[Feat]: encode sft data with mt5 tokenizer

README.md

@@ -1,31 +1,31 @@
 ---
 dataset_info:
   features:
-  - name: instruction
-    dtype: string
-  - name: input
-    dtype: string
-  - name: response
-    dtype: string
+  - name: input_ids
+    sequence: int32
+  - name: attention_mask
+    sequence: int8
+  - name: labels
+    sequence: int64
   splits:
   - name: train
-    num_bytes: 187357244
+    num_bytes: 146421104
     num_examples: 37516
   - name: validation
-    num_bytes: 20974221
+    num_bytes: 18453961
     num_examples: 4689
   - name: test
-    num_bytes: 20928020
+    num_bytes: 18468141
     num_examples: 4689
-  download_size: 108682624
-  dataset_size: 229259485
+  download_size: 59047247
+  dataset_size: 183343206
 configs:
 - config_name: default
   data_files:
   - split: train
-    path: sft/train-*
+    path: mt5base-encoded/train-*
   - split: validation
-    path: sft/validation-*
+    path: mt5base-encoded/validation-*
   - split: test
-    path: sft/test-*
+    path: mt5base-encoded/test-*
 ---

mt5base-encoded/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4516bcd21567da2c7bacf51e3675c5e7c632e877e0ffc10460f87afb0c55ca66
+size 5964742

mt5base-encoded/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c8285c0d466c16089e905fe8573c961efb456a8df1e1da786fa3e13e0f1f5b2
+size 47124814

mt5base-encoded/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:234491067a730e79c9ac63f758b1bda2f0460bc86a92b191e94cabde2ec26e50
+size 5957691