[Sync] update taco (#1030)

configs/summarizers/compass_knowledge.py

@@ -12,7 +12,6 @@ compassbench_v1_knowledge_groups = [
     {'name': 'knowledge_perf_4_and_cloze', 'subsets': [['knowledge_cn', 'perf_4'], ['compassbench_v1_knowledge-mixed-cloze_en', 'score']]},
 ]
 
-'compassbench_v1_knowledge-mixed-cloze_en'
 summarizer = dict(
     dataset_abbrs=[
         'knowledge_perf_4_and_cloze',

configs/summarizers/compassbench_v1_objective.py

+
+from mmengine.config import read_base
+
+with read_base():
+    from .groups.cibench import cibench_summary_groups
+    from .groups.plugineval import plugineval_summary_groups
+
+
+compassbench_v1_language_names = [
+    # ['information_retrieval_en', 'score'],
+    # ['information_retrieval_zh', 'score'],
+    ['intention_recognition_en_circular', 'acc_origin'],
+    ['intention_recognition_en_circular', 'perf_circular'],
+    ['intention_recognition_zh_circular', 'acc_origin'],
+    ['intention_recognition_zh_circular', 'perf_circular'],
+    ['sentiment_analysis_en_circular', 'acc_origin'],
+    ['sentiment_analysis_en_circular', 'perf_circular'],
+    ['sentiment_analysis_zh_circular', 'acc_origin'],
+    ['sentiment_analysis_zh_circular', 'perf_circular'],
+    ['translation', 'score'],
+    ['content_critic_en_circular', 'acc_origin'],
+    ['content_critic_en_circular', 'perf_circular'],
+    ['content_critic_zh_circular', 'acc_origin'],
+    ['content_critic_zh_circular', 'perf_circular'],
+    ['content_summarization_en', 'rouge1'],
+    ['content_summarization_zh', 'rouge1'],
+    ['traditional_cultural_understanding_zh_circular', 'acc_origin'],
+    ['traditional_cultural_understanding_zh_circular', 'perf_circular'],
+    ['chinese_semantic_understanding_zh_circular', 'acc_origin'],
+    ['chinese_semantic_understanding_zh_circular', 'perf_circular'],
+]
+
+compassbench_v1_language_summary_groups = [
+    {'name': 'language_zh_acc_1_and_non_mcq', 'subsets': [[name, metric] for name, metric in compassbench_v1_language_names if '_zh' in name and metric != 'perf_circular']},
+    {'name': 'language_en_acc_1_and_non_mcq', 'subsets': [[name, metric] for name, metric in compassbench_v1_language_names if '_en' in name and metric != 'perf_circular']},
+    {'name': 'language_acc_1_and_non_mcq', 'subsets': ['language_zh_acc_1_and_non_mcq', 'language_en_acc_1_and_non_mcq']},
+
+    {'name': 'language_zh_perf_4_and_non_mcq', 'subsets': [[name, metric] for name, metric in compassbench_v1_language_names if '_zh' in name and metric != 'acc_origin']},
+    {'name': 'language_en_perf_4_and_non_mcq', 'subsets': [[name, metric] for name, metric in compassbench_v1_language_names if '_en' in name and metric != 'acc_origin']},
+    {'name': 'language_perf_4_and_non_mcq', 'subsets': ['language_zh_perf_4_and_non_mcq', 'language_en_perf_4_and_non_mcq']},
+]
+
+# This summarizer is used for `./datasets/compassbench_v1_knowledge/compassbench_v1_knowledge_gen`
+compassbench_v1_knowledge_names = [
+    'compassbench_v1_knowledge-common_knowledge-single_choice_cn_circular',
+    'compassbench_v1_knowledge-humanity-single_choice_cn_circular',
+    'compassbench_v1_knowledge-natural_science-single_choice_cn_circular',
+    'compassbench_v1_knowledge-social_science-single_choice_cn_circular',
+]
+
+compassbench_v1_knowledge_summary_groups = [
+    {'name': 'knowledge_cn', 'subsets': compassbench_v1_knowledge_names},
+    {'name': 'knowledge_acc_1_and_cloze', 'subsets': [['knowledge_cn', 'acc_1'], ['compassbench_v1_knowledge-mixed-cloze_en', 'score']]},
+    {'name': 'knowledge_perf_4_and_cloze', 'subsets': [['knowledge_cn', 'perf_4'], ['compassbench_v1_knowledge-mixed-cloze_en', 'score']]},
+]
+
+compassbench_v1_reason_summary_groups = [
+    {'name': 'reasonbench_cn_abductive_circular', 'subsets': ['reasonbench_cn_abductive_alphanlg_translated_circular']},
+    {'name': 'reasonbench_en_abductive_circular', 'subsets': ['reasonbench_en_abductive_alphanlg_circular']},
+    {'name': 'reasonbench_cn_deductive_circular', 'subsets': ['reasonbench_cn_deductive_bbh3obj_translated_circular', 'reasonbench_cn_deductive_logiqa_zh_circular']},
+    {'name': 'reasonbench_cn_inductive_circular', 'subsets': ['reasonbench_cn_inductive_deer_translated_circular', 'reasonbench_cn_inductive_selfgenerated_circular']},
+    {'name': 'reasonbench_en_inductive_circular', 'subsets': ['reasonbench_en_inductive_deer_circular', 'reasonbench_en_inductive_selfgenerated_circular']},
+
+    {'name': 'reasonbench_cn_circular', 'subsets': ['reasonbench_cn_commonsense_circular', 'reasonbench_cn_abductive_circular', 'reasonbench_cn_deductive_circular', 'reasonbench_cn_inductive_circular']},
+    {'name': 'reasonbench_en_circular', 'subsets': ['reasonbench_en_commonsense_circular', 'reasonbench_en_abductive_circular', 'reasonbench_en_deductive_logiqa_zh_translated_circular', 'reasonbench_en_inductive_circular']},
+    {'name': 'reasonbench', 'subsets': ['reasonbench_cn_circular', 'reasonbench_en_circular']},
+]
+
+compassbench_v1_math_summary_groups = [
+    {'name': 'math_acc_1_and_fill_in_blank', 'subsets': [['compassbench_v1_math-high-single_choice_cn', 'acc_1'], ['compassbench_v1_math-high-single_choice_en', 'acc_1'], ['compassbench_v1_math-middle-single_choice_cn', 'acc_1'], ['compassbench_v1_math-middle-single_choice_en', 'acc_1'], ['compassbench_v1_math-primary-cloze_cn', 'accuracy'], ['compassbench_v1_math-primary-cloze_en', 'accuracy']]},
+    {'name': 'math_perf_4_and_fill_in_blank', 'subsets': [['compassbench_v1_math-high-single_choice_cn', 'perf_4'], ['compassbench_v1_math-high-single_choice_en', 'perf_4'], ['compassbench_v1_math-middle-single_choice_cn', 'perf_4'], ['compassbench_v1_math-middle-single_choice_en', 'perf_4'], ['compassbench_v1_math-primary-cloze_cn', 'accuracy'], ['compassbench_v1_math-primary-cloze_en', 'accuracy']]},
+    {'name': 'math_perf_4_and_fill_in_blank_cn', 'subsets': [['compassbench_v1_math-high-single_choice_cn', 'perf_4'], ['compassbench_v1_math-middle-single_choice_cn', 'perf_4'], ['compassbench_v1_math-primary-cloze_cn', 'accuracy']]},
+    {'name': 'math_perf_4_and_fill_in_blank_en', 'subsets': [['compassbench_v1_math-high-single_choice_en', 'perf_4'], ['compassbench_v1_math-middle-single_choice_en', 'perf_4'], ['compassbench_v1_math-primary-cloze_en', 'accuracy']]},
+]
+
+
+code_passk_summary_groups = [
+    # rename
+    {'name': 'humaneval_pass@1(greedy)', 'subsets': [['openai_humaneval', 'humaneval_pass@1']]},
+    {'name': 'humaneval_pass@10', 'subsets': [['openai_humaneval_passk', 'humaneval_pass@10']]},
+    {'name': 'humaneval_pass@10', 'subsets': [['openai_humaneval_repeat10', 'humaneval_pass@10']]},
+    {'name': 'humaneval_cn_pass@1(greedy)', 'subsets': [['openai_humaneval_cn', 'humaneval_pass@1']]},
+    {'name': 'humaneval_cn_pass@10', 'subsets': [['openai_humaneval_cn_passk', 'humaneval_pass@10']]},
+    {'name': 'humaneval_cn_pass@10', 'subsets': [['openai_humaneval_cn_repeat10', 'humaneval_pass@10']]},
+    {'name': 'humaneval_plus_pass@1(greedy)', 'subsets': [['humaneval_plus', 'humaneval_plus_pass@1']]},
+    {'name': 'humaneval_plus_pass@10', 'subsets': [['humaneval_plus_passk', 'humaneval_plus_pass@10']]},
+    {'name': 'humaneval_plus_pass@10', 'subsets': [['humaneval_plus_repeat10', 'humaneval_plus_pass@10']]},
+    {'name': 'mbpp_pass@1(greedy)', 'subsets': [['mbpp', 'score']]},
+    {'name': 'mbpp_pass@10', 'subsets': [['mbpp_passk', 'pass@10']]},
+    {'name': 'mbpp_pass@10', 'subsets': [['mbpp_repeat10', 'pass@10']]},
+    {'name': 'mbpp_cn_pass@1(greedy)', 'subsets': [['mbpp_cn', 'score']]},
+    {'name': 'mbpp_cn_pass@10', 'subsets': [['mbpp_cn_passk', 'pass@10']]},
+    {'name': 'mbpp_cn_pass@10', 'subsets': [['mbpp_cn_repeat10', 'pass@10']]},
+    {'name': 'sanitized_mbpp_pass@1(greedy)', 'subsets': [['sanitized_mbpp', 'score']]},
+    {'name': 'sanitized_mbpp_pass@10', 'subsets': [['sanitized_mbpp_passk', 'pass@10']]},
+    {'name': 'sanitized_mbpp_pass@10', 'subsets': [['sanitized_mbpp_repeat10', 'pass@10']]},
+    # real add
+    {'name': 'humanevalx', 'subsets': ['humanevalx-python', 'humanevalx-cpp', 'humanevalx-go', 'humanevalx-java', 'humanevalx-js']},
+    {'name': 'code_cn', 'subsets': ['humaneval_cn_pass@1(greedy)', 'mbpp_cn_pass@1(greedy)']},
+    {'name': 'code_en', 'subsets': ['humaneval_plus_pass@1(greedy)', 'sanitized_mbpp_pass@1(greedy)', 'humanevalx']},
+    {'name': 'code', 'subsets': ['humaneval_cn_pass@1(greedy)', 'mbpp_cn_pass@1(greedy)', 'humaneval_plus_pass@1(greedy)', 'sanitized_mbpp_pass@1(greedy)', 'humanevalx']},
+]
+
+agent_summary_groups = [
+    # dict(name='cibench_template', subsets=['cibench_template:executable', 'cibench_template:numeric_correct', 'cibench_template:text_score', 'cibench_template:vis_sim']),
+    # dict(name='cibench_template_cn', subsets=['cibench_template_cn:executable', 'cibench_template_cn:numeric_correct', 'cibench_template_cn:text_score', 'cibench_template_cn:vis_sim']),
+
+    dict(name='cibench_template', subsets=['cibench_template_wo_nltk:executable', 'cibench_template_wo_nltk:numeric_correct', 'cibench_template_wo_nltk:vis_sim']),
+    dict(name='cibench_template_cn', subsets=['cibench_template_cn_wo_nltk:executable', 'cibench_template_cn_wo_nltk:numeric_correct', 'cibench_template_cn_wo_nltk:vis_sim']),
+
+    dict(name='agent_cn', subsets=['cibench_template_cn', 'plugin_eval-mus-p10_one_review_zh']),
+    dict(name='agent_en', subsets=['cibench_template', 'plugin_eval-mus-p10_one_review']),
+    dict(name='agent', subsets=['agent_cn', 'agent_en']),
+]
+
+other_summary_groups = [
+    {
+        "name": "average_cn",
+        "subsets": [
+            ["language_zh_perf_4_and_non_mcq", "naive_average"],
+            ["knowledge_cn", "perf_4"],
+            ["reasonbench_cn_circular", "perf_circular"],
+            ["math_perf_4_and_fill_in_blank_cn", "naive_average"],
+            ["code_cn", "naive_average"],
+            ["agent_cn", "naive_average"],
+        ],
+    },
+    {
+        "name": "average_en",
+        "subsets": [
+            ["language_en_perf_4_and_non_mcq", "naive_average"],
+            ["compassbench_v1_knowledge-mixed-cloze_en", "score"],
+            ["reasonbench_en_circular", "perf_circular"],
+            ["math_perf_4_and_fill_in_blank_en", "naive_average"],
+            ["code_en", "naive_average"],
+            ["agent_en", "naive_average"],
+        ],
+    },
+    {
+        "name": "average",
+        "subsets": [
+            ["language_perf_4_and_non_mcq", "naive_average"],
+            ["knowledge_perf_4_and_cloze", "naive_average"],
+            ["reasonbench", "perf_circular"],
+            ["math_perf_4_and_fill_in_blank", "naive_average"],
+            ["code", "naive_average"],
+            ["agent", "naive_average"],
+        ],
+    },
+]
+
+
+
+
+summarizer = dict(
+    dataset_abbrs=[
+        ['average', 'naive_average'],
+        ['average_cn', 'naive_average'],
+        ['average_en', 'naive_average'],
+        '',
+        '',
+        '',
+
+        ['language_perf_4_and_non_mcq', 'naive_average'],
+        ['language_zh_perf_4_and_non_mcq', 'naive_average'],
+        ['language_en_perf_4_and_non_mcq', 'naive_average'],
+        ['intention_recognition_zh_circular', 'perf_circular'],
+        ['intention_recognition_en_circular', 'perf_circular'],
+        ['sentiment_analysis_zh_circular', 'perf_circular'],
+        ['sentiment_analysis_en_circular', 'perf_circular'],
+        ['translation', 'score'],
+        ['content_critic_zh_circular', 'perf_circular'],
+        ['content_critic_en_circular', 'perf_circular'],
+        ['content_summarization_zh', 'rouge1'],
+        ['content_summarization_en', 'rouge1'],
+        ['traditional_cultural_understanding_zh_circular', 'perf_circular'],
+        ['chinese_semantic_understanding_zh_circular', 'perf_circular'],
+
+        ['knowledge_perf_4_and_cloze', 'naive_average'],
+        ['knowledge_cn', 'perf_4'],
+        ['compassbench_v1_knowledge-mixed-cloze_en', 'score'],
+        ['compassbench_v1_knowledge-common_knowledge-single_choice_cn_circular', 'perf_4'],
+        ['compassbench_v1_knowledge-humanity-single_choice_cn_circular', 'perf_4'],
+        ['compassbench_v1_knowledge-natural_science-single_choice_cn_circular', 'perf_4'],
+        ['compassbench_v1_knowledge-social_science-single_choice_cn_circular', 'perf_4'],
+
+        ['reasonbench', 'perf_circular'],
+        ['reasonbench_cn_circular', 'perf_circular'],
+        ['reasonbench_en_circular', 'perf_circular'],
+        ['reasonbench_cn_commonsense_circular', 'perf_circular'],
+        ['reasonbench_cn_abductive_circular', 'perf_circular'],
+        ['reasonbench_cn_deductive_circular', 'perf_circular'],
+        ['reasonbench_cn_inductive_circular', 'perf_circular'],
+        ['reasonbench_en_commonsense_circular', 'perf_circular'],
+        ['reasonbench_en_abductive_circular', 'perf_circular'],
+        ['reasonbench_en_deductive_logiqa_zh_translated_circular', 'perf_circular'],
+        ['reasonbench_en_inductive_circular', 'perf_circular'],
+
+        ['math_perf_4_and_fill_in_blank', 'naive_average'],
+        ['math_perf_4_and_fill_in_blank_cn', 'naive_average'],
+        ['math_perf_4_and_fill_in_blank_en', 'naive_average'],
+        ['compassbench_v1_math-high-single_choice_cn', 'perf_4'],
+        ['compassbench_v1_math-high-single_choice_en', 'perf_4'],
+        ['compassbench_v1_math-middle-single_choice_cn', 'perf_4'],
+        ['compassbench_v1_math-middle-single_choice_en', 'perf_4'],
+        ['compassbench_v1_math-primary-cloze_cn', 'accuracy'],
+        ['compassbench_v1_math-primary-cloze_en', 'accuracy'],
+
+        ['code', 'naive_average'],
+        ['code_cn', 'naive_average'],
+        ['code_en', 'naive_average'],
+        ['humaneval_cn_pass@1(greedy)', 'naive_average'],
+        ['humaneval_plus_pass@1(greedy)', 'naive_average'],
+        ['mbpp_cn_pass@1(greedy)', 'naive_average'],
+        ['sanitized_mbpp_pass@1(greedy)', 'naive_average'],
+        ['humanevalx', 'naive_average'],
+
+        ['agent', 'naive_average'],
+        ['agent_cn', 'naive_average'],
+        ['agent_en', 'naive_average'],
+        ['cibench_template_cn', 'naive_average'],
+        ['cibench_template', 'naive_average'],
+        ['plugin_eval-mus-p10_one_review_zh', 'naive_average'],
+        ['plugin_eval-mus-p10_one_review', 'naive_average'],
+    ],
+    summary_groups=sum([v for k, v in locals().items() if k.endswith("_summary_groups")], []),
+)

configs/summarizers/groups/lcbench.py

+lcbench_summary_groups = [
+    {'name': 'lcbench', 'subsets': ['lcbench_en', 'lcbench_cn']},
+]

configs/summarizers/groups/mathbench_v1.py

+mathbench_v1_summary_groups = [
+    {'name': 'mathbench-college_application', 'subsets': ['mathbench-college-single_choice_cn', 'mathbench-college-single_choice_en']},
+    {'name': 'mathbench-high_application', 'subsets': ['mathbench-high-single_choice_cn', 'mathbench-high-single_choice_en']},
+    {'name': 'mathbench-middle_application', 'subsets': ['mathbench-middle-single_choice_cn', 'mathbench-middle-single_choice_en']},
+    {'name': 'mathbench-primary_application', 'subsets': ['mathbench-primary-cloze_cn', 'mathbench-primary-cloze_en', 'mathbench-calculate-cloze_en'], 'weights': {'mathbench-primary-cloze_cn': 1, 'mathbench-primary-cloze_en': 1, 'mathbench-calculate-cloze_en': 2}},
+    {'name': 'mathbench-college_knowledge', 'subsets': ['mathbench-college_knowledge-single_choice_cn', 'mathbench-college_knowledge-single_choice_en']},
+    {'name': 'mathbench-high_knowledge', 'subsets': ['mathbench-high_knowledge-single_choice_cn', 'mathbench-high_knowledge-single_choice_en']},
+    {'name': 'mathbench-middle_knowledge', 'subsets': ['mathbench-middle_knowledge-single_choice_cn', 'mathbench-middle_knowledge-single_choice_en']},
+    {'name': 'mathbench-primary_knowledge', 'subsets': ['mathbench-primary_knowledge-single_choice_cn', 'mathbench-primary_knowledge-single_choice_en']},
+    {'name': 'mathbench_application', 'subsets': ['mathbench-college_application', 'mathbench-high_application', 'mathbench-middle_application', 'mathbench-primary_application']},
+    {'name': 'mathbench_knowledge', 'subsets': ['mathbench-college_knowledge', 'mathbench-high_knowledge', 'mathbench-middle_knowledge', 'mathbench-primary_knowledge']},
+    {'name': 'mathbench', 'subsets': ['mathbench_application', 'mathbench_knowledge']},
+]

configs/summarizers/groups/plugineval.py

@@ -71,6 +71,40 @@ _base_summary_groups = [
             ['plugin_eval-review_str_v1', 'review_quality'],
         ]
     },
+
+    # special treatment for first 10% data points
+    {
+        'name': 'plugin_eval-p10-instruct_v1',
+        'metric': 'format_metric',
+        'subsets': [
+            ['plugin_eval-p10-instruct_v1', 'string_format_metric'],
+            ['plugin_eval-p10-instruct_v1', 'json_format_metric'],
+        ]
+    },
+    {
+        'name': 'plugin_eval-p10-instruct_v1',
+        'metric': 'args_em_metric',
+        'subsets': [
+            ['plugin_eval-p10-instruct_v1', 'string_args_em_metric'],
+            ['plugin_eval-p10-instruct_v1', 'json_args_em_metric'],
+        ]
+    },
+    {
+        'name': 'plugin_eval-p10',
+        'subsets': [
+            ['plugin_eval-p10-instruct_v1', 'format_metric'],
+            ['plugin_eval-p10-instruct_v1', 'args_em_metric'],
+            ['plugin_eval-p10-plan_str_v1', 'f1_score'],
+            ['plugin_eval-p10-plan_json_v1', 'f1_score'],
+            ['plugin_eval-p10-reason_str_v2', 'thought'],
+            ['plugin_eval-p10-reason_retrieve_understand_json_v2', 'thought'],
+            ['plugin_eval-p10-retrieve_str_v2', 'name'],
+            ['plugin_eval-p10-reason_retrieve_understand_json_v2', 'name'],
+            ['plugin_eval-p10-understand_str_v2', 'args'],
+            ['plugin_eval-p10-reason_retrieve_understand_json_v2', 'args'],
+            ['plugin_eval-p10-review_str_v6', 'review_quality'],
+        ]
+    },
 ]
 
 plugineval_summary_groups = []

configs/summarizers/mathbench_v1.py

 summarizer = dict(
     dataset_abbrs=[
-        '######## MathBench Accuracy ########', # category
+        '######## MathBench Application Accuracy ########', # category
         ['mathbench-college-single_choice_cn', 'acc_1'],
         ['mathbench-college-single_choice_en', 'acc_1'],
         ['mathbench-high-single_choice_cn', 'acc_1'],
@@ -9,15 +9,15 @@ summarizer = dict(
         ['mathbench-middle-single_choice_en', 'acc_1'],
         ['mathbench-primary-cloze_cn', 'accuracy'],
         ['mathbench-primary-cloze_en', 'accuracy'],
-        ['mathbench-calculate-cloze_en', 'accuracy'],
-        '######## MathBench CircularEval ########', # category
+        ['mathbench-arithmetic-cloze_en', 'accuracy'],
+        '######## MathBench Application CircularEval ########', # category
         ['mathbench-college-single_choice_cn', 'perf_4'],
         ['mathbench-college-single_choice_en', 'perf_4'],
         ['mathbench-high-single_choice_cn', 'perf_4'],
         ['mathbench-high-single_choice_en', 'perf_4'],
         ['mathbench-middle-single_choice_cn', 'perf_4'],
         ['mathbench-middle-single_choice_en', 'perf_4'],
-        '######## MathBench Knowledge ########', # category
+        '######## MathBench Knowledge CircularEval ########', # category
         ['mathbench-college_knowledge-single_choice_cn', 'perf_4'],
         ['mathbench-college_knowledge-single_choice_en', 'perf_4'],
         ['mathbench-high_knowledge-single_choice_cn', 'perf_4'],
@@ -26,6 +26,15 @@ summarizer = dict(
         ['mathbench-middle_knowledge-single_choice_en', 'perf_4'],
         ['mathbench-primary_knowledge-single_choice_cn', 'perf_4'],
         ['mathbench-primary_knowledge-single_choice_en', 'perf_4'],
+        '######## MathBench Knowledge Accuracy ########', # category
+        ['mathbench-college_knowledge-single_choice_cn', 'acc_1'],
+        ['mathbench-college_knowledge-single_choice_en', 'acc_1'],
+        ['mathbench-high_knowledge-single_choice_cn', 'acc_1'],
+        ['mathbench-high_knowledge-single_choice_en', 'acc_1'],
+        ['mathbench-middle_knowledge-single_choice_cn', 'acc_1'],
+        ['mathbench-middle_knowledge-single_choice_en', 'acc_1'],
+        ['mathbench-primary_knowledge-single_choice_cn', 'acc_1'],
+        ['mathbench-primary_knowledge-single_choice_en', 'acc_1'],
     ],
     summary_groups=sum(
         [v for k, v in locals().items() if k.endswith("_summary_groups")], [])

opencompass/cli/main.py

+import argparse
+import getpass
+import os
+import os.path as osp
+from datetime import datetime
+
+from mmengine.config import Config, DictAction
+
+from opencompass.partitioners import MultimodalNaivePartitioner
+from opencompass.registry import PARTITIONERS, RUNNERS, build_from_cfg
+from opencompass.runners import SlurmRunner
+from opencompass.summarizers import DefaultSummarizer
+from opencompass.utils import LarkReporter, get_logger
+from opencompass.utils.run import (exec_mm_infer_runner, fill_eval_cfg,
+                                   fill_infer_cfg, get_config_from_arg)
+
+
+def parse_args():
+    parser = argparse.ArgumentParser(description='Run an evaluation task')
+    parser.add_argument('config', nargs='?', help='Train config file path')
+
+    # add mutually exclusive args `--slurm` `--dlc`, defaults to local runner
+    # if "infer" or "eval" not specified
+    launch_method = parser.add_mutually_exclusive_group()
+    launch_method.add_argument('--slurm',
+                               action='store_true',
+                               default=False,
+                               help='Whether to force tasks to run with srun. '
+                               'If True, `--partition(-p)` must be set. '
+                               'Defaults to False')
+    launch_method.add_argument('--dlc',
+                               action='store_true',
+                               default=False,
+                               help='Whether to force tasks to run on dlc. If '
+                               'True, `--aliyun-cfg` must be set. Defaults'
+                               ' to False')
+    # multi-modal support
+    parser.add_argument('--mm-eval',
+                        help='Whether or not enable multimodal evaluation',
+                        action='store_true',
+                        default=False)
+    # Add shortcut parameters (models, datasets and summarizer)
+    parser.add_argument('--models', nargs='+', help='', default=None)
+    parser.add_argument('--datasets', nargs='+', help='', default=None)
+    parser.add_argument('--summarizer', help='', default=None)
+    # add general args
+    parser.add_argument('--debug',
+                        help='Debug mode, in which scheduler will run tasks '
+                        'in the single process, and output will not be '
+                        'redirected to files',
+                        action='store_true',
+                        default=False)
+    parser.add_argument('--dry-run',
+                        help='Dry run mode, in which the scheduler will not '
+                        'actually run the tasks, but only print the commands '
+                        'to run',
+                        action='store_true',
+                        default=False)
+    parser.add_argument('-m',
+                        '--mode',
+                        help='Running mode. You can choose "infer" if you '
+                        'only want the inference results, or "eval" if you '
+                        'already have the results and want to evaluate them, '
+                        'or "viz" if you want to visualize the results.',
+                        choices=['all', 'infer', 'eval', 'viz'],
+                        default='all',
+                        type=str)
+    parser.add_argument('-r',
+                        '--reuse',
+                        nargs='?',
+                        type=str,
+                        const='latest',
+                        help='Reuse previous outputs & results, and run any '
+                        'missing jobs presented in the config. If its '
+                        'argument is not specified, the latest results in '
+                        'the work_dir will be reused. The argument should '
+                        'also be a specific timestamp, e.g. 20230516_144254')
+    parser.add_argument('-w',
+                        '--work-dir',
+                        help='Work path, all the outputs will be '
+                        'saved in this path, including the slurm logs, '
+                        'the evaluation results, the summary results, etc.'
+                        'If not specified, the work_dir will be set to '
+                        './outputs/default.',
+                        default=None,
+                        type=str)
+    parser.add_argument(
+        '--config-dir',
+        default='configs',
+        help='Use the custom config directory instead of config/ to '
+        'search the configs for datasets, models and summarizers',
+        type=str)
+    parser.add_argument('-l',
+                        '--lark',
+                        help='Report the running status to lark bot',
+                        action='store_true',
+                        default=False)
+    parser.add_argument('--max-partition-size',
+                        help='The maximum size of an infer task. Only '
+                        'effective when "infer" is missing from the config.',
+                        type=int,
+                        default=40000),
+    parser.add_argument(
+        '--gen-task-coef',
+        help='The dataset cost measurement coefficient for generation tasks, '
+        'Only effective when "infer" is missing from the config.',
+        type=int,
+        default=20)
+    parser.add_argument('--max-num-workers',
+                        help='Max number of workers to run in parallel. '
+                        'Will be overrideen by the "max_num_workers" argument '
+                        'in the config.',
+                        type=int,
+                        default=32)
+    parser.add_argument('--max-workers-per-gpu',
+                        help='Max task to run in parallel on one GPU. '
+                        'It will only be used in the local runner.',
+                        type=int,
+                        default=1)
+    parser.add_argument(
+        '--retry',
+        help='Number of retries if the job failed when using slurm or dlc. '
+        'Will be overrideen by the "retry" argument in the config.',
+        type=int,
+        default=2)
+    parser.add_argument(
+        '--dump-eval-details',
+        help='Whether to dump the evaluation details, including the '
+        'correctness of each sample, bpb, etc.',
+        action='store_true',
+    )
+    # set srun args
+    slurm_parser = parser.add_argument_group('slurm_args')
+    parse_slurm_args(slurm_parser)
+    # set dlc args
+    dlc_parser = parser.add_argument_group('dlc_args')
+    parse_dlc_args(dlc_parser)
+    # set hf args
+    hf_parser = parser.add_argument_group('hf_args')
+    parse_hf_args(hf_parser)
+    # set custom dataset args
+    custom_dataset_parser = parser.add_argument_group('custom_dataset_args')
+    parse_custom_dataset_args(custom_dataset_parser)
+    args = parser.parse_args()
+    if args.slurm:
+        assert args.partition is not None, (
+            '--partition(-p) must be set if you want to use slurm')
+    if args.dlc:
+        assert os.path.exists(args.aliyun_cfg), (
+            'When launching tasks using dlc, it needs to be configured '
+            'in "~/.aliyun.cfg", or use "--aliyun-cfg $ALiYun-CFG_Path"'
+            ' to specify a new path.')
+    return args
+
+
+def parse_slurm_args(slurm_parser):
+    """These args are all for slurm launch."""
+    slurm_parser.add_argument('-p',
+                              '--partition',
+                              help='Slurm partition name',
+                              default=None,
+                              type=str)
+    slurm_parser.add_argument('-q',
+                              '--quotatype',
+                              help='Slurm quota type',
+                              default=None,
+                              type=str)
+    slurm_parser.add_argument('--qos',
+                              help='Slurm quality of service',
+                              default=None,
+                              type=str)
+
+
+def parse_dlc_args(dlc_parser):
+    """These args are all for dlc launch."""
+    dlc_parser.add_argument('--aliyun-cfg',
+                            help='The config path for aliyun config',
+                            default='~/.aliyun.cfg',
+                            type=str)
+
+
+def parse_hf_args(hf_parser):
+    """These args are all for the quick construction of HuggingFace models."""
+    hf_parser.add_argument('--hf-path', type=str)
+    hf_parser.add_argument('--peft-path', type=str)
+    hf_parser.add_argument('--tokenizer-path', type=str)
+    hf_parser.add_argument('--model-kwargs',
+                           nargs='+',
+                           action=DictAction,
+                           default={})
+    hf_parser.add_argument('--tokenizer-kwargs',
+                           nargs='+',
+                           action=DictAction,
+                           default={})
+    hf_parser.add_argument('--max-out-len', type=int)
+    hf_parser.add_argument('--max-seq-len', type=int)
+    hf_parser.add_argument('--no-batch-padding',
+                           action='store_true',
+                           default=False)
+    hf_parser.add_argument('--batch-size', type=int)
+    hf_parser.add_argument('--num-gpus', type=int)
+    hf_parser.add_argument('--pad-token-id', type=int)
+
+
+def parse_custom_dataset_args(custom_dataset_parser):
+    """These args are all for the quick construction of custom datasets."""
+    custom_dataset_parser.add_argument('--custom-dataset-path', type=str)
+    custom_dataset_parser.add_argument('--custom-dataset-meta-path', type=str)
+    custom_dataset_parser.add_argument('--custom-dataset-data-type',
+                                       type=str,
+                                       choices=['mcq', 'qa'])
+    custom_dataset_parser.add_argument('--custom-dataset-infer-method',
+                                       type=str,
+                                       choices=['gen', 'ppl'])
+
+
+def main():
+    args = parse_args()
+    if args.dry_run:
+        args.debug = True
+    # initialize logger
+    logger = get_logger(log_level='DEBUG' if args.debug else 'INFO')
+
+    cfg = get_config_from_arg(args)
+    if args.work_dir is not None:
+        cfg['work_dir'] = args.work_dir
+    else:
+        cfg.setdefault('work_dir', './outputs/default/')
+
+    # cfg_time_str defaults to the current time
+    cfg_time_str = dir_time_str = datetime.now().strftime('%Y%m%d_%H%M%S')
+    if args.reuse:
+        if args.reuse == 'latest':
+            if not os.path.exists(cfg.work_dir) or not os.listdir(
+                    cfg.work_dir):
+                logger.warning('No previous results to reuse!')
+            else:
+                dirs = os.listdir(cfg.work_dir)
+                dir_time_str = sorted(dirs)[-1]
+        else:
+            dir_time_str = args.reuse
+        logger.info(f'Reusing experiements from {dir_time_str}')
+    elif args.mode in ['eval', 'viz']:
+        raise ValueError('You must specify -r or --reuse when running in eval '
+                         'or viz mode!')
+
+    # update "actual" work_dir
+    cfg['work_dir'] = osp.join(cfg.work_dir, dir_time_str)
+    os.makedirs(osp.join(cfg.work_dir, 'configs'), exist_ok=True)
+
+    # dump config
+    output_config_path = osp.join(cfg.work_dir, 'configs',
+                                  f'{cfg_time_str}.py')
+    cfg.dump(output_config_path)
+    # Config is intentally reloaded here to avoid initialized
+    # types cannot be serialized
+    cfg = Config.fromfile(output_config_path, format_python_code=False)
+
+    # report to lark bot if specify --lark
+    if not args.lark:
+        cfg['lark_bot_url'] = None
+    elif cfg.get('lark_bot_url', None):
+        content = f'{getpass.getuser()}\'s task has been launched!'
+        LarkReporter(cfg['lark_bot_url']).post(content)
+
+    if args.mode in ['all', 'infer']:
+        # When user have specified --slurm or --dlc, or have not set
+        # "infer" in config, we will provide a default configuration
+        # for infer
+        if (args.dlc or args.slurm) and cfg.get('infer', None):
+            logger.warning('You have set "infer" in the config, but '
+                           'also specified --slurm or --dlc. '
+                           'The "infer" configuration will be overridden by '
+                           'your runtime arguments.')
+        # Check whether run multimodal evaluation
+        if args.mm_eval:
+            partitioner = MultimodalNaivePartitioner(
+                osp.join(cfg['work_dir'], 'predictions/'))
+            tasks = partitioner(cfg)
+            exec_mm_infer_runner(tasks, args, cfg)
+            return
+
+        if args.dlc or args.slurm or cfg.get('infer', None) is None:
+            fill_infer_cfg(cfg, args)
+
+        if args.partition is not None:
+            if RUNNERS.get(cfg.infer.runner.type) == SlurmRunner:
+                cfg.infer.runner.partition = args.partition
+                cfg.infer.runner.quotatype = args.quotatype
+        else:
+            logger.warning('SlurmRunner is not used, so the partition '
+                           'argument is ignored.')
+        if args.debug:
+            cfg.infer.runner.debug = True
+        if args.lark:
+            cfg.infer.runner.lark_bot_url = cfg['lark_bot_url']
+        cfg.infer.partitioner['out_dir'] = osp.join(cfg['work_dir'],
+                                                    'predictions/')
+        partitioner = PARTITIONERS.build(cfg.infer.partitioner)
+        tasks = partitioner(cfg)
+        if args.dry_run:
+            return
+        runner = RUNNERS.build(cfg.infer.runner)
+        # Add extra attack config if exists
+        if hasattr(cfg, 'attack'):
+            for task in tasks:
+                cfg.attack.dataset = task.datasets[0][0].abbr
+                task.attack = cfg.attack
+        runner(tasks)
+
+    # evaluate
+    if args.mode in ['all', 'eval']:
+        # When user have specified --slurm or --dlc, or have not set
+        # "eval" in config, we will provide a default configuration
+        # for eval
+        if (args.dlc or args.slurm) and cfg.get('eval', None):
+            logger.warning('You have set "eval" in the config, but '
+                           'also specified --slurm or --dlc. '
+                           'The "eval" configuration will be overridden by '
+                           'your runtime arguments.')
+
+        if args.dlc or args.slurm or cfg.get('eval', None) is None:
+            fill_eval_cfg(cfg, args)
+        if args.dump_eval_details:
+            cfg.eval.runner.task.dump_details = True
+
+        if args.partition is not None:
+            if RUNNERS.get(cfg.eval.runner.type) == SlurmRunner:
+                cfg.eval.runner.partition = args.partition
+                cfg.eval.runner.quotatype = args.quotatype
+            else:
+                logger.warning('SlurmRunner is not used, so the partition '
+                               'argument is ignored.')
+        if args.debug:
+            cfg.eval.runner.debug = True
+        if args.lark:
+            cfg.eval.runner.lark_bot_url = cfg['lark_bot_url']
+        cfg.eval.partitioner['out_dir'] = osp.join(cfg['work_dir'], 'results/')
+        partitioner = PARTITIONERS.build(cfg.eval.partitioner)
+        tasks = partitioner(cfg)
+        if args.dry_run:
+            return
+        runner = RUNNERS.build(cfg.eval.runner)
+
+        # For meta-review-judge in subjective evaluation
+        if isinstance(tasks, list) and len(tasks) != 0 and isinstance(
+                tasks[0], list):
+            for task_part in tasks:
+                runner(task_part)
+        else:
+            runner(tasks)
+
+    # visualize
+    if args.mode in ['all', 'eval', 'viz']:
+        summarizer_cfg = cfg.get('summarizer', {})
+        if not summarizer_cfg or summarizer_cfg.get('type', None) is None:
+            summarizer_cfg['type'] = DefaultSummarizer
+        summarizer_cfg['config'] = cfg
+        summarizer = build_from_cfg(summarizer_cfg)
+        summarizer.summarize(time_str=cfg_time_str)
+
+
+if __name__ == '__main__':
+    main()

opencompass/datasets/NPHardEval/cmp_GCP_D.py

 import ast
 
-import networkx as nx
+try:
+    import networkx as nx
+except ImportError:
+    nx = None
+
 from datasets import Dataset
 
 from opencompass.openicl.icl_evaluator import BaseEvaluator

opencompass/datasets/NPHardEval/cmp_TSP_D.py

 import ast
 import json
 
-import networkx as nx
+try:
+    import networkx as nx
+except ImportError:
+    nx = None
+
 import pandas as pd
 from datasets import Dataset
 

opencompass/datasets/NPHardEval/p_SPP.py

 import ast
 import json
 
-import networkx as nx
+try:
+    import networkx as nx
+except ImportError:
+    nx = None
+
 from datasets import Dataset
 
 from opencompass.openicl.icl_evaluator import BaseEvaluator

opencompass/datasets/__init__.py

@@ -3,6 +3,7 @@ from .afqmcd import *  # noqa: F401, F403
 from .agieval import *  # noqa: F401, F403
 from .anli import AnliDataset  # noqa: F401, F403
 from .anthropics_evals import *  # noqa: F401, F403
+from .apps import *  # noqa: F401, F403
 from .arc import *  # noqa: F401, F403
 from .ax import *  # noqa: F401, F403
 from .bbh import *  # noqa: F401, F403
@@ -94,6 +95,7 @@ from .summedits import *  # noqa: F401, F403
 from .summscreen import *  # noqa: F401, F403
 from .svamp import *  # noqa: F401, F403
 from .tabmwp import *  # noqa: F401, F403
+from .taco import *  # noqa: F401, F403
 from .teval import *  # noqa: F401, F403
 from .TheoremQA import *  # noqa: F401, F403
 from .tnews import *  # noqa: F401, F403

opencompass/datasets/apps.py

@@ -19,13 +19,19 @@ from unittest.mock import mock_open, patch
 
 import numpy as np
 from datasets import Dataset, DatasetDict, load_dataset, load_from_disk
-from pyext import RuntimeModule
+
+try:
+    from pyext import RuntimeModule
+except ImportError:
+    RuntimeModule = None
 
 from opencompass.openicl.icl_evaluator import BaseEvaluator
 from opencompass.registry import ICL_EVALUATORS, LOAD_DATASET
+from opencompass.utils.logging import get_logger
 
 from .base import BaseDataset
 
+logger = get_logger()
 TIMEOUT = 10
 
 
@@ -67,18 +73,20 @@ class APPSDataset(BaseDataset):
             new_dataset[split] = Dataset.from_dict(new_data)
 
         # num_repeats duplicate
-        train_repeated = []
+        # train_repeated = []
         test_repeated = []
-        for sample in new_dataset['train']:
-            train_repeated.extend([sample] * num_repeats)
+        # for sample in new_dataset['train']:
+        #     train_repeated.extend([sample] * num_repeats)
         for sample in new_dataset['test']:
             test_repeated.extend([sample] * num_repeats)
 
-        dataset_train_repeated = new_dataset['train'].from_list(train_repeated)
+        # dataset_train_repeated = new_dataset['train'].from_list(
+        #    train_repeated
+        # )
         dataset_test_repeated = new_dataset['test'].from_list(test_repeated)
 
         return DatasetDict({
-            'train': dataset_train_repeated,
+            # 'train': dataset_train_repeated,
             'test': dataset_test_repeated
         })
 
@@ -121,18 +129,20 @@ class APPS_miniDataset(BaseDataset):
             new_dataset[split] = Dataset.from_dict(new_data)
 
         # num_repeats duplicate
-        train_repeated = []
+        # train_repeated = []
         test_repeated = []
-        for sample in new_dataset['train']:
-            train_repeated.extend([sample] * num_repeats)
+        # for sample in new_dataset['train']:
+        #     train_repeated.extend([sample] * num_repeats)
         for sample in new_dataset['test']:
             test_repeated.extend([sample] * num_repeats)
 
-        dataset_train_repeated = new_dataset['train'].from_list(train_repeated)
+        # dataset_train_repeated = new_dataset['train'].from_list(
+        #     train_repeated
+        # )
         dataset_test_repeated = new_dataset['test'].from_list(test_repeated)
 
         return DatasetDict({
-            'train': dataset_train_repeated,
+            # 'train': dataset_train_repeated,
             'test': dataset_test_repeated
         })
 
@@ -308,7 +318,10 @@ def timeout_handler(signum, frame):
     raise TimeoutException
 
 
-signal.signal(signal.SIGALRM, timeout_handler)
+try:
+    signal.signal(signal.SIGALRM, timeout_handler)
+except AttributeError:
+    logger.warning('signal.SIGALRM is not available on this platform')
 timeout = 4  # seconds
 
 

opencompass/datasets/custom.py

@@ -210,6 +210,8 @@ def make_mcq_gen_config(meta):
         input_columns=meta['input_columns'],
         output_column=meta['output_column'],
     )
+    if 'test_range' in meta:
+        reader_cfg['test_range'] = meta['test_range']
     infer_cfg = dict(
         prompt_template=dict(
             type=PromptTemplate,
@@ -255,6 +257,8 @@ def make_circular_mcq_gen_config(meta):
         input_columns=meta['input_columns'],
         output_column=meta['output_column'],
     )
+    if 'test_range' in meta:
+        reader_cfg['test_range'] = meta['test_range']
     infer_cfg = dict(
         prompt_template=dict(
             type=PromptTemplate,
@@ -304,6 +308,8 @@ def make_qa_gen_config(meta):
         input_columns=meta['input_columns'],
         output_column=meta['output_column'],
     )
+    if 'test_range' in meta:
+        reader_cfg['test_range'] = meta['test_range']
     infer_cfg = dict(
         prompt_template=dict(
             type=PromptTemplate,
@@ -353,6 +359,8 @@ def make_mcq_ppl_config(meta):
         input_columns=meta['input_columns'],
         output_column=meta['output_column'],
     )
+    if 'test_range' in meta:
+        reader_cfg['test_range'] = meta['test_range']
     infer_cfg = dict(
         prompt_template=dict(
             type=PromptTemplate,
@@ -399,6 +407,8 @@ def make_circular_mcq_ppl_config(meta):
         input_columns=meta['input_columns'],
         output_column=meta['output_column'],
     )
+    if 'test_range' in meta:
+        reader_cfg['test_range'] = meta['test_range']
     infer_cfg = dict(
         prompt_template=dict(
             type=PromptTemplate,

opencompass/datasets/humanevalx.py

@@ -168,9 +168,12 @@ def _clean_up_code(text: str, language_type: str, reference) -> str:
     """Cleans up the generated code."""
     try:
         # for chatGLM related text
-        text = eval(text)
+        eval_text = eval(text)
     except Exception:
         pass
+    else:
+        if isinstance(eval_text, str):
+            text = eval_text
     # extract code from code block
     text = text.lstrip('\n')
     if '```' in text:

opencompass/datasets/lawbench/evaluation_functions/ljp_article.py

 import re
-import cn2an
 
 """
 task: law article prediction
@@ -15,6 +14,7 @@ def compute_ljp_article(data_dict):
     A reference contains a list of articles of the Criminal Law of the People's Republic of China.
     We compute the F1-score between the prediction and the reference.
     """
+    import cn2an
 
     score_list, abstentions = [], 0
 

opencompass/datasets/lawbench/evaluation_functions/ljp_imprison.py

 import math
-import cn2an
+
 import re
 
 #法律判决预测-刑期预测
 def compute_ljp_imprison(data_dict):
+    import cn2an
+
     score_list, abstentions = [], 0
 
     for example in data_dict:

opencompass/datasets/math.py

@@ -85,6 +85,7 @@ def normalize_final_answer(final_answer: str) -> str:
 
     # Extract answer that is in LaTeX math, is bold,
     # is surrounded by a box, etc.
+    final_answer = re.sub(r'(\\text\{)\((.*?)\)(\})', '\\2', final_answer)
     final_answer = re.sub(r'(\\text\{)(.*?)(\})', '\\2', final_answer)
     final_answer = re.sub(r'(\\textbf\{)(.*?)(\})', '\\2', final_answer)
     final_answer = re.sub(r'(\\overline\{)(.*?)(\})', '\\2', final_answer)
@@ -178,10 +179,7 @@ class MATHEvaluator(BaseEvaluator):
 
     def score(self, predictions, references):
         if len(predictions) != len(references):
-            return {
-                'error': 'predictions and references have different '
-                'length'
-            }
+            return {'error': 'preds and refrs have different length'}
         correct = 0
         count = 0
         details = []
@@ -457,8 +455,23 @@ class MATHEvaluator(BaseEvaluator):
             ss2 = strip_string_func(str2)
             if verbose:
                 print(ss1, ss2)
-            return ss1 == ss2
+            if ss1 == ss2:
+                return True
+            ss1 = normalize_final_answer(ss1)
+            ss2 = normalize_final_answer(ss2)
+            if ss1 == ss2:
+                return True
+        except Exception:
+            pass
+
+        try:
+            ss1 = normalize_final_answer(str1)
+            ss2 = normalize_final_answer(str2)
+            if ss1 == ss2:
+                return True
         except Exception:
+            pass
+
         return str1 == str2
 
 

opencompass/datasets/mathbench.py

@@ -57,7 +57,7 @@ class MathBenchDataset(BaseDataset):
         """
         data = []
         filename = osp.join(path, f'{name}.jsonl')
-        with open(filename, 'r') as infile:
+        with open(filename, 'r', encoding='utf-8') as infile:
             for id, line in enumerate(infile):
                 entry = json.loads(line)
                 if 'cloze' in name:

opencompass/datasets/mbpp.py

@@ -244,6 +244,7 @@ class MBPPEvaluator(BaseEvaluator):
                 if not isinstance(preds, list):
                     preds = [preds]
                 for pred in preds:
+                    pred = self._process_answer(pred)
                     mbpp_preds.append({'task_id': refer, 'solution': pred})
             with tempfile.TemporaryDirectory() as tmp_dir:
                 out_dir = osp.join(tmp_dir, 'mbpp_eval.jsonl')

opencompass/datasets/taco.py

@@ -18,14 +18,20 @@ from io import StringIO
 from unittest.mock import mock_open, patch
 
 import numpy as np
-from datasets import Dataset, DatasetDict, load_dataset
-from pyext import RuntimeModule
+from datasets import Dataset, DatasetDict, load_from_disk
+
+try:
+    from pyext import RuntimeModule
+except ImportError:
+    RuntimeModule = None
 
 from opencompass.openicl.icl_evaluator import BaseEvaluator
 from opencompass.registry import ICL_EVALUATORS, LOAD_DATASET
+from opencompass.utils.logging import get_logger
 
 from .base import BaseDataset
 
+logger = get_logger()
 TIMEOUT = 10
 
 
@@ -34,7 +40,7 @@ class TACODataset(BaseDataset):
 
     @staticmethod
     def load(path: str, num_repeats: int = 1):
-        dataset = load_dataset(path)
+        dataset = load_from_disk(path)
         new_dataset = DatasetDict()
         # add new column "starter" in the prompt
         for split in dataset.keys():
@@ -69,18 +75,20 @@ class TACODataset(BaseDataset):
             new_dataset[split] = Dataset.from_dict(new_data)
 
         # num_repeats duplicate
-        train_repeated = []
+        # train_repeated = []
         test_repeated = []
-        for sample in new_dataset['train']:
-            train_repeated.extend([sample] * num_repeats)
+        # for sample in new_dataset['train']:
+        #     train_repeated.extend([sample] * num_repeats)
         for sample in new_dataset['test']:
             test_repeated.extend([sample] * num_repeats)
 
-        dataset_train_repeated = new_dataset['train'].from_list(train_repeated)
+        # dataset_train_repeated = new_dataset['train'].from_list(
+        #     train_repeated
+        # )
         dataset_test_repeated = new_dataset['test'].from_list(test_repeated)
 
         return DatasetDict({
-            'train': dataset_train_repeated,
+            # 'train': dataset_train_repeated,
             'test': dataset_test_repeated
         })
 
@@ -256,7 +264,10 @@ def timeout_handler(signum, frame):
     raise TimeoutException
 
 
-signal.signal(signal.SIGALRM, timeout_handler)
+try:
+    signal.signal(signal.SIGALRM, timeout_handler)
+except AttributeError:
+    logger.warning('signal.SIGALRM is not available on this platform')
 timeout = 4  # seconds