add ir4xor2-deepseek

README.md

@@ -41,9 +41,9 @@ configs:
 - config_name: ir4xor2-deepseek
   data_files:
   - split: train
-    path: ir4xor2-deepseek/train-*
+    path: data/ir4xor2-deepseek/train*
   - split: test
-    path: ir4xor2-deepseek/test-*
+    path: data/ir4xor2-deepseek/test*
 dataset_info:
 - config_name: default
   features:
@@ -54,21 +54,6 @@ dataset_info:
   splits:
   - name: train
   - name: test
-- config_name: ir4xor2-deepseek
-  features:
-  - name: input
-    dtype: string
-  - name: output
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 258915789
-    num_examples: 64643
-  - name: test
-    num_bytes: 3984288
-    num_examples: 1000
-  download_size: 86012849
-  dataset_size: 262900077
 language:
 - code
 size_categories:
@@ -104,4 +89,4 @@ If you use RepairLLaMA in academic research, please cite "[RepairLLaMA: Efficien
   number = {2312.15698},
   institution = {arXiv},
 }
-```
+```

data/ir4xor2-deepseek/README.md

@@ -0,0 +1,7 @@
+`input`: Buggy function w/ infilling mask and buggy code (IR4)
+`output`: Fixed chunk (OR2)
+
+The dataset is target at DeepSeek FIM models (i.e., uses their special tokens)
+
+The size of training data is: 64643
+The size of testing data is: 1000

ir4xor2-deepseek/test-00000-of-00001.parquet → data/ir4xor2-deepseek/test.jsonl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7efc2e76d0722c350b4b966c8fe36585ae65b93b63efc096866244d69a54a2d6
-size 1303879
+oid sha256:fe61005222ee8275754e39d0dddf82911e644b9b8e7ee02ecf18b1ea84fd4140
+size 4301365

ir4xor2-deepseek/train-00000-of-00001.parquet → data/ir4xor2-deepseek/train.jsonl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:48a3ccba7860e95ac902ed052a83b6d1862e1abd44123fe16f5096d9e63af1aa
-size 84708970
+oid sha256:5699c781e7cc13a68e0e167e450e0cd13b2fe4586f2ca6516968d48906babe21
+size 280600693