From dd9c18167495cc918448df2f6956296c81ff3478 Mon Sep 17 00:00:00 2001
From: Dmitry Petrov <dmitry.petrov@nevesomo.com>
Date: Mon, 23 Jan 2023 23:36:54 +0000
Subject: [PATCH] Try a 40K dataset (4x data)

---
 data/data.xml.dvc |  9 +++++----
 dvc.lock          | 38 +++++++++++++++++++-------------------
 params.yaml       |  2 +-
 3 files changed, 25 insertions(+), 24 deletions(-)

diff --git a/data/data.xml.dvc b/data/data.xml.dvc
index 114d7c814..ca146c5a0 100644
--- a/data/data.xml.dvc
+++ b/data/data.xml.dvc
@@ -1,12 +1,13 @@
-md5: f66db4fac66a93d4feaa939b4506c3ab
+md5: 25c4a84510a41557840c61692dd14c11
 frozen: true
 deps:
 - path: get-started/data.xml
   repo:
     url: https://github.com/iterative/dataset-registry
-    rev_lock: 705bc71a0a13c47b9e5147a3524fafc41f8ac7fa
+    rev_lock: cf6481baf56f156aa0876709cc231aaf3f3a3c29
+    rev: get-started-40K
 outs:
-- md5: 22a1a2931c8370d3aeedd7183606fd7f
-  size: 14445097
+- md5: 4bd325a30d5f1d5ea1a451d98767ddde
+  size: 59918667
   hash: md5
   path: data.xml
diff --git a/dvc.lock b/dvc.lock
index 447fcd573..ee3b44ae3 100644
--- a/dvc.lock
+++ b/dvc.lock
@@ -5,8 +5,8 @@ stages:
     deps:
     - path: data/data.xml
       hash: md5
-      md5: 22a1a2931c8370d3aeedd7183606fd7f
-      size: 14445097
+      md5: 4bd325a30d5f1d5ea1a451d98767ddde
+      size: 59918667
     - path: src/prepare.py
       hash: md5
       md5: f54d670ac8a4f63206781fc31d1f2651
@@ -18,16 +18,16 @@ stages:
     outs:
     - path: data/prepared
       hash: md5
-      md5: 153aad06d376b6595932470e459ef42a.dir
-      size: 8437363
+      md5: f8934609be51496ee500f80eea539c6f.dir
+      size: 35339221
       nfiles: 2
   featurize:
     cmd: python src/featurization.py data/prepared data/features
     deps:
     - path: data/prepared
       hash: md5
-      md5: 153aad06d376b6595932470e459ef42a.dir
-      size: 8437363
+      md5: f8934609be51496ee500f80eea539c6f.dir
+      size: 35339221
       nfiles: 2
     - path: src/featurization.py
       hash: md5
@@ -35,21 +35,21 @@ stages:
       size: 4158
     params:
       params.yaml:
-        featurize.max_features: 200
+        featurize.max_features: 500
         featurize.ngrams: 2
     outs:
     - path: data/features
       hash: md5
-      md5: 4281fdd8e973e3bbe5abc0ae10adebc7.dir
-      size: 2232588
+      md5: c9308f114f6a8f06fb5ba2b40ea81678.dir
+      size: 12597137
       nfiles: 2
   train:
     cmd: python src/train.py data/features model.pkl
     deps:
     - path: data/features
       hash: md5
-      md5: 4281fdd8e973e3bbe5abc0ae10adebc7.dir
-      size: 2232588
+      md5: c9308f114f6a8f06fb5ba2b40ea81678.dir
+      size: 12597137
       nfiles: 2
     - path: src/train.py
       hash: md5
@@ -63,20 +63,20 @@ stages:
     outs:
     - path: model.pkl
       hash: md5
-      md5: 46f38c08e3d5174e5e3fb8753994d38b
-      size: 1957931
+      md5: b568c889ca6a5719632188daa0bfd513
+      size: 3365545
   evaluate:
     cmd: python src/evaluate.py model.pkl data/features
     deps:
     - path: data/features
       hash: md5
-      md5: 4281fdd8e973e3bbe5abc0ae10adebc7.dir
-      size: 2232588
+      md5: c9308f114f6a8f06fb5ba2b40ea81678.dir
+      size: 12597137
       nfiles: 2
     - path: model.pkl
       hash: md5
-      md5: 46f38c08e3d5174e5e3fb8753994d38b
-      size: 1957931
+      md5: b568c889ca6a5719632188daa0bfd513
+      size: 3365545
     - path: src/evaluate.py
       hash: md5
       md5: a1a59f55636170fb56e0c6afd3e28fa4
@@ -84,6 +84,6 @@ stages:
     outs:
     - path: eval
       hash: md5
-      md5: 0deddc7fb86151e1cb4684e93be58f70.dir
-      size: 1292365
+      md5: 6fe98138454e84433ffaf097fc5cfd51.dir
+      size: 4964239
       nfiles: 8
diff --git a/params.yaml b/params.yaml
index 954cc22b5..b16b7b66f 100644
--- a/params.yaml
+++ b/params.yaml
@@ -3,7 +3,7 @@ prepare:
   seed: 20170428
 
 featurize:
-  max_features: 200
+  max_features: 500
   ngrams: 2
 
 train: