Upload results for model microsoft/phi-2

data/microsoft/phi-2/base/24-02-05-18:00:45.json

@@ -0,0 +1,1162 @@
+{
+  "results": {
+    "repellendus-laborum_lsat-rc_base": {
+      "acc,none": 0.26765799256505574,
+      "acc_stderr,none": 0.0270445453145873,
+      "alias": "repellendus-laborum_lsat-rc_base"
+    },
+    "repellendus-laborum_lsat-lr_base": {
+      "acc,none": 0.23137254901960785,
+      "acc_stderr,none": 0.018691965462419517,
+      "alias": "repellendus-laborum_lsat-lr_base"
+    },
+    "repellendus-laborum_lsat-ar_base": {
+      "acc,none": 0.25217391304347825,
+      "acc_stderr,none": 0.028696745294493366,
+      "alias": "repellendus-laborum_lsat-ar_base"
+    },
+    "repellendus-laborum_logiqa_base": {
+      "acc,none": 0.29073482428115016,
+      "acc_stderr,none": 0.0181640562091778,
+      "alias": "repellendus-laborum_logiqa_base"
+    },
+    "repellendus-laborum_logiqa2_base": {
+      "acc,none": 0.29389312977099236,
+      "acc_stderr,none": 0.011493223255677107,
+      "alias": "repellendus-laborum_logiqa2_base"
+    },
+    "possimus-voluptate_lsat-rc_base": {
+      "acc,none": 0.2527881040892193,
+      "acc_stderr,none": 0.026548061072649957,
+      "alias": "possimus-voluptate_lsat-rc_base"
+    },
+    "possimus-voluptate_lsat-lr_base": {
+      "acc,none": 0.22941176470588234,
+      "acc_stderr,none": 0.01863631913244453,
+      "alias": "possimus-voluptate_lsat-lr_base"
+    },
+    "possimus-voluptate_lsat-ar_base": {
+      "acc,none": 0.21739130434782608,
+      "acc_stderr,none": 0.02725685083881996,
+      "alias": "possimus-voluptate_lsat-ar_base"
+    },
+    "possimus-voluptate_logiqa_base": {
+      "acc,none": 0.2795527156549521,
+      "acc_stderr,none": 0.017951178003680606,
+      "alias": "possimus-voluptate_logiqa_base"
+    },
+    "possimus-voluptate_logiqa2_base": {
+      "acc,none": 0.2926208651399491,
+      "acc_stderr,none": 0.01147864633663911,
+      "alias": "possimus-voluptate_logiqa2_base"
+    },
+    "maxime-expedita_lsat-rc_base": {
+      "acc,none": 0.2527881040892193,
+      "acc_stderr,none": 0.026548061072649953,
+      "alias": "maxime-expedita_lsat-rc_base"
+    },
+    "maxime-expedita_lsat-lr_base": {
+      "acc,none": 0.24705882352941178,
+      "acc_stderr,none": 0.019117091440867724,
+      "alias": "maxime-expedita_lsat-lr_base"
+    },
+    "maxime-expedita_lsat-ar_base": {
+      "acc,none": 0.23478260869565218,
+      "acc_stderr,none": 0.028009647070930132,
+      "alias": "maxime-expedita_lsat-ar_base"
+    },
+    "maxime-expedita_logiqa_base": {
+      "acc,none": 0.2971246006389776,
+      "acc_stderr,none": 0.018279674935144995,
+      "alias": "maxime-expedita_logiqa_base"
+    },
+    "maxime-expedita_logiqa2_base": {
+      "acc,none": 0.27099236641221375,
+      "acc_stderr,none": 0.011213894711527516,
+      "alias": "maxime-expedita_logiqa2_base"
+    },
+    "eveniet-ea_lsat-rc_base": {
+      "acc,none": 0.2899628252788104,
+      "acc_stderr,none": 0.027716877855226897,
+      "alias": "eveniet-ea_lsat-rc_base"
+    },
+    "eveniet-ea_lsat-lr_base": {
+      "acc,none": 0.23333333333333334,
+      "acc_stderr,none": 0.01874704371659074,
+      "alias": "eveniet-ea_lsat-lr_base"
+    },
+    "eveniet-ea_lsat-ar_base": {
+      "acc,none": 0.2217391304347826,
+      "acc_stderr,none": 0.02745149660405891,
+      "alias": "eveniet-ea_lsat-ar_base"
+    },
+    "eveniet-ea_logiqa_base": {
+      "acc,none": 0.2955271565495208,
+      "acc_stderr,none": 0.018251174484565112,
+      "alias": "eveniet-ea_logiqa_base"
+    },
+    "eveniet-ea_logiqa2_base": {
+      "acc,none": 0.2837150127226463,
+      "acc_stderr,none": 0.011373548669758796,
+      "alias": "eveniet-ea_logiqa2_base"
+    },
+    "distinctio-unde_lsat-rc_base": {
+      "acc,none": 0.26394052044609667,
+      "acc_stderr,none": 0.026924155643902548,
+      "alias": "distinctio-unde_lsat-rc_base"
+    },
+    "distinctio-unde_lsat-lr_base": {
+      "acc,none": 0.2647058823529412,
+      "acc_stderr,none": 0.019554803257850088,
+      "alias": "distinctio-unde_lsat-lr_base"
+    },
+    "distinctio-unde_lsat-ar_base": {
+      "acc,none": 0.24347826086956523,
+      "acc_stderr,none": 0.028361099300075063,
+      "alias": "distinctio-unde_lsat-ar_base"
+    },
+    "distinctio-unde_logiqa_base": {
+      "acc,none": 0.3003194888178914,
+      "acc_stderr,none": 0.01833587493212361,
+      "alias": "distinctio-unde_logiqa_base"
+    },
+    "distinctio-unde_logiqa2_base": {
+      "acc,none": 0.2970737913486005,
+      "acc_stderr,none": 0.011529193947365896,
+      "alias": "distinctio-unde_logiqa2_base"
+    },
+    "aspernatur-sint_lsat-rc_base": {
+      "acc,none": 0.26765799256505574,
+      "acc_stderr,none": 0.027044545314587293,
+      "alias": "aspernatur-sint_lsat-rc_base"
+    },
+    "aspernatur-sint_lsat-lr_base": {
+      "acc,none": 0.27058823529411763,
+      "acc_stderr,none": 0.0196916426487322,
+      "alias": "aspernatur-sint_lsat-lr_base"
+    },
+    "aspernatur-sint_lsat-ar_base": {
+      "acc,none": 0.2217391304347826,
+      "acc_stderr,none": 0.027451496604058913,
+      "alias": "aspernatur-sint_lsat-ar_base"
+    },
+    "aspernatur-sint_logiqa_base": {
+      "acc,none": 0.2955271565495208,
+      "acc_stderr,none": 0.018251174484565112,
+      "alias": "aspernatur-sint_logiqa_base"
+    },
+    "aspernatur-sint_logiqa2_base": {
+      "acc,none": 0.2868956743002545,
+      "acc_stderr,none": 0.01141170254782954,
+      "alias": "aspernatur-sint_logiqa2_base"
+    }
+  },
+  "configs": {
+    "aspernatur-sint_logiqa2_base": {
+      "task": "aspernatur-sint_logiqa2_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "aspernatur-sint-logiqa2/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "aspernatur-sint_logiqa_base": {
+      "task": "aspernatur-sint_logiqa_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "aspernatur-sint-logiqa/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "aspernatur-sint_lsat-ar_base": {
+      "task": "aspernatur-sint_lsat-ar_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "aspernatur-sint-lsat-ar/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "aspernatur-sint_lsat-lr_base": {
+      "task": "aspernatur-sint_lsat-lr_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "aspernatur-sint-lsat-lr/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "aspernatur-sint_lsat-rc_base": {
+      "task": "aspernatur-sint_lsat-rc_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "aspernatur-sint-lsat-rc/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "distinctio-unde_logiqa2_base": {
+      "task": "distinctio-unde_logiqa2_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "distinctio-unde-logiqa2/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "distinctio-unde_logiqa_base": {
+      "task": "distinctio-unde_logiqa_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "distinctio-unde-logiqa/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "distinctio-unde_lsat-ar_base": {
+      "task": "distinctio-unde_lsat-ar_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "distinctio-unde-lsat-ar/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "distinctio-unde_lsat-lr_base": {
+      "task": "distinctio-unde_lsat-lr_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "distinctio-unde-lsat-lr/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "distinctio-unde_lsat-rc_base": {
+      "task": "distinctio-unde_lsat-rc_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "distinctio-unde-lsat-rc/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "eveniet-ea_logiqa2_base": {
+      "task": "eveniet-ea_logiqa2_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "eveniet-ea-logiqa2/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "eveniet-ea_logiqa_base": {
+      "task": "eveniet-ea_logiqa_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "eveniet-ea-logiqa/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "eveniet-ea_lsat-ar_base": {
+      "task": "eveniet-ea_lsat-ar_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "eveniet-ea-lsat-ar/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "eveniet-ea_lsat-lr_base": {
+      "task": "eveniet-ea_lsat-lr_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "eveniet-ea-lsat-lr/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "eveniet-ea_lsat-rc_base": {
+      "task": "eveniet-ea_lsat-rc_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "eveniet-ea-lsat-rc/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "maxime-expedita_logiqa2_base": {
+      "task": "maxime-expedita_logiqa2_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "maxime-expedita-logiqa2/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "maxime-expedita_logiqa_base": {
+      "task": "maxime-expedita_logiqa_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "maxime-expedita-logiqa/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "maxime-expedita_lsat-ar_base": {
+      "task": "maxime-expedita_lsat-ar_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "maxime-expedita-lsat-ar/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "maxime-expedita_lsat-lr_base": {
+      "task": "maxime-expedita_lsat-lr_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "maxime-expedita-lsat-lr/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "maxime-expedita_lsat-rc_base": {
+      "task": "maxime-expedita_lsat-rc_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "maxime-expedita-lsat-rc/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "possimus-voluptate_logiqa2_base": {
+      "task": "possimus-voluptate_logiqa2_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "possimus-voluptate-logiqa2/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "possimus-voluptate_logiqa_base": {
+      "task": "possimus-voluptate_logiqa_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "possimus-voluptate-logiqa/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "possimus-voluptate_lsat-ar_base": {
+      "task": "possimus-voluptate_lsat-ar_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "possimus-voluptate-lsat-ar/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "possimus-voluptate_lsat-lr_base": {
+      "task": "possimus-voluptate_lsat-lr_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "possimus-voluptate-lsat-lr/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "possimus-voluptate_lsat-rc_base": {
+      "task": "possimus-voluptate_lsat-rc_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "possimus-voluptate-lsat-rc/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "repellendus-laborum_logiqa2_base": {
+      "task": "repellendus-laborum_logiqa2_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "repellendus-laborum-logiqa2/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "repellendus-laborum_logiqa_base": {
+      "task": "repellendus-laborum_logiqa_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "repellendus-laborum-logiqa/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "repellendus-laborum_lsat-ar_base": {
+      "task": "repellendus-laborum_lsat-ar_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "repellendus-laborum-lsat-ar/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "repellendus-laborum_lsat-lr_base": {
+      "task": "repellendus-laborum_lsat-lr_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "repellendus-laborum-lsat-lr/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "repellendus-laborum_lsat-rc_base": {
+      "task": "repellendus-laborum_lsat-rc_base",
+      "group": "logikon-bench",
+      "dataset_path": "logikon/cot-eval-traces",
+      "dataset_kwargs": {
+        "data_files": {
+          "test": "repellendus-laborum-lsat-rc/test-00000-of-00001.parquet"
+        }
+      },
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc) -> str:\n    \"\"\"\n    Answer the following question about the given passage.\n    \n    Passage: <passage>\n    \n    Question: <question>\n    A. <choice1>\n    B. <choice2>\n    C. <choice3>\n    D. <choice4>\n    [E. <choice5>]\n        \n    Answer:\n    \"\"\"\n    k = len(doc[\"options\"])\n    choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n    prompt = \"Answer the following question about the given passage.\\n\\n\"\n    prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n    prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n    for choice, option in zip(choices, doc[\"options\"]):\n        prompt += f\"{choice.upper()}. {option}\\n\"\n    prompt += \"\\n\"\n    prompt += \"Answer:\"\n    return prompt\n",
+      "doc_to_target": "{{answer}}",
+      "doc_to_choice": "{{options}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    }
+  },
+  "versions": {
+    "aspernatur-sint_logiqa2_base": 0.0,
+    "aspernatur-sint_logiqa_base": 0.0,
+    "aspernatur-sint_lsat-ar_base": 0.0,
+    "aspernatur-sint_lsat-lr_base": 0.0,
+    "aspernatur-sint_lsat-rc_base": 0.0,
+    "distinctio-unde_logiqa2_base": 0.0,
+    "distinctio-unde_logiqa_base": 0.0,
+    "distinctio-unde_lsat-ar_base": 0.0,
+    "distinctio-unde_lsat-lr_base": 0.0,
+    "distinctio-unde_lsat-rc_base": 0.0,
+    "eveniet-ea_logiqa2_base": 0.0,
+    "eveniet-ea_logiqa_base": 0.0,
+    "eveniet-ea_lsat-ar_base": 0.0,
+    "eveniet-ea_lsat-lr_base": 0.0,
+    "eveniet-ea_lsat-rc_base": 0.0,
+    "maxime-expedita_logiqa2_base": 0.0,
+    "maxime-expedita_logiqa_base": 0.0,
+    "maxime-expedita_lsat-ar_base": 0.0,
+    "maxime-expedita_lsat-lr_base": 0.0,
+    "maxime-expedita_lsat-rc_base": 0.0,
+    "possimus-voluptate_logiqa2_base": 0.0,
+    "possimus-voluptate_logiqa_base": 0.0,
+    "possimus-voluptate_lsat-ar_base": 0.0,
+    "possimus-voluptate_lsat-lr_base": 0.0,
+    "possimus-voluptate_lsat-rc_base": 0.0,
+    "repellendus-laborum_logiqa2_base": 0.0,
+    "repellendus-laborum_logiqa_base": 0.0,
+    "repellendus-laborum_lsat-ar_base": 0.0,
+    "repellendus-laborum_lsat-lr_base": 0.0,
+    "repellendus-laborum_lsat-rc_base": 0.0
+  },
+  "n-shot": {
+    "aspernatur-sint_logiqa2_base": 0,
+    "aspernatur-sint_logiqa_base": 0,
+    "aspernatur-sint_lsat-ar_base": 0,
+    "aspernatur-sint_lsat-lr_base": 0,
+    "aspernatur-sint_lsat-rc_base": 0,
+    "distinctio-unde_logiqa2_base": 0,
+    "distinctio-unde_logiqa_base": 0,
+    "distinctio-unde_lsat-ar_base": 0,
+    "distinctio-unde_lsat-lr_base": 0,
+    "distinctio-unde_lsat-rc_base": 0,
+    "eveniet-ea_logiqa2_base": 0,
+    "eveniet-ea_logiqa_base": 0,
+    "eveniet-ea_lsat-ar_base": 0,
+    "eveniet-ea_lsat-lr_base": 0,
+    "eveniet-ea_lsat-rc_base": 0,
+    "maxime-expedita_logiqa2_base": 0,
+    "maxime-expedita_logiqa_base": 0,
+    "maxime-expedita_lsat-ar_base": 0,
+    "maxime-expedita_lsat-lr_base": 0,
+    "maxime-expedita_lsat-rc_base": 0,
+    "possimus-voluptate_logiqa2_base": 0,
+    "possimus-voluptate_logiqa_base": 0,
+    "possimus-voluptate_lsat-ar_base": 0,
+    "possimus-voluptate_lsat-lr_base": 0,
+    "possimus-voluptate_lsat-rc_base": 0,
+    "repellendus-laborum_logiqa2_base": 0,
+    "repellendus-laborum_logiqa_base": 0,
+    "repellendus-laborum_lsat-ar_base": 0,
+    "repellendus-laborum_lsat-lr_base": 0,
+    "repellendus-laborum_lsat-rc_base": 0
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": "pretrained=microsoft/phi-2,revision=main,dtype=auto,tensor_parallel_size=1,gpu_memory_utilization=0.9,trust_remote_code=true,max_length=2048",
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": null,
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": null
+  },
+  "git_hash": "3d5b980"
+}