adding csv files from internal kaggle

README.md

@@ -1,11 +1,17 @@
 ---
 configs:
-- config_name: default
+- config_name: cs4782
   data_files:
   - split: train
     path: "train/4782_train_x.csv"
   - split: test
     path: "test/4782_test_x.csv"
+- config_name: default
+  data_files:
+  - split: train
+    path: "internal_kaggle/ar-xive-project-baselines/train.csv"
+  - split: test
+    path: "internal_kaggle/ar-xive-project-baselines/test_papers.csv"
 ---
 
 # arXiv Classifier Data
@@ -27,6 +33,13 @@ Working doc: https://docs.google.com/document/d/1ZTiR5M7NBpsfYbJdlP6qcIldBJVzJI1
 
 ## Preprocessing
 
-Code: [4782_preprocess_data.ipynb](4782_preprocess_data.ipynb)
+4782 Kaggle code: [4782_preprocess_data.ipynb](4782_preprocess_data.ipynb)
+
+Internal Kaggle code: [clean_data_for_hugging_face.ipynb](https://huggingface.co/datasets/mlcore/arxiv-classifier/blob/main/clean_data_for_hugging_face.ipynb)
+
+NOTE: We only include v1 of papers (to avoid any data leakage between train and test splits). There might be a better way of deduplicating (TODO).
+
+## Todo
 
-We only include v1 of papers (to avoid any data leakage between train and test splits)
+- Can we get the full text for the 2023 corpus/all papers?
+- Can we incorporate additional metadata (e.g., authors)?

internal_kaggle/ar-xive-project-baselines/test_papers.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:719faa8663e5704422a868b506e12158ae0c5a069eea2214fa8e085ad0488c6d
+size 763540298

internal_kaggle/ar-xive-project-baselines/train.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9efb39f9f26aa896c41b8697ab9a9d249d8f34ea16a1a3ffe2c3d8485af1241e
+size 3101585187