From 7d8cb9afec8017fa1006df377028f9648c40d3b5 Mon Sep 17 00:00:00 2001
From: Yizhou Chi <chiyizhou@fuzhi.ai>
Date: Fri, 13 Sep 2024 19:10:45 +0800
Subject: [PATCH] add image datasets config

---
 expo/data/hf_data.py |  6 +++---
 expo/datasets.yaml   | 24 ++++++++++++++++++++++++
 2 files changed, 27 insertions(+), 3 deletions(-)

diff --git a/expo/data/hf_data.py b/expo/data/hf_data.py
index 6f615c8cb..df3a6ed20 100644
--- a/expo/data/hf_data.py
+++ b/expo/data/hf_data.py
@@ -115,8 +115,8 @@ class HFExpDataset(ExpDataset):
 
 if __name__ == "__main__":
     dataset_dir = "D:/work/automl/datasets"
-    save_analysis_pool = True
-    force_update = True
+    save_analysis_pool = False
+    force_update = False
     datasets_dict = {"datasets": {}}
     solution_designer = SolutionDesigner()
     for dataset_meta in HFDATSETS:
@@ -125,7 +125,7 @@ if __name__ == "__main__":
             dataset_dir,
             dataset_meta["dataset_name"],
             target_col=dataset_meta["target_col"],
-            image_col=dataset_meta["image_col"],
+            image_col=dataset_meta.get("image_col", ""),
             force_update=force_update,
             modality=dataset_meta["modality"],
         )
diff --git a/expo/datasets.yaml b/expo/datasets.yaml
index 051e8232d..92e004c6d 100644
--- a/expo/datasets.yaml
+++ b/expo/datasets.yaml
@@ -183,6 +183,14 @@ datasets:
       \ column `label`.\nPerform data analysis, data preprocessing, feature engineering,\
       \ and modeling to predict the target. \nReport f1 weighted on the eval data.\
       \ Do not plot or make any visualizations.\n"
+  fashion_mnist:
+    dataset: fashion_mnist
+    metric: f1 weighted
+    target_col: label
+    user_requirement: "This is a fashion_mnist dataset. Your goal is to predict the\
+      \ target column `label`.\nPerform data analysis, data preprocessing, feature\
+      \ engineering, and modeling to predict the target. \nReport f1 weighted on the\
+      \ eval data. Do not plot or make any visualizations.\n"
   gnad10:
     dataset: gnad10
     metric: f1 weighted
@@ -191,6 +199,14 @@ datasets:
       \ column `label`.\nPerform data analysis, data preprocessing, feature engineering,\
       \ and modeling to predict the target. \nReport f1 weighted on the eval data.\
       \ Do not plot or make any visualizations.\n"
+  oxford-iiit-pet:
+    dataset: oxford-iiit-pet
+    metric: f1
+    target_col: label_cat_dog
+    user_requirement: "This is a oxford-iiit-pet dataset. Your goal is to predict\
+      \ the target column `label_cat_dog`.\nPerform data analysis, data preprocessing,\
+      \ feature engineering, and modeling to predict the target. \nReport f1 on the\
+      \ eval data. Do not plot or make any visualizations.\n"
   sms_spam:
     dataset: sms_spam
     metric: f1
@@ -199,3 +215,11 @@ datasets:
       \ column `label`.\nPerform data analysis, data preprocessing, feature engineering,\
       \ and modeling to predict the target. \nReport f1 on the eval data. Do not plot\
       \ or make any visualizations.\n"
+  stanford_cars:
+    dataset: stanford_cars
+    metric: f1 weighted
+    target_col: label
+    user_requirement: "This is a stanford_cars dataset. Your goal is to predict the\
+      \ target column `label`.\nPerform data analysis, data preprocessing, feature\
+      \ engineering, and modeling to predict the target. \nReport f1 weighted on the\
+      \ eval data. Do not plot or make any visualizations.\n"