-
Notifications
You must be signed in to change notification settings - Fork 5
/
dvc.yaml
53 lines (40 loc) · 1.6 KB
/
dvc.yaml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
## [PIPELINE STAGES](https://dvc.org/doc/user-guide/pipelines/defining-pipelines#defining-pipelines)
# What are stages ?
## Each one of the data workflows that
## we reproduce reliably to ensure consistent results.
## Might include Feature engineering , train , evaluation
# How are stages defined ?
## The stages entries are
## cmd: python path that wraps executable shell command.
## params: loads parameters. Related to params.yaml file
## deps: dependencies. Might contain Stages that might be executed first
## outs: outputs, they can be data or even the model
# What do you need to set up your experiments ?
## For setting up the experiments, we are going to follow the data science
## cookie template and we are going to use the folders already given.
stages:
load:
cmd: python ExperimentsDVC/src/data/load_dataset.py
featureselection:
cmd: python ExperimentsDVC/src/features/feature_selection.py
params:
- featureselection.features
- featureselection.labels
deps:
- ExperimentsDVC/src/data/load_dataset.py
train:
cmd: python ExperimentsDVC/src/models/train_model.py
params:
- train.n_estimators
- train.max_samples
- train.max_depth
- train.min_samples_split
- train.min_impurity_decrease
deps:
- ExperimentsDVC/src/features/feature_engineering.py
metrics:
- dvclive/metrics.json:
cache: false
plots:
- dvclive/plots:
cache: false