[Fix] Fix IFEval (#906)

* fix ifeval

* fix ifeval

* fix ifeval

* fix ifeval

configs/datasets/IFEval/IFEval_gen.py

@@ -1,4 +1,4 @@
 from mmengine.config import read_base
 
 with read_base():
-    from .IFEval_gen_3321a3 import humaneval_datasets  # noqa: F401, F403
+    from .IFEval_gen_3321a3 import ifeval_datasets  # noqa: F401, F403

configs/datasets/IFEval/IFEval_gen_3321a3.py

@@ -26,7 +26,7 @@ ifeval_datasets = [
     dict(
         abbr='IFEval',
         type=IFEvalDataset,
-        path='./data_1/ifeval/input_data.jsonl',
+        path='data/ifeval/input_data.jsonl',
         reader_cfg=ifeval_reader_cfg,
         infer_cfg=ifeval_infer_cfg,
         eval_cfg=ifeval_eval_cfg)

opencompass/datasets/IFEval/evaluation_main.py

@@ -20,7 +20,7 @@ from typing import Dict, Optional, Union
 
 from absl import flags
 
-import opencompass.datasets.ifeval.instructions_registry as instructions_registry
+import opencompass.datasets.IFEval.instructions_registry as instructions_registry
 
 _INPUT_DATA = flags.DEFINE_string('input_data',
                                   None,
@@ -69,8 +69,6 @@ def test_instruction_following_strict(
         instruction_cls = instructions_registry.INSTRUCTION_DICT[
             instruction_id]
         instruction = instruction_cls(instruction_id)
-        print(index)
-        print(inp)
         instruction.build_description(**inp.kwargs[index])
         args = instruction.get_instruction_args()
         if args and 'prompt' in args:

opencompass/datasets/IFEval/ifeval.py

@@ -29,7 +29,6 @@ class IFEvaluator(BaseEvaluator):
     def score(self, predictions, references):
         results = []
         for pred, refer in zip(predictions, references):
-            print(refer)
             input = InputExample(
                 key=refer['key'],
                 instruction_id_list=refer['instruction_id_list'],

opencompass/datasets/IFEval/instructions.py

@@ -22,10 +22,14 @@ import re
 import string
 from typing import Dict, Optional, Sequence, Union
 
-import langdetect
+try:
+    import langdetect
+except ImportError:
+    langdetect = None
+
 from absl import logging
 
-import opencompass.datasets.ifeval.instructions_util as instructions_util
+import opencompass.datasets.IFEval.instructions_util as instructions_util
 
 _InstructionArgsDtype = Optional[Dict[str, Union[int, str, Sequence[str]]]]
 

opencompass/datasets/IFEval/instructions_registry.py

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """Registry of all instructions."""
-import opencompass.datasets.ifeval.instructions as instructions
+import opencompass.datasets.IFEval.instructions as instructions
 
 _KEYWORD = 'keywords:'
 

opencompass/datasets/__init__.py

@@ -47,7 +47,7 @@ from .huggingface import *  # noqa: F401, F403
 from .humaneval import *  # noqa: F401, F403
 from .humanevalx import *  # noqa: F401, F403
 from .hungarian_math import *  # noqa: F401, F403
-from .IFEval import *  # noqa: F401, F403
+from .IFEval.ifeval import IFEvalDataset, IFEvaluator  # noqa: F401, F403
 from .infinitebench import *  # noqa: F401, F403
 from .iwslt2017 import *  # noqa: F401, F403
 from .jigsawmultilingual import *  # noqa: F401, F403

requirements/runtime.txt

@@ -9,7 +9,9 @@ evaluate>=0.3.0
 fairscale
 func_timeout
 fuzzywuzzy
+immutabledict
 jieba
+langdetect
 ltp
 mmengine-lite
 nltk==3.8