -121,7 +121,7 @@ spec:
app: mlflow
spec:
containers:
- image: docker.io/library/numalogic-simple-pipeline:mlflow-v0.3.0a1
- image: quay.io/numaio/numalogic/example-mlflow
name: mlflow
args:
- server
Expand Down
4 changes: 2 additions & 2 deletions poetry.lock

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

4 changes: 2 additions & 2 deletions pyproject.toml
@@ -1,6 +1,6 @@
[tool.poetry]
name = "numalogic"
version = "0.3.0a1"
version = "0.3.0"
description = "Collection of operational Machine Learning models and tools."
authors = ["Numalogic Developers"]
packages = [{ include = "numalogic" }]
Expand Down Expand Up @@ -28,7 +28,7 @@ scikit-learn = "^1.0"
mlflow-skinny = { version = "~2.0.1", optional = true }

[tool.poetry.extras]
mlflow-skinny = ["mlflow-skinny"]
mlflow = ["mlflow-skinny"]

[tool.poetry.group.torch]
optional = true
Expand Down

0 comments on commit 612a101

Please sign in to comment.