pre-tokenized with gpt2-medium

README.md

@@ -76,10 +76,10 @@ dataset_info:
     dtype: int64
   splits:
   - name: train
-    num_bytes: 296997624
+    num_bytes: 585290772
     num_examples: 41890
-  download_size: 101120287
-  dataset_size: 296997624
+  download_size: 158058450
+  dataset_size: 585290772
 configs:
 - config_name: julia
   data_files:

scheme/{train-00000-of-00001.parquet → train-00000-of-00002.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fc113cf6a9a816974bbe58a62b16c6a4eb29adf47b955156020c16de3af2d30c
-size 101120287
+oid sha256:86901011480f325dfef6d17d2e3e16c50282ad41f4bf6e0eb0a0ea997f9ae8c2
+size 76797636

scheme/train-00001-of-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d12cb9065e656b70ee20c9da618f5276c0288d7860f7cec101d71640095e164b
+size 81260814