diff --git a/.dvc/.gitignore b/.dvc/.gitignore
new file mode 100644
index 00000000..528f30c7
--- /dev/null
+++ b/.dvc/.gitignore
@@ -0,0 +1,3 @@
+/config.local
+/tmp
+/cache
diff --git a/.dvc/config b/.dvc/config
new file mode 100644
index 00000000..7a5426e9
--- /dev/null
+++ b/.dvc/config
@@ -0,0 +1,5 @@
+[core]
+    analytics = false
+    remote = remote_storage
+['remote "remote_storage"']
+    url = /home/mlops/dvc_remote
diff --git a/.dvcignore b/.dvcignore
new file mode 100644
index 00000000..51973055
--- /dev/null
+++ b/.dvcignore
@@ -0,0 +1,3 @@
+# Add patterns of files dvc should ignore, which could improve
+# the performance. Learn more at
+# https://dvc.org/doc/user-guide/dvcignore
diff --git a/data/prepared/.gitignore b/data/prepared/.gitignore
index e69de29b..22a65dd9 100644
--- a/data/prepared/.gitignore
+++ b/data/prepared/.gitignore
@@ -0,0 +1,2 @@
+/train.csv
+/test.csv
diff --git a/data/prepared/test.csv.dvc b/data/prepared/test.csv.dvc
new file mode 100644
index 00000000..3b350283
--- /dev/null
+++ b/data/prepared/test.csv.dvc
@@ -0,0 +1,4 @@
+outs:
+- md5: cbd4ba69ced15e40820a635e7e741627
+  size: 71491
+  path: test.csv
diff --git a/data/prepared/train.csv.dvc b/data/prepared/train.csv.dvc
new file mode 100644
index 00000000..ae08b157
--- /dev/null
+++ b/data/prepared/train.csv.dvc
@@ -0,0 +1,4 @@
+outs:
+- md5: 7f79cf9a4ab1316f7d7246d0b92ea16c
+  size: 178060
+  path: train.csv
diff --git a/data/raw/.gitignore b/data/raw/.gitignore
index e69de29b..a5d9d98f 100644
--- a/data/raw/.gitignore
+++ b/data/raw/.gitignore
@@ -0,0 +1,2 @@
+/train
+/val
diff --git a/data/raw/train.dvc b/data/raw/train.dvc
new file mode 100644
index 00000000..14162364
--- /dev/null
+++ b/data/raw/train.dvc
@@ -0,0 +1,5 @@
+outs:
+- md5: 7adc7abb69056f4d7afb512c78f2fce9.dir
+  size: 75309082
+  nfiles: 9470
+  path: train
diff --git a/data/raw/val.dvc b/data/raw/val.dvc
new file mode 100644
index 00000000..237148f4
--- /dev/null
+++ b/data/raw/val.dvc
@@ -0,0 +1,5 @@
+outs:
+- md5: 0ad4dcf197b452735726bf8d8777201d.dir
+  size: 31248080
+  nfiles: 3925
+  path: val
diff --git a/metrics/accuracy.json b/metrics/accuracy.json
new file mode 100644
index 00000000..39176609
--- /dev/null
+++ b/metrics/accuracy.json
@@ -0,0 +1 @@
+{"accuracy": 0.7490494296577946}
\ No newline at end of file
diff --git a/model/.gitignore b/model/.gitignore
index e69de29b..565a9d50 100644
--- a/model/.gitignore
+++ b/model/.gitignore
@@ -0,0 +1 @@
+/model.joblib
diff --git a/model/model.joblib.dvc b/model/model.joblib.dvc
new file mode 100644
index 00000000..4b96d593
--- /dev/null
+++ b/model/model.joblib.dvc
@@ -0,0 +1,4 @@
+outs:
+- md5: 22490d1b369e3f7423c5b6ebd4db4234
+  size: 241075
+  path: model.joblib
diff --git a/src/train.py b/src/train.py
index e5feeda6..5d3c7c2c 100644
--- a/src/train.py
+++ b/src/train.py
@@ -37,7 +37,7 @@ def load_data(data_path):
 def main(repo_path):
     train_csv_path = repo_path / "data/prepared/train.csv"
     train_data, labels = load_data(train_csv_path)
-    sgd = SGDClassifier(max_iter=10)
+    sgd = SGDClassifier(max_iter=100)
     trained_model = sgd.fit(train_data, labels)
     dump(trained_model, repo_path / "model/model.joblib")