Reproduce model using bigrams

Author: shcheklein

dvc.lock

@@ -35,21 +35,21 @@ stages:
       size: 4158
     params:
       params.yaml:
-        featurize.max_features: 100
-        featurize.ngrams: 1
+        featurize.max_features: 200
+        featurize.ngrams: 2
     outs:
     - path: data/features
       hash: md5
-      md5: f8f5cbc3188008a7542d02d63054d9d2.dir
-      size: 1556290
+      md5: f35d4cc2c552ac959ae602162b8543f3.dir
+      size: 2232588
       nfiles: 2
   train:
     cmd: python src/train.py data/features model.pkl
     deps:
     - path: data/features
       hash: md5
-      md5: f8f5cbc3188008a7542d02d63054d9d2.dir
-      size: 1556290
+      md5: f35d4cc2c552ac959ae602162b8543f3.dir
+      size: 2232588
       nfiles: 2
     - path: src/train.py
       hash: md5
@@ -63,8 +63,8 @@ stages:
     outs:
     - path: model.pkl
       hash: md5
-      md5: cfa72ff6e2575c44f78f423cada5b783
-      size: 1855075
+      md5: d1f6e055f7f5e2827fcfae68d9b64d4c
+      size: 1958115
   evaluate:
     cmd: python src/evaluate.py model.pkl data/features
     deps:

params.yaml

@@ -3,8 +3,8 @@ prepare:
   seed: 20170428
 
 featurize:
-  max_features: 100
-  ngrams: 1
+  max_features: 200
+  ngrams: 2
 
 train:
   seed: 20170428