facebook old models result

lm-eval-output/facebook/opt-6.7b/lambada_multilingual/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/result-jsonl.tar.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e72a14d8844198e15463336ca4652d645a880a5ff2568a18d5bd8e33177c7796
+size 5205711

lm-eval-output/facebook/opt-6.7b/lambada_multilingual/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/results.json

@@ -0,0 +1,252 @@
+{
+  "results": {
+    "lambada_multilingual": {
+      "perplexity,none": 94.24299539565746,
+      "perplexity_stderr,none": 30.028815681145435,
+      "acc,none": 0.3881622355909179,
+      "acc_stderr,none": 0.08375300012280831,
+      "alias": "lambada_multilingual"
+    },
+    "lambada_openai_mt_de": {
+      "perplexity,none": 140.0133340076818,
+      "perplexity_stderr,none": 8.652865280352644,
+      "acc,none": 0.28041917329710847,
+      "acc_stderr,none": 0.006258290472632361,
+      "alias": " - lambada_openai_mt_de"
+    },
+    "lambada_openai_mt_en": {
+      "perplexity,none": 4.25205289918128,
+      "perplexity_stderr,none": 0.09260941569288471,
+      "acc,none": 0.6741703861828061,
+      "acc_stderr,none": 0.006529684317476097,
+      "alias": " - lambada_openai_mt_en"
+    },
+    "lambada_openai_mt_es": {
+      "perplexity,none": 124.4520996487743,
+      "perplexity_stderr,none": 7.245651841086687,
+      "acc,none": 0.3060353192315156,
+      "acc_stderr,none": 0.006420465728110188,
+      "alias": " - lambada_openai_mt_es"
+    },
+    "lambada_openai_mt_fr": {
+      "perplexity,none": 76.7582261318957,
+      "perplexity_stderr,none": 4.452752530346086,
+      "acc,none": 0.36095478362119154,
+      "acc_stderr,none": 0.006691203954844635,
+      "alias": " - lambada_openai_mt_fr"
+    },
+    "lambada_openai_mt_it": {
+      "perplexity,none": 125.73926429075424,
+      "perplexity_stderr,none": 7.919251868167341,
+      "acc,none": 0.3192315156219678,
+      "acc_stderr,none": 0.0064947834277386774,
+      "alias": " - lambada_openai_mt_it"
+    }
+  },
+  "groups": {
+    "lambada_multilingual": {
+      "perplexity,none": 94.24299539565746,
+      "perplexity_stderr,none": 30.028815681145435,
+      "acc,none": 0.3881622355909179,
+      "acc_stderr,none": 0.08375300012280831,
+      "alias": "lambada_multilingual"
+    }
+  },
+  "configs": {
+    "lambada_openai_mt_de": {
+      "task": "lambada_openai_mt_de",
+      "group": [
+        "lambada_multilingual"
+      ],
+      "dataset_path": "EleutherAI/lambada_openai",
+      "dataset_name": "de",
+      "test_split": "test",
+      "doc_to_text": "{{text.split(' ')[:-1]|join(' ')}}",
+      "doc_to_target": "{{' '+text.split(' ')[-1]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "perplexity",
+          "aggregation": "perplexity",
+          "higher_is_better": false
+        },
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "loglikelihood",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{text}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "lambada_openai_mt_en": {
+      "task": "lambada_openai_mt_en",
+      "group": [
+        "lambada_multilingual"
+      ],
+      "dataset_path": "EleutherAI/lambada_openai",
+      "dataset_name": "en",
+      "test_split": "test",
+      "doc_to_text": "{{text.split(' ')[:-1]|join(' ')}}",
+      "doc_to_target": "{{' '+text.split(' ')[-1]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "perplexity",
+          "aggregation": "perplexity",
+          "higher_is_better": false
+        },
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "loglikelihood",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{text}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "lambada_openai_mt_es": {
+      "task": "lambada_openai_mt_es",
+      "group": [
+        "lambada_multilingual"
+      ],
+      "dataset_path": "EleutherAI/lambada_openai",
+      "dataset_name": "es",
+      "test_split": "test",
+      "doc_to_text": "{{text.split(' ')[:-1]|join(' ')}}",
+      "doc_to_target": "{{' '+text.split(' ')[-1]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "perplexity",
+          "aggregation": "perplexity",
+          "higher_is_better": false
+        },
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "loglikelihood",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{text}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "lambada_openai_mt_fr": {
+      "task": "lambada_openai_mt_fr",
+      "group": [
+        "lambada_multilingual"
+      ],
+      "dataset_path": "EleutherAI/lambada_openai",
+      "dataset_name": "fr",
+      "test_split": "test",
+      "doc_to_text": "{{text.split(' ')[:-1]|join(' ')}}",
+      "doc_to_target": "{{' '+text.split(' ')[-1]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "perplexity",
+          "aggregation": "perplexity",
+          "higher_is_better": false
+        },
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "loglikelihood",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{text}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "lambada_openai_mt_it": {
+      "task": "lambada_openai_mt_it",
+      "group": [
+        "lambada_multilingual"
+      ],
+      "dataset_path": "EleutherAI/lambada_openai",
+      "dataset_name": "it",
+      "test_split": "test",
+      "doc_to_text": "{{text.split(' ')[:-1]|join(' ')}}",
+      "doc_to_target": "{{' '+text.split(' ')[-1]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "perplexity",
+          "aggregation": "perplexity",
+          "higher_is_better": false
+        },
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "loglikelihood",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{text}}",
+      "metadata": {
+        "version": 1.0
+      }
+    }
+  },
+  "versions": {
+    "lambada_multilingual": "N/A",
+    "lambada_openai_mt_de": 1.0,
+    "lambada_openai_mt_en": 1.0,
+    "lambada_openai_mt_es": 1.0,
+    "lambada_openai_mt_fr": 1.0,
+    "lambada_openai_mt_it": 1.0
+  },
+  "n-shot": {
+    "lambada_multilingual": 0,
+    "lambada_openai_mt_de": 0,
+    "lambada_openai_mt_en": 0,
+    "lambada_openai_mt_es": 0,
+    "lambada_openai_mt_fr": 0,
+    "lambada_openai_mt_it": 0
+  },
+  "config": {
+    "model": "hf",
+    "model_args": "pretrained=facebook/opt-6.7b,dtype=bfloat16,trust_remote_code=True",
+    "batch_size": "auto",
+    "batch_sizes": [
+      32
+    ],
+    "device": null,
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": null
+  },
+  "git_hash": "2c0a875"
+}

lm-eval-output/facebook/opt-6.7b/lambada_multilingual/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/taskrun.log

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c8ff9006ae067f27909e498f559b22c853379627ddc7df5df9d80011e73b529
+size 46328

lm-eval-output/facebook/opt-6.7b/pawsx/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/result-jsonl.tar.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d27c7f7fb81487d7254ffc5d4c06837d1e7853f9d41789a653da4192f567bb2e
+size 2134656

lm-eval-output/facebook/opt-6.7b/pawsx/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/results.json

@@ -0,0 +1,283 @@
+{
+  "results": {
+    "pawsx": {
+      "acc,none": 0.4877142857142857,
+      "acc_stderr,none": 0.049786073170358214,
+      "alias": "pawsx"
+    },
+    "paws_de": {
+      "acc,none": 0.417,
+      "acc_stderr,none": 0.011027978425535509,
+      "alias": " - paws_de"
+    },
+    "paws_en": {
+      "acc,none": 0.382,
+      "acc_stderr,none": 0.010867246593514927,
+      "alias": " - paws_en"
+    },
+    "paws_es": {
+      "acc,none": 0.4455,
+      "acc_stderr,none": 0.011116504096687393,
+      "alias": " - paws_es"
+    },
+    "paws_fr": {
+      "acc,none": 0.522,
+      "acc_stderr,none": 0.011172305500884877,
+      "alias": " - paws_fr"
+    },
+    "paws_ja": {
+      "acc,none": 0.558,
+      "acc_stderr,none": 0.011107641056719627,
+      "alias": " - paws_ja"
+    },
+    "paws_ko": {
+      "acc,none": 0.551,
+      "acc_stderr,none": 0.011124809242874423,
+      "alias": " - paws_ko"
+    },
+    "paws_zh": {
+      "acc,none": 0.5385,
+      "acc_stderr,none": 0.011149934327957058,
+      "alias": " - paws_zh"
+    }
+  },
+  "groups": {
+    "pawsx": {
+      "acc,none": 0.4877142857142857,
+      "acc_stderr,none": 0.049786073170358214,
+      "alias": "pawsx"
+    }
+  },
+  "configs": {
+    "paws_de": {
+      "task": "paws_de",
+      "group": "pawsx",
+      "dataset_path": "paws-x",
+      "dataset_name": "de",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[sentence1+\", richtig? Ja, \"+sentence2, sentence1+\", richtig? Nein, \"+sentence2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "paws_en": {
+      "task": "paws_en",
+      "group": "pawsx",
+      "dataset_path": "paws-x",
+      "dataset_name": "en",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[sentence1+\", right? Yes, \"+sentence2, sentence1+\", right? No, \"+sentence2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "paws_es": {
+      "task": "paws_es",
+      "group": "pawsx",
+      "dataset_path": "paws-x",
+      "dataset_name": "es",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[sentence1+\", verdad? Sí, \"+sentence2, sentence1+\", verdad? No, \"+sentence2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "paws_fr": {
+      "task": "paws_fr",
+      "group": "pawsx",
+      "dataset_path": "paws-x",
+      "dataset_name": "fr",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[sentence1+\", n'est-ce pas? Oui, \"+sentence2, sentence1+\", n'est-ce pas? No, \"+sentence2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "paws_ja": {
+      "task": "paws_ja",
+      "group": "pawsx",
+      "dataset_path": "paws-x",
+      "dataset_name": "ja",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[sentence1+\", ですね? はい, \"+sentence2, sentence1+\", ですね? いいえ, \"+sentence2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "paws_ko": {
+      "task": "paws_ko",
+      "group": "pawsx",
+      "dataset_path": "paws-x",
+      "dataset_name": "ko",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[sentence1+\", 맞죠? 예, \"+sentence2, sentence1+\", 맞죠? 아니요, \"+sentence2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    },
+    "paws_zh": {
+      "task": "paws_zh",
+      "group": "pawsx",
+      "dataset_path": "paws-x",
+      "dataset_name": "zh",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[sentence1+\", 对吧? 是, \"+sentence2, sentence1+\", 对吧? 不是, \"+sentence2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 0.0
+      }
+    }
+  },
+  "versions": {
+    "paws_de": 0.0,
+    "paws_en": 0.0,
+    "paws_es": 0.0,
+    "paws_fr": 0.0,
+    "paws_ja": 0.0,
+    "paws_ko": 0.0,
+    "paws_zh": 0.0,
+    "pawsx": "N/A"
+  },
+  "n-shot": {
+    "paws_de": 0,
+    "paws_en": 0,
+    "paws_es": 0,
+    "paws_fr": 0,
+    "paws_ja": 0,
+    "paws_ko": 0,
+    "paws_zh": 0,
+    "pawsx": 0
+  },
+  "config": {
+    "model": "hf",
+    "model_args": "pretrained=facebook/opt-6.7b,dtype=bfloat16,trust_remote_code=True",
+    "batch_size": "auto",
+    "batch_sizes": [
+      16
+    ],
+    "device": null,
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": null
+  },
+  "git_hash": "2c0a875"
+}

lm-eval-output/facebook/opt-6.7b/pawsx/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/taskrun.log

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6e2dc8b0566562b770dac9f7d69ac61eb63b1655ce4db1df89bd9b7cd930fcf
+size 31153

lm-eval-output/facebook/opt-6.7b/xcopa/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/result-jsonl.tar.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1708e0657980a8a37f0fce08798949b10853b8b9d7be5ac1309630f7160a5b24
+size 532183

lm-eval-output/facebook/opt-6.7b/xcopa/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/results.json

@@ -0,0 +1,390 @@
+{
+  "results": {
+    "xcopa": {
+      "acc,none": 0.5229090909090909,
+      "acc_stderr,none": 0.027215770707865124,
+      "alias": "xcopa"
+    },
+    "xcopa_et": {
+      "acc,none": 0.496,
+      "acc_stderr,none": 0.02238235778196214,
+      "alias": " - xcopa_et"
+    },
+    "xcopa_ht": {
+      "acc,none": 0.51,
+      "acc_stderr,none": 0.022378596989230785,
+      "alias": " - xcopa_ht"
+    },
+    "xcopa_id": {
+      "acc,none": 0.528,
+      "acc_stderr,none": 0.022347949832668086,
+      "alias": " - xcopa_id"
+    },
+    "xcopa_it": {
+      "acc,none": 0.544,
+      "acc_stderr,none": 0.022296238348407053,
+      "alias": " - xcopa_it"
+    },
+    "xcopa_qu": {
+      "acc,none": 0.506,
+      "acc_stderr,none": 0.022381462412439324,
+      "alias": " - xcopa_qu"
+    },
+    "xcopa_sw": {
+      "acc,none": 0.526,
+      "acc_stderr,none": 0.02235279165091416,
+      "alias": " - xcopa_sw"
+    },
+    "xcopa_ta": {
+      "acc,none": 0.552,
+      "acc_stderr,none": 0.022261697292270132,
+      "alias": " - xcopa_ta"
+    },
+    "xcopa_th": {
+      "acc,none": 0.544,
+      "acc_stderr,none": 0.022296238348407046,
+      "alias": " - xcopa_th"
+    },
+    "xcopa_tr": {
+      "acc,none": 0.532,
+      "acc_stderr,none": 0.022337186479044296,
+      "alias": " - xcopa_tr"
+    },
+    "xcopa_vi": {
+      "acc,none": 0.524,
+      "acc_stderr,none": 0.022357273881016403,
+      "alias": " - xcopa_vi"
+    },
+    "xcopa_zh": {
+      "acc,none": 0.49,
+      "acc_stderr,none": 0.02237859698923078,
+      "alias": " - xcopa_zh"
+    }
+  },
+  "groups": {
+    "xcopa": {
+      "acc,none": 0.5229090909090909,
+      "acc_stderr,none": 0.027215770707865124,
+      "alias": "xcopa"
+    }
+  },
+  "configs": {
+    "xcopa_et": {
+      "task": "xcopa_et",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "et",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e199f6a0>, connector={'cause': 'sest', 'effect': 'seetõttu'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_ht": {
+      "task": "xcopa_ht",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "ht",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e1ab3600>, connector={'cause': 'poukisa', 'effect': 'donk sa'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_id": {
+      "task": "xcopa_id",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "id",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e199e020>, connector={'cause': 'karena', 'effect': 'maka'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_it": {
+      "task": "xcopa_it",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "it",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e1ab2480>, connector={'cause': 'perché', 'effect': 'quindi'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_qu": {
+      "task": "xcopa_qu",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "qu",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e3968c20>, connector={'cause': 'imataq', 'effect': 'chaymi'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_sw": {
+      "task": "xcopa_sw",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "sw",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e1ab2340>, connector={'cause': 'kwa sababu', 'effect': 'kwa hiyo'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_ta": {
+      "task": "xcopa_ta",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "ta",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e1ab3100>, connector={'cause': 'காரணமாக', 'effect': 'எனவே'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_th": {
+      "task": "xcopa_th",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "th",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e199d300>, connector={'cause': 'เพราะ', 'effect': 'ดังนั้น'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_tr": {
+      "task": "xcopa_tr",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "tr",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e3968900>, connector={'cause': 'çünkü', 'effect': 'bu yüzden'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_vi": {
+      "task": "xcopa_vi",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "vi",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e196aac0>, connector={'cause': 'bởi vì', 'effect': 'vì vậy'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xcopa_zh": {
+      "task": "xcopa_zh",
+      "group": "xcopa",
+      "dataset_path": "xcopa",
+      "dataset_name": "zh",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "functools.partial(<function doc_to_text at 0x7f21e3968040>, connector={'cause': '因为', 'effect': '所以'})",
+      "doc_to_target": "label",
+      "doc_to_choice": "def doc_to_choice(doc):\n    return [convert_choice(doc[\"choice1\"]), convert_choice(doc[\"choice2\"])]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc"
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    }
+  },
+  "versions": {
+    "xcopa": "N/A",
+    "xcopa_et": 1.0,
+    "xcopa_ht": 1.0,
+    "xcopa_id": 1.0,
+    "xcopa_it": 1.0,
+    "xcopa_qu": 1.0,
+    "xcopa_sw": 1.0,
+    "xcopa_ta": 1.0,
+    "xcopa_th": 1.0,
+    "xcopa_tr": 1.0,
+    "xcopa_vi": 1.0,
+    "xcopa_zh": 1.0
+  },
+  "n-shot": {
+    "xcopa": 0,
+    "xcopa_et": 0,
+    "xcopa_ht": 0,
+    "xcopa_id": 0,
+    "xcopa_it": 0,
+    "xcopa_qu": 0,
+    "xcopa_sw": 0,
+    "xcopa_ta": 0,
+    "xcopa_th": 0,
+    "xcopa_tr": 0,
+    "xcopa_vi": 0,
+    "xcopa_zh": 0
+  },
+  "config": {
+    "model": "hf",
+    "model_args": "pretrained=facebook/opt-6.7b,dtype=bfloat16,trust_remote_code=True",
+    "batch_size": "auto",
+    "batch_sizes": [
+      16
+    ],
+    "device": null,
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": null
+  },
+  "git_hash": "2c0a875"
+}

lm-eval-output/facebook/opt-6.7b/xcopa/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/taskrun.log

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4dab5e1674c21385daada89fc15c75ecaf6ce5d6bbc474d5faf2a6c7da2ccf62
+size 50306

lm-eval-output/facebook/opt-6.7b/xnli/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/result-jsonl.tar.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf564cf1ecfc9d77543610daa5c3e91873a22dd624e32b3d6e269a73ea8aeb51
+size 6013120

lm-eval-output/facebook/opt-6.7b/xnli/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/results.json

@@ -0,0 +1,548 @@
+{
+  "results": {
+    "xnli": {
+      "acc,none": 0.3784738955823293,
+      "acc_stderr,none": 0.054985199168171614,
+      "alias": "xnli"
+    },
+    "xnli_ar": {
+      "acc,none": 0.3373493975903614,
+      "acc_stderr,none": 0.009476976849778591,
+      "alias": " - xnli_ar"
+    },
+    "xnli_bg": {
+      "acc,none": 0.3550200803212851,
+      "acc_stderr,none": 0.009591512730974288,
+      "alias": " - xnli_bg"
+    },
+    "xnli_de": {
+      "acc,none": 0.42811244979919677,
+      "acc_stderr,none": 0.009917947421067469,
+      "alias": " - xnli_de"
+    },
+    "xnli_el": {
+      "acc,none": 0.3345381526104418,
+      "acc_stderr,none": 0.009457404390939166,
+      "alias": " - xnli_el"
+    },
+    "xnli_en": {
+      "acc,none": 0.5602409638554217,
+      "acc_stderr,none": 0.009949067285169363,
+      "alias": " - xnli_en"
+    },
+    "xnli_es": {
+      "acc,none": 0.4493975903614458,
+      "acc_stderr,none": 0.009970615649588137,
+      "alias": " - xnli_es"
+    },
+    "xnli_fr": {
+      "acc,none": 0.4506024096385542,
+      "acc_stderr,none": 0.009973042774811681,
+      "alias": " - xnli_fr"
+    },
+    "xnli_hi": {
+      "acc,none": 0.3417670682730924,
+      "acc_stderr,none": 0.009506977398287632,
+      "alias": " - xnli_hi"
+    },
+    "xnli_ru": {
+      "acc,none": 0.37670682730923694,
+      "acc_stderr,none": 0.009712599529552994,
+      "alias": " - xnli_ru"
+    },
+    "xnli_sw": {
+      "acc,none": 0.3385542168674699,
+      "acc_stderr,none": 0.009485250208516887,
+      "alias": " - xnli_sw"
+    },
+    "xnli_th": {
+      "acc,none": 0.3317269076305221,
+      "acc_stderr,none": 0.00943745490032912,
+      "alias": " - xnli_th"
+    },
+    "xnli_tr": {
+      "acc,none": 0.3610441767068273,
+      "acc_stderr,none": 0.009627269742195714,
+      "alias": " - xnli_tr"
+    },
+    "xnli_ur": {
+      "acc,none": 0.3365461847389558,
+      "acc_stderr,none": 0.009471423054177131,
+      "alias": " - xnli_ur"
+    },
+    "xnli_vi": {
+      "acc,none": 0.3449799196787149,
+      "acc_stderr,none": 0.009528219800053311,
+      "alias": " - xnli_vi"
+    },
+    "xnli_zh": {
+      "acc,none": 0.3305220883534137,
+      "acc_stderr,none": 0.00942878910928982,
+      "alias": " - xnli_zh"
+    }
+  },
+  "groups": {
+    "xnli": {
+      "acc,none": 0.3784738955823293,
+      "acc_stderr,none": 0.054985199168171614,
+      "alias": "xnli"
+    }
+  },
+  "configs": {
+    "xnli_ar": {
+      "task": "xnli_ar",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "ar",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", صحيح? نعم, \"+hypothesis,premise+\", صحيح? لذا, \"+hypothesis,premise+\", صحيح? رقم, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_bg": {
+      "task": "xnli_bg",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "bg",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", правилно? да, \"+hypothesis,premise+\", правилно? така, \"+hypothesis,premise+\", правилно? не, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_de": {
+      "task": "xnli_de",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "de",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", richtig? Ja, \"+hypothesis,premise+\", richtig? Auch, \"+hypothesis,premise+\", richtig? Nein, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_el": {
+      "task": "xnli_el",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "el",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", σωστός? Ναί, \"+hypothesis,premise+\", σωστός? Έτσι, \"+hypothesis,premise+\", σωστός? όχι, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_en": {
+      "task": "xnli_en",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "en",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", right? Yes, \"+hypothesis,premise+\", right? Also, \"+hypothesis,premise+\", right? No, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_es": {
+      "task": "xnli_es",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "es",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", correcto? Sí, \"+hypothesis,premise+\", correcto? Asi que, \"+hypothesis,premise+\", correcto? No, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_fr": {
+      "task": "xnli_fr",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "fr",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", correct? Oui, \"+hypothesis,premise+\", correct? Aussi, \"+hypothesis,premise+\", correct? Non, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_hi": {
+      "task": "xnli_hi",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "hi",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", सही? हाँ, \"+hypothesis,premise+\", सही? इसलिए, \"+hypothesis,premise+\", सही? नहीं, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_ru": {
+      "task": "xnli_ru",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "ru",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", правильно? Да, \"+hypothesis,premise+\", правильно? Так, \"+hypothesis,premise+\", правильно? Нет, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_sw": {
+      "task": "xnli_sw",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "sw",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", sahihi? Ndiyo, \"+hypothesis,premise+\", sahihi? Hivyo, \"+hypothesis,premise+\", sahihi? Hapana, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_th": {
+      "task": "xnli_th",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "th",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", ถูกต้อง? ใช่, \"+hypothesis,premise+\", ถูกต้อง? ดังนั้น, \"+hypothesis,premise+\", ถูกต้อง? ไม่, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_tr": {
+      "task": "xnli_tr",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "tr",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", doğru? Evet, \"+hypothesis,premise+\", doğru? Böylece, \"+hypothesis,premise+\", doğru? Hayır, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_ur": {
+      "task": "xnli_ur",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "ur",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", صحیح? جی ہاں, \"+hypothesis,premise+\", صحیح? اس لئے, \"+hypothesis,premise+\", صحیح? نہیں, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_vi": {
+      "task": "xnli_vi",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "vi",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", đúng? Vâng, \"+hypothesis,premise+\", đúng? Vì vậy, \"+hypothesis,premise+\", đúng? Không, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xnli_zh": {
+      "task": "xnli_zh",
+      "group": "xnli",
+      "dataset_path": "xnli",
+      "dataset_name": "zh",
+      "training_split": "train",
+      "validation_split": "validation",
+      "doc_to_text": "",
+      "doc_to_target": "label",
+      "doc_to_choice": "{{[premise+\", 正确? 是的, \"+hypothesis,premise+\", 正确? 所以, \"+hypothesis,premise+\", 正确? 不是的, \"+hypothesis]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    }
+  },
+  "versions": {
+    "xnli": "N/A",
+    "xnli_ar": 1.0,
+    "xnli_bg": 1.0,
+    "xnli_de": 1.0,
+    "xnli_el": 1.0,
+    "xnli_en": 1.0,
+    "xnli_es": 1.0,
+    "xnli_fr": 1.0,
+    "xnli_hi": 1.0,
+    "xnli_ru": 1.0,
+    "xnli_sw": 1.0,
+    "xnli_th": 1.0,
+    "xnli_tr": 1.0,
+    "xnli_ur": 1.0,
+    "xnli_vi": 1.0,
+    "xnli_zh": 1.0
+  },
+  "n-shot": {
+    "xnli": 0,
+    "xnli_ar": 0,
+    "xnli_bg": 0,
+    "xnli_de": 0,
+    "xnli_el": 0,
+    "xnli_en": 0,
+    "xnli_es": 0,
+    "xnli_fr": 0,
+    "xnli_hi": 0,
+    "xnli_ru": 0,
+    "xnli_sw": 0,
+    "xnli_th": 0,
+    "xnli_tr": 0,
+    "xnli_ur": 0,
+    "xnli_vi": 0,
+    "xnli_zh": 0
+  },
+  "config": {
+    "model": "hf",
+    "model_args": "pretrained=facebook/opt-6.7b,dtype=bfloat16,trust_remote_code=True",
+    "batch_size": "auto",
+    "batch_sizes": [
+      16
+    ],
+    "device": null,
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": null
+  },
+  "git_hash": "2c0a875"
+}

lm-eval-output/facebook/opt-6.7b/xnli/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/taskrun.log

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcdb60ce05adb597c21d45d643aa60b56e8181b964ba1199dbfe2d850cb474f9
+size 76734

lm-eval-output/facebook/opt-6.7b/xstorycloze/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/result-jsonl.tar.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f75eae201d4e3a7bfe779b77cc6c5563a9e456596f003c970e7dd6847ffa1e9
+size 4064740

lm-eval-output/facebook/opt-6.7b/xstorycloze/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/results.json

@@ -0,0 +1,423 @@
+{
+  "results": {
+    "xstorycloze": {
+      "acc,none": 0.5250586607304013,
+      "acc_stderr,none": 0.06681237456634356,
+      "alias": "xstorycloze"
+    },
+    "xstorycloze_ar": {
+      "acc,none": 0.4870946393117141,
+      "acc_stderr,none": 0.01286283860572848,
+      "alias": " - xstorycloze_ar"
+    },
+    "xstorycloze_en": {
+      "acc,none": 0.7352746525479815,
+      "acc_stderr,none": 0.01135361301038959,
+      "alias": " - xstorycloze_en"
+    },
+    "xstorycloze_es": {
+      "acc,none": 0.5837193911317009,
+      "acc_stderr,none": 0.012685473350967527,
+      "alias": " - xstorycloze_es"
+    },
+    "xstorycloze_eu": {
+      "acc,none": 0.4937127729980146,
+      "acc_stderr,none": 0.01286610802121821,
+      "alias": " - xstorycloze_eu"
+    },
+    "xstorycloze_hi": {
+      "acc,none": 0.5095962938451357,
+      "acc_stderr,none": 0.012864755260408957,
+      "alias": " - xstorycloze_hi"
+    },
+    "xstorycloze_id": {
+      "acc,none": 0.4877564526803441,
+      "acc_stderr,none": 0.012863267059205548,
+      "alias": " - xstorycloze_id"
+    },
+    "xstorycloze_my": {
+      "acc,none": 0.46591661151555264,
+      "acc_stderr,none": 0.012837195610619431,
+      "alias": " - xstorycloze_my"
+    },
+    "xstorycloze_ru": {
+      "acc,none": 0.4930509596293845,
+      "acc_stderr,none": 0.012865882570960722,
+      "alias": " - xstorycloze_ru"
+    },
+    "xstorycloze_sw": {
+      "acc,none": 0.5043017868960953,
+      "acc_stderr,none": 0.012866649085718848,
+      "alias": " - xstorycloze_sw"
+    },
+    "xstorycloze_te": {
+      "acc,none": 0.5268034414295168,
+      "acc_stderr,none": 0.012848623899505763,
+      "alias": " - xstorycloze_te"
+    },
+    "xstorycloze_zh": {
+      "acc,none": 0.4884182660489742,
+      "acc_stderr,none": 0.012863672949335882,
+      "alias": " - xstorycloze_zh"
+    }
+  },
+  "groups": {
+    "xstorycloze": {
+      "acc,none": 0.5250586607304013,
+      "acc_stderr,none": 0.06681237456634356,
+      "alias": "xstorycloze"
+    }
+  },
+  "configs": {
+    "xstorycloze_ar": {
+      "task": "xstorycloze_ar",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "ar",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_en": {
+      "task": "xstorycloze_en",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "en",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_es": {
+      "task": "xstorycloze_es",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "es",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_eu": {
+      "task": "xstorycloze_eu",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "eu",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_hi": {
+      "task": "xstorycloze_hi",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "hi",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_id": {
+      "task": "xstorycloze_id",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "id",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_my": {
+      "task": "xstorycloze_my",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "my",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_ru": {
+      "task": "xstorycloze_ru",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "ru",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_sw": {
+      "task": "xstorycloze_sw",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "sw",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_te": {
+      "task": "xstorycloze_te",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "te",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xstorycloze_zh": {
+      "task": "xstorycloze_zh",
+      "group": "xstorycloze",
+      "dataset_path": "juletxara/xstory_cloze",
+      "dataset_name": "zh",
+      "training_split": "train",
+      "validation_split": "eval",
+      "doc_to_text": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "doc_to_target": "{{answer_right_ending-1}}",
+      "doc_to_choice": "{{[sentence_quiz1, sentence_quiz2]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{[input_sentence_1, input_sentence_2, input_sentence_3, input_sentence_4]|join(' ')}}",
+      "metadata": {
+        "version": 1.0
+      }
+    }
+  },
+  "versions": {
+    "xstorycloze": "N/A",
+    "xstorycloze_ar": 1.0,
+    "xstorycloze_en": 1.0,
+    "xstorycloze_es": 1.0,
+    "xstorycloze_eu": 1.0,
+    "xstorycloze_hi": 1.0,
+    "xstorycloze_id": 1.0,
+    "xstorycloze_my": 1.0,
+    "xstorycloze_ru": 1.0,
+    "xstorycloze_sw": 1.0,
+    "xstorycloze_te": 1.0,
+    "xstorycloze_zh": 1.0
+  },
+  "n-shot": {
+    "xstorycloze": 0,
+    "xstorycloze_ar": 0,
+    "xstorycloze_en": 0,
+    "xstorycloze_es": 0,
+    "xstorycloze_eu": 0,
+    "xstorycloze_hi": 0,
+    "xstorycloze_id": 0,
+    "xstorycloze_my": 0,
+    "xstorycloze_ru": 0,
+    "xstorycloze_sw": 0,
+    "xstorycloze_te": 0,
+    "xstorycloze_zh": 0
+  },
+  "config": {
+    "model": "hf",
+    "model_args": "pretrained=facebook/opt-6.7b,dtype=bfloat16,trust_remote_code=True",
+    "batch_size": "auto",
+    "batch_sizes": [
+      4
+    ],
+    "device": null,
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": null
+  },
+  "git_hash": "2c0a875"
+}

lm-eval-output/facebook/opt-6.7b/xstorycloze/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/taskrun.log

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbf3d178a7da61b1b225ea4abc0fb4456cc9e8d830b30ce68f21fc02e0aeb5e8
+size 60974

lm-eval-output/facebook/opt-6.7b/xwinograd/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/result-jsonl.tar.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec202cf302b3027efe92b4c44f4b8cc9065413154cdecbb722e1ee20f2d5ce2e
+size 512992

lm-eval-output/facebook/opt-6.7b/xwinograd/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/results.json

@@ -0,0 +1,248 @@
+{
+  "results": {
+    "xwinograd": {
+      "acc,none": 0.7104967408406383,
+      "acc_stderr,none": 0.08278874226692942,
+      "alias": "xwinograd"
+    },
+    "xwinograd_en": {
+      "acc,none": 0.8520430107526882,
+      "acc_stderr,none": 0.007365130391822567,
+      "alias": " - xwinograd_en"
+    },
+    "xwinograd_fr": {
+      "acc,none": 0.6626506024096386,
+      "acc_stderr,none": 0.05221260262032129,
+      "alias": " - xwinograd_fr"
+    },
+    "xwinograd_jp": {
+      "acc,none": 0.5370177267987487,
+      "acc_stderr,none": 0.016109933135138657,
+      "alias": " - xwinograd_jp"
+    },
+    "xwinograd_pt": {
+      "acc,none": 0.6615969581749049,
+      "acc_stderr,none": 0.029232316577302644,
+      "alias": " - xwinograd_pt"
+    },
+    "xwinograd_ru": {
+      "acc,none": 0.5333333333333333,
+      "acc_stderr,none": 0.0281538589456489,
+      "alias": " - xwinograd_ru"
+    },
+    "xwinograd_zh": {
+      "acc,none": 0.5317460317460317,
+      "acc_stderr,none": 0.02224891667424951,
+      "alias": " - xwinograd_zh"
+    }
+  },
+  "groups": {
+    "xwinograd": {
+      "acc,none": 0.7104967408406383,
+      "acc_stderr,none": 0.08278874226692942,
+      "alias": "xwinograd"
+    }
+  },
+  "configs": {
+    "xwinograd_en": {
+      "task": "xwinograd_en",
+      "group": [
+        "xwinograd"
+      ],
+      "dataset_path": "Muennighoff/xwinograd",
+      "dataset_name": "en",
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc: Dict) -> int:\n    \"\"\"\n    Return index of the correct choice.\n\n    Note: We are using the \"multiple input\" mode of the multiple-choice\n        output-type, which means we use different contexts with the same target\n        for the different choices, rather than the same context and different targets.\n    \"\"\"\n    answer_to_num = {\"1\": 0, \"2\": 1}\n    return answer_to_num[doc[\"answer\"]]\n",
+      "doc_to_target": "def doc_to_target(doc: Dict) -> str:\n    \"\"\"\n    Return the target completion.\n\n    Note that this does not depend on the correct choice as we are using\n    \"multiple input\" mode.\n    \"\"\"\n    idx = doc[\"sentence\"].index(\"_\") + 1\n    return doc[\"sentence\"][idx:].strip()\n",
+      "doc_to_choice": "def doc_to_choice(doc: Dict) -> List[str]:\n    \"\"\"Return the choices that will be used as contexts in \"multiple input\" mode.\"\"\"\n    idx = doc[\"sentence\"].index(\"_\")\n    options = [doc[\"option1\"], doc[\"option2\"]]\n    return [doc[\"sentence\"][:idx] + opt for opt in options]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xwinograd_fr": {
+      "task": "xwinograd_fr",
+      "group": [
+        "xwinograd"
+      ],
+      "dataset_path": "Muennighoff/xwinograd",
+      "dataset_name": "fr",
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc: Dict) -> int:\n    \"\"\"\n    Return index of the correct choice.\n\n    Note: We are using the \"multiple input\" mode of the multiple-choice\n        output-type, which means we use different contexts with the same target\n        for the different choices, rather than the same context and different targets.\n    \"\"\"\n    answer_to_num = {\"1\": 0, \"2\": 1}\n    return answer_to_num[doc[\"answer\"]]\n",
+      "doc_to_target": "def doc_to_target(doc: Dict) -> str:\n    \"\"\"\n    Return the target completion.\n\n    Note that this does not depend on the correct choice as we are using\n    \"multiple input\" mode.\n    \"\"\"\n    idx = doc[\"sentence\"].index(\"_\") + 1\n    return doc[\"sentence\"][idx:].strip()\n",
+      "doc_to_choice": "def doc_to_choice(doc: Dict) -> List[str]:\n    \"\"\"Return the choices that will be used as contexts in \"multiple input\" mode.\"\"\"\n    idx = doc[\"sentence\"].index(\"_\")\n    options = [doc[\"option1\"], doc[\"option2\"]]\n    return [doc[\"sentence\"][:idx] + opt for opt in options]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xwinograd_jp": {
+      "task": "xwinograd_jp",
+      "group": [
+        "xwinograd"
+      ],
+      "dataset_path": "Muennighoff/xwinograd",
+      "dataset_name": "jp",
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc: Dict) -> int:\n    \"\"\"\n    Return index of the correct choice.\n\n    Note: We are using the \"multiple input\" mode of the multiple-choice\n        output-type, which means we use different contexts with the same target\n        for the different choices, rather than the same context and different targets.\n    \"\"\"\n    answer_to_num = {\"1\": 0, \"2\": 1}\n    return answer_to_num[doc[\"answer\"]]\n",
+      "doc_to_target": "def doc_to_target(doc: Dict) -> str:\n    \"\"\"\n    Return the target completion.\n\n    Note that this does not depend on the correct choice as we are using\n    \"multiple input\" mode.\n    \"\"\"\n    idx = doc[\"sentence\"].index(\"_\") + 1\n    return doc[\"sentence\"][idx:].strip()\n",
+      "doc_to_choice": "def doc_to_choice(doc: Dict) -> List[str]:\n    \"\"\"Return the choices that will be used as contexts in \"multiple input\" mode.\"\"\"\n    idx = doc[\"sentence\"].index(\"_\")\n    options = [doc[\"option1\"], doc[\"option2\"]]\n    return [doc[\"sentence\"][:idx] + opt for opt in options]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xwinograd_pt": {
+      "task": "xwinograd_pt",
+      "group": [
+        "xwinograd"
+      ],
+      "dataset_path": "Muennighoff/xwinograd",
+      "dataset_name": "pt",
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc: Dict) -> int:\n    \"\"\"\n    Return index of the correct choice.\n\n    Note: We are using the \"multiple input\" mode of the multiple-choice\n        output-type, which means we use different contexts with the same target\n        for the different choices, rather than the same context and different targets.\n    \"\"\"\n    answer_to_num = {\"1\": 0, \"2\": 1}\n    return answer_to_num[doc[\"answer\"]]\n",
+      "doc_to_target": "def doc_to_target(doc: Dict) -> str:\n    \"\"\"\n    Return the target completion.\n\n    Note that this does not depend on the correct choice as we are using\n    \"multiple input\" mode.\n    \"\"\"\n    idx = doc[\"sentence\"].index(\"_\") + 1\n    return doc[\"sentence\"][idx:].strip()\n",
+      "doc_to_choice": "def doc_to_choice(doc: Dict) -> List[str]:\n    \"\"\"Return the choices that will be used as contexts in \"multiple input\" mode.\"\"\"\n    idx = doc[\"sentence\"].index(\"_\")\n    options = [doc[\"option1\"], doc[\"option2\"]]\n    return [doc[\"sentence\"][:idx] + opt for opt in options]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xwinograd_ru": {
+      "task": "xwinograd_ru",
+      "group": [
+        "xwinograd"
+      ],
+      "dataset_path": "Muennighoff/xwinograd",
+      "dataset_name": "ru",
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc: Dict) -> int:\n    \"\"\"\n    Return index of the correct choice.\n\n    Note: We are using the \"multiple input\" mode of the multiple-choice\n        output-type, which means we use different contexts with the same target\n        for the different choices, rather than the same context and different targets.\n    \"\"\"\n    answer_to_num = {\"1\": 0, \"2\": 1}\n    return answer_to_num[doc[\"answer\"]]\n",
+      "doc_to_target": "def doc_to_target(doc: Dict) -> str:\n    \"\"\"\n    Return the target completion.\n\n    Note that this does not depend on the correct choice as we are using\n    \"multiple input\" mode.\n    \"\"\"\n    idx = doc[\"sentence\"].index(\"_\") + 1\n    return doc[\"sentence\"][idx:].strip()\n",
+      "doc_to_choice": "def doc_to_choice(doc: Dict) -> List[str]:\n    \"\"\"Return the choices that will be used as contexts in \"multiple input\" mode.\"\"\"\n    idx = doc[\"sentence\"].index(\"_\")\n    options = [doc[\"option1\"], doc[\"option2\"]]\n    return [doc[\"sentence\"][:idx] + opt for opt in options]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    },
+    "xwinograd_zh": {
+      "task": "xwinograd_zh",
+      "group": [
+        "xwinograd"
+      ],
+      "dataset_path": "Muennighoff/xwinograd",
+      "dataset_name": "zh",
+      "test_split": "test",
+      "doc_to_text": "def doc_to_text(doc: Dict) -> int:\n    \"\"\"\n    Return index of the correct choice.\n\n    Note: We are using the \"multiple input\" mode of the multiple-choice\n        output-type, which means we use different contexts with the same target\n        for the different choices, rather than the same context and different targets.\n    \"\"\"\n    answer_to_num = {\"1\": 0, \"2\": 1}\n    return answer_to_num[doc[\"answer\"]]\n",
+      "doc_to_target": "def doc_to_target(doc: Dict) -> str:\n    \"\"\"\n    Return the target completion.\n\n    Note that this does not depend on the correct choice as we are using\n    \"multiple input\" mode.\n    \"\"\"\n    idx = doc[\"sentence\"].index(\"_\") + 1\n    return doc[\"sentence\"][idx:].strip()\n",
+      "doc_to_choice": "def doc_to_choice(doc: Dict) -> List[str]:\n    \"\"\"Return the choices that will be used as contexts in \"multiple input\" mode.\"\"\"\n    idx = doc[\"sentence\"].index(\"_\")\n    options = [doc[\"option1\"], doc[\"option2\"]]\n    return [doc[\"sentence\"][:idx] + opt for opt in options]\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0
+      }
+    }
+  },
+  "versions": {
+    "xwinograd": "N/A",
+    "xwinograd_en": 1.0,
+    "xwinograd_fr": 1.0,
+    "xwinograd_jp": 1.0,
+    "xwinograd_pt": 1.0,
+    "xwinograd_ru": 1.0,
+    "xwinograd_zh": 1.0
+  },
+  "n-shot": {
+    "xwinograd": 0,
+    "xwinograd_en": 0,
+    "xwinograd_fr": 0,
+    "xwinograd_jp": 0,
+    "xwinograd_pt": 0,
+    "xwinograd_ru": 0,
+    "xwinograd_zh": 0
+  },
+  "config": {
+    "model": "hf",
+    "model_args": "pretrained=facebook/opt-6.7b,dtype=bfloat16,trust_remote_code=True",
+    "batch_size": "auto",
+    "batch_sizes": [
+      32
+    ],
+    "device": null,
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": null
+  },
+  "git_hash": "2c0a875"
+}

lm-eval-output/facebook/opt-6.7b/xwinograd/dtype=bfloat16,trust_remote_code=True-num_fewshot=-1-nvidia-gpu/taskrun.log

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5e409ba16c33b98e326f4cab0c0afe998eed06689e8e22e095d181634444c1c
+size 36604