Skip to content

Commit

Permalink
stage -1-dvc updated
Browse files Browse the repository at this point in the history
  • Loading branch information
shivpalSW committed Jul 13, 2023
1 parent 821af53 commit bc36dae
Show file tree
Hide file tree
Showing 3 changed files with 48 additions and 7 deletions.
2 changes: 2 additions & 0 deletions artifacts/prepared/.gitignore
Original file line number Diff line number Diff line change
@@ -0,0 +1,2 @@
/train.tsv
/test.tsv
31 changes: 31 additions & 0 deletions dvc.lock
Original file line number Diff line number Diff line change
@@ -0,0 +1,31 @@
schema: '2.0'
stages:
prepare_data:
cmd: python src/stage_01_prepare.py --config=configs/config.yaml --params=params.yaml
deps:
- path: configs/config.yaml
md5: ef78262a4a0652b7f26da188dec74419
size: 230
- path: data/data.xml
md5: a304afb96060aad90176268345e10355
size: 37891850
- path: src/stage_01_prepare.py
md5: 47a1121c82f1e79ff3004abb9c4d5d4d
size: 2096
- path: src/utils/common.py
md5: 48b676b0d599169fc5e054b0829a5a4f
size: 519
- path: src/utils/data_mgmt.py
md5: c985e579010a81c5462ebf8184649fe4
size: 871
params:
params.yaml:
prepare.seed: 2021
prepare.split: 0.2
outs:
- path: artifacts/prepared/test.tsv
md5: 60c757f3b30604473de7fa775dfc00d9
size: 4899778
- path: artifacts/prepared/train.tsv
md5: 553034c4cf40efc63c99c19fe98610d0
size: 18986541
22 changes: 15 additions & 7 deletions dvc.yaml
Original file line number Diff line number Diff line change
@@ -1,8 +1,16 @@
# add stages here
# stages:
# stage_name:
# cmd: python src/stage_01______.py
# deps:
# - src/stage_01______.py
# outs:
# - output_data_of_file_path
stages:
prepare_data:
cmd: python src/stage_01_prepare.py --config=configs/config.yaml --params=params.yaml
deps:
- src/stage_01_prepare.py
- data/data.xml
- src/utils/common.py
- src/utils/data_mgmt.py
- configs/config.yaml
params:
- prepare.seed
- prepare.split
outs:
- artifacts/prepared/train.tsv
- artifacts/prepared/test.tsv

0 comments on commit bc36dae

Please sign in to comment.