junte/ML
1
0
Fork 0
mirror of https://github.com/Hestia-Homes/ML.git synced 2026-07-12 13:28:58 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

new makefile

Browse source
This commit is contained in:
Michael Duong 2023-09-22 13:38:22 +00:00
parent 1e70a3a582
commit 59cfa971c7
4 changed files with 28 additions and 31 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
modules/ml-pipeline/src/pipeline/configs/build_model.yaml
View file

@ -11,6 +11,6 @@ AutogluonAutoML:
output_filepath: ./data/model/autogluonmodel/ output_filepath: ./data/model/autogluonmodel/
problem_type: regression problem_type: regression
eval_metric: mean_absolute_error eval_metric: mean_absolute_error
time_limit: 400 time_limit: 60
presets: good_quality presets: medium_quality
excluded_model_types: ['KNN'] excluded_model_types: ['KNN']

1
modules/ml-pipeline/src/pipeline/configs/feature_processor_logic.py
View file

@ -17,3 +17,4 @@ def SAP_ENDING(df):
new_feature_funcs = {"SAP_ENDING": SAP_ENDING} new_feature_funcs = {"SAP_ENDING": SAP_ENDING}
# new_feature_funcs = {}

52
modules/ml-pipeline/src/pipeline/dvc.lock
View file

@ -5,8 +5,8 @@ stages:
deps: deps:
- path: prepare_data.py - path: prepare_data.py
hash: md5 hash: md5
md5: 934d774e67f38e440b621ce71152f5f6 md5: 2648d7d407dca857a1d20a11a88d3d98
size: 5031 size: 5116
params: params:
configs/prepare_data.yaml: configs/prepare_data.yaml:
output_test_filepath: ./data/prepared_data/test.parquet output_test_filepath: ./data/prepared_data/test.parquet
@ -15,8 +15,8 @@ stages:
outs: outs:
- path: data/prepared_data/ - path: data/prepared_data/
hash: md5 hash: md5
md5: 3767eec56906f5ac724a3f07433645ef.dir md5: ec064b0274e2e6a0864580a748e6bb6a.dir
size: 13442342 size: 24067192
nfiles: 2 nfiles: 2
build_model: build_model:
cmd: python build_model.py cmd: python build_model.py
@ -27,8 +27,8 @@ stages:
size: 5134 size: 5134
- path: data/prepared_data - path: data/prepared_data
hash: md5 hash: md5
md5: 3767eec56906f5ac724a3f07433645ef.dir md5: ec064b0274e2e6a0864580a748e6bb6a.dir
size: 13442342 size: 24067192
nfiles: 2 nfiles: 2
params: params:
configs/build_model.yaml: configs/build_model.yaml:
@ -36,8 +36,8 @@ stages:
output_filepath: ./data/model/autogluonmodel/ output_filepath: ./data/model/autogluonmodel/
problem_type: regression problem_type: regression
eval_metric: mean_absolute_error eval_metric: mean_absolute_error
time_limit: 400 time_limit: 60
presets: good_quality presets: medium_quality
excluded_model_types: excluded_model_types:
- KNN - KNN
SKLearnLinearRegression: SKLearnLinearRegression:
@ -49,25 +49,25 @@ stages:
outs: outs:
- path: data/model/ - path: data/model/
hash: md5 hash: md5
md5: 7b2f8334c81fb5ff23e42e77741b31d1.dir md5: dc73587056f07735719bfee464a5f898.dir
size: 118227750 size: 285397707
nfiles: 71 nfiles: 16
- path: metrics/fit_metrics.json - path: metrics/fit_metrics.json
hash: md5 hash: md5
md5: e1c9a16617804f48e8ffac7cec6575ca md5: f6d03cb197a3d78e61f6fef023ed8d7f
size: 185 size: 184
generate_predictions: generate_predictions:
cmd: python generate_predictions.py cmd: python generate_predictions.py
deps: deps:
- path: data/model - path: data/model
hash: md5 hash: md5
md5: 7b2f8334c81fb5ff23e42e77741b31d1.dir md5: dc73587056f07735719bfee464a5f898.dir
size: 118227750 size: 285397707
nfiles: 71 nfiles: 16
- path: data/prepared_data - path: data/prepared_data
hash: md5 hash: md5
md5: 3767eec56906f5ac724a3f07433645ef.dir md5: ec064b0274e2e6a0864580a748e6bb6a.dir
size: 13442342 size: 24067192
nfiles: 2 nfiles: 2
- path: generate_predictions.py - path: generate_predictions.py
hash: md5 hash: md5
@ -83,21 +83,21 @@ stages:
outs: outs:
- path: data/predictions/ - path: data/predictions/
hash: md5 hash: md5
md5: fb7cf3f4a90598ec1e43a1b7a4af3bef.dir md5: 3d70862c00e38f383a06c3e4df5ebb38.dir
size: 536774 size: 386958
nfiles: 1 nfiles: 1
generate_metrics: generate_metrics:
cmd: python generate_metrics.py cmd: python generate_metrics.py
deps: deps:
- path: data/predictions - path: data/predictions
hash: md5 hash: md5
md5: fb7cf3f4a90598ec1e43a1b7a4af3bef.dir md5: 3d70862c00e38f383a06c3e4df5ebb38.dir
size: 536774 size: 386958
nfiles: 1 nfiles: 1
- path: data/prepared_data - path: data/prepared_data
hash: md5 hash: md5
md5: 3767eec56906f5ac724a3f07433645ef.dir md5: ec064b0274e2e6a0864580a748e6bb6a.dir
size: 13442342 size: 24067192
nfiles: 2 nfiles: 2
- path: generate_metrics.py - path: generate_metrics.py
hash: md5 hash: md5
@ -106,14 +106,12 @@ stages:
params: params:
configs/generate_metrics.yaml: configs/generate_metrics.yaml:
dataclient_type: local dataclient_type: local
input_datahandler_type: parquet
metrics_output_filepath: ./metrics/metrics.json metrics_output_filepath: ./metrics/metrics.json
metrics_type: Regression metrics_type: Regression
output_datahandler_type: json
outs: outs:
- path: metrics/metrics.json - path: metrics/metrics.json
hash: md5 hash: md5
md5: 852ef4cf2ca5e7f89d70420a9df7a596 md5: 2507f756ea68768185ebeaf66db2ebbd
size: 183 size: 183
startup_cleanup: startup_cleanup:
cmd: python startup_cleanup.py cmd: python startup_cleanup.py

2
modules/ml-pipeline/src/pipeline/requirements/training/requirements-dev.txt
View file

@ -5,5 +5,3 @@ autogluon==0.8.2
alibi==0.9.4 alibi==0.9.4
pyarrow==13.0.0 pyarrow==13.0.0
pre-commit==3.3.3 pre-commit==3.3.3
sphinx==7.2.5
sphinx_rtd_theme==1.3.0