Merge remote-tracking branch 'upstream/master' into mmlu_fix

results/bloom/bloom-7b1/README.md

+# bloom-7b1
+
+## bloom-7b1_bbh_3-shot.json
+|                      Task                      |Version|       Metric        |Value|   |Stderr|
+|------------------------------------------------|------:|---------------------|----:|---|-----:|
+|bigbench_causal_judgement                       |      0|multiple_choice_grade|52.11|±  |  3.63|
+|bigbench_date_understanding                     |      0|multiple_choice_grade|36.59|±  |  2.51|
+|bigbench_disambiguation_qa                      |      0|multiple_choice_grade|26.36|±  |  2.75|
+|bigbench_dyck_languages                         |      0|multiple_choice_grade|14.40|±  |  1.11|
+|bigbench_formal_fallacies_syllogisms_negation   |      0|multiple_choice_grade|50.06|±  |  0.42|
+|bigbench_geometric_shapes                       |      0|multiple_choice_grade|20.06|±  |  2.12|
+|                                                |       |exact_str_match      | 0.00|±  |  0.00|
+|bigbench_hyperbaton                             |      0|multiple_choice_grade|48.62|±  |  0.22|
+|bigbench_logical_deduction_five_objects         |      0|multiple_choice_grade|26.00|±  |  1.96|
+|bigbench_logical_deduction_seven_objects        |      0|multiple_choice_grade|19.14|±  |  1.49|
+|bigbench_logical_deduction_three_objects        |      0|multiple_choice_grade|37.00|±  |  2.79|
+|bigbench_movie_recommendation                   |      0|multiple_choice_grade|26.40|±  |  1.97|
+|bigbench_navigate                               |      0|multiple_choice_grade|49.90|±  |  1.58|
+|bigbench_reasoning_about_colored_objects        |      0|multiple_choice_grade|24.85|±  |  0.97|
+|bigbench_ruin_names                             |      0|multiple_choice_grade|34.38|±  |  2.25|
+|bigbench_salient_translation_error_detection    |      0|multiple_choice_grade|19.14|±  |  1.25|
+|bigbench_snarks                                 |      0|multiple_choice_grade|49.72|±  |  3.73|
+|bigbench_sports_understanding                   |      0|multiple_choice_grade|50.30|±  |  1.59|
+|bigbench_temporal_sequences                     |      0|multiple_choice_grade|24.80|±  |  1.37|
+|bigbench_tracking_shuffled_objects_five_objects |      0|multiple_choice_grade|18.40|±  |  1.10|
+|bigbench_tracking_shuffled_objects_seven_objects|      0|multiple_choice_grade|14.00|±  |  0.83|
+|bigbench_tracking_shuffled_objects_three_objects|      0|multiple_choice_grade|37.00|±  |  2.79|
+
+## bloom-7b1_common_sense_reasoning_0-shot.json
+|    Task     |Version| Metric |Value|   |Stderr|
+|-------------|------:|--------|----:|---|-----:|
+|arc_challenge|      0|acc     |30.38|±  |  1.34|
+|             |       |acc_norm|33.53|±  |  1.38|
+|arc_easy     |      0|acc     |64.94|±  |  0.98|
+|             |       |acc_norm|57.32|±  |  1.01|
+|boolq        |      1|acc     |62.87|±  |  0.85|
+|copa         |      0|acc     |72.00|±  |  4.51|
+|hellaswag    |      0|acc     |46.24|±  |  0.50|
+|             |       |acc_norm|59.68|±  |  0.49|
+|mc_taco      |      0|em      |13.59|   |      |
+|             |       |f1      |50.53|   |      |
+|openbookqa   |      0|acc     |25.20|±  |  1.94|
+|             |       |acc_norm|35.80|±  |  2.15|
+|piqa         |      0|acc     |72.74|±  |  1.04|
+|             |       |acc_norm|73.67|±  |  1.03|
+|prost        |      0|acc     |26.18|±  |  0.32|
+|             |       |acc_norm|30.57|±  |  0.34|
+|swag         |      0|acc     |50.25|±  |  0.35|
+|             |       |acc_norm|68.26|±  |  0.33|
+|winogrande   |      0|acc     |64.33|±  |  1.35|
+|wsc273       |      0|acc     |81.32|±  |  2.36|
+
+## bloom-7b1_gsm8k_8-shot.json
+|Task |Version|Metric|Value|   |Stderr|
+|-----|------:|------|----:|---|-----:|
+|gsm8k|      0|acc   |  1.9|±  |  0.38|
+
+## bloom-7b1_mathematical_reasoning_few_shot_5-shot.json
+|          Task           |Version| Metric |Value|   |Stderr|
+|-------------------------|------:|--------|----:|---|-----:|
+|drop                     |      1|em      | 2.51|±  |  0.16|
+|                         |       |f1      | 5.09|±  |  0.18|
+|gsm8k                    |      0|acc     | 0.00|±  |  0.00|
+|math_algebra             |      1|acc     | 0.00|±  |  0.00|
+|math_counting_and_prob   |      1|acc     | 0.00|±  |  0.00|
+|math_geometry            |      1|acc     | 0.00|±  |  0.00|
+|math_intermediate_algebra|      1|acc     | 0.00|±  |  0.00|
+|math_num_theory          |      1|acc     | 0.00|±  |  0.00|
+|math_prealgebra          |      1|acc     | 0.00|±  |  0.00|
+|math_precalc             |      1|acc     | 0.00|±  |  0.00|
+|mathqa                   |      0|acc     |26.57|±  |  0.81|
+|                         |       |acc_norm|26.53|±  |  0.81|
+
+## bloom-7b1_pawsx_0-shot.json
+|  Task  |Version|Metric|Value|   |Stderr|
+|--------|------:|------|----:|---|-----:|
+|pawsx_de|      0|acc   |52.85|±  |  1.12|
+|pawsx_en|      0|acc   |61.30|±  |  1.09|
+|pawsx_es|      0|acc   |59.35|±  |  1.10|
+|pawsx_fr|      0|acc   |50.90|±  |  1.12|
+|pawsx_ja|      0|acc   |45.45|±  |  1.11|
+|pawsx_ko|      0|acc   |45.10|±  |  1.11|
+|pawsx_zh|      0|acc   |47.35|±  |  1.12|
+
+## bloom-7b1_question_answering_0-shot.json
+|    Task     |Version|   Metric   |Value|   |Stderr|
+|-------------|------:|------------|----:|---|-----:|
+|headqa_en    |      0|acc         |31.18|±  |  0.88|
+|             |       |acc_norm    |35.56|±  |  0.91|
+|headqa_es    |      0|acc         |29.54|±  |  0.87|
+|             |       |acc_norm    |34.32|±  |  0.91|
+|logiqa       |      0|acc         |20.28|±  |  1.58|
+|             |       |acc_norm    |28.11|±  |  1.76|
+|squad2       |      1|exact       | 7.82|   |      |
+|             |       |f1          |12.64|   |      |
+|             |       |HasAns_exact|14.84|   |      |
+|             |       |HasAns_f1   |24.51|   |      |
+|             |       |NoAns_exact | 0.81|   |      |
+|             |       |NoAns_f1    | 0.81|   |      |
+|             |       |best_exact  |50.07|   |      |
+|             |       |best_f1     |50.07|   |      |
+|triviaqa     |      1|acc         | 5.52|±  |  0.21|
+|truthfulqa_mc|      1|mc1         |22.40|±  |  1.46|
+|             |       |mc2         |38.90|±  |  1.40|
+|webqs        |      0|acc         | 2.26|±  |  0.33|
+
+## bloom-7b1_reading_comprehension_0-shot.json
+|Task|Version|Metric|Value|   |Stderr|
+|----|------:|------|----:|---|-----:|
+|coqa|      1|f1    |68.83|±  |  1.63|
+|    |       |em    |53.87|±  |  2.00|
+|drop|      1|em    | 2.57|±  |  0.16|
+|    |       |f1    | 9.85|±  |  0.21|
+|race|      1|acc   |36.56|±  |  1.49|
+
+## bloom-7b1_xcopa_0-shot.json
+|  Task  |Version|Metric|Value|   |Stderr|
+|--------|------:|------|----:|---|-----:|
+|xcopa_et|      0|acc   | 48.2|±  |  2.24|
+|xcopa_ht|      0|acc   | 50.8|±  |  2.24|
+|xcopa_id|      0|acc   | 69.8|±  |  2.06|
+|xcopa_it|      0|acc   | 52.8|±  |  2.23|
+|xcopa_qu|      0|acc   | 50.8|±  |  2.24|
+|xcopa_sw|      0|acc   | 51.6|±  |  2.24|
+|xcopa_ta|      0|acc   | 59.2|±  |  2.20|
+|xcopa_th|      0|acc   | 55.4|±  |  2.23|
+|xcopa_tr|      0|acc   | 51.2|±  |  2.24|
+|xcopa_vi|      0|acc   | 70.8|±  |  2.04|
+|xcopa_zh|      0|acc   | 65.2|±  |  2.13|
+
+## bloom-7b1_xnli_0-shot.json
+| Task  |Version|Metric|Value|   |Stderr|
+|-------|------:|------|----:|---|-----:|
+|xnli_ar|      0|acc   |33.83|±  |  0.67|
+|xnli_bg|      0|acc   |39.70|±  |  0.69|
+|xnli_de|      0|acc   |39.86|±  |  0.69|
+|xnli_el|      0|acc   |35.75|±  |  0.68|
+|xnli_en|      0|acc   |53.91|±  |  0.70|
+|xnli_es|      0|acc   |48.70|±  |  0.71|
+|xnli_fr|      0|acc   |49.68|±  |  0.71|
+|xnli_hi|      0|acc   |46.51|±  |  0.70|
+|xnli_ru|      0|acc   |43.05|±  |  0.70|
+|xnli_sw|      0|acc   |37.92|±  |  0.69|
+|xnli_th|      0|acc   |34.99|±  |  0.67|
+|xnli_tr|      0|acc   |35.09|±  |  0.67|
+|xnli_ur|      0|acc   |42.10|±  |  0.70|
+|xnli_vi|      0|acc   |47.05|±  |  0.71|
+|xnli_zh|      0|acc   |35.43|±  |  0.68|
+
+## bloom-7b1_xstory_cloze_0-shot.json
+|     Task      |Version|Metric|Value|   |Stderr|
+|---------------|------:|------|----:|---|-----:|
+|xstory_cloze_ar|      0|acc   |58.57|±  |  1.27|
+|xstory_cloze_en|      0|acc   |70.75|±  |  1.17|
+|xstory_cloze_es|      0|acc   |66.12|±  |  1.22|
+|xstory_cloze_eu|      0|acc   |57.18|±  |  1.27|
+|xstory_cloze_hi|      0|acc   |60.56|±  |  1.26|
+|xstory_cloze_id|      0|acc   |64.46|±  |  1.23|
+|xstory_cloze_my|      0|acc   |48.97|±  |  1.29|
+|xstory_cloze_ru|      0|acc   |52.75|±  |  1.28|
+|xstory_cloze_sw|      0|acc   |53.94|±  |  1.28|
+|xstory_cloze_te|      0|acc   |57.45|±  |  1.27|
+|xstory_cloze_zh|      0|acc   |61.88|±  |  1.25|
+
+## bloom-7b1_xwinograd_0-shot.json
+|    Task    |Version|Metric|Value|   |Stderr|
+|------------|------:|------|----:|---|-----:|
+|xwinograd_en|      0|acc   |82.15|±  |  0.79|
+|xwinograd_fr|      0|acc   |71.08|±  |  5.01|
+|xwinograd_jp|      0|acc   |58.50|±  |  1.59|
+|xwinograd_pt|      0|acc   |76.81|±  |  2.61|
+|xwinograd_ru|      0|acc   |56.83|±  |  2.80|
+|xwinograd_zh|      0|acc   |74.40|±  |  1.95|

results/bloom/bloom-7b1/bloom-7b1_bbh_3-shot.json

+{
+  "results": {
+    "bigbench_disambiguation_qa": {
+      "multiple_choice_grade": 0.26356589147286824,
+      "multiple_choice_grade_stderr": 0.027481788262218698
+    },
+    "bigbench_logical_deduction_three_objects": {
+      "multiple_choice_grade": 0.37,
+      "multiple_choice_grade_stderr": 0.027921294063982
+    },
+    "bigbench_causal_judgement": {
+      "multiple_choice_grade": 0.5210526315789473,
+      "multiple_choice_grade_stderr": 0.03633739504773335
+    },
+    "bigbench_date_understanding": {
+      "multiple_choice_grade": 0.36585365853658536,
+      "multiple_choice_grade_stderr": 0.025108717905729792
+    },
+    "bigbench_navigate": {
+      "multiple_choice_grade": 0.499,
+      "multiple_choice_grade_stderr": 0.015819268290576817
+    },
+    "bigbench_salient_translation_error_detection": {
+      "multiple_choice_grade": 0.19138276553106212,
+      "multiple_choice_grade_stderr": 0.012458774650265594
+    },
+    "bigbench_temporal_sequences": {
+      "multiple_choice_grade": 0.248,
+      "multiple_choice_grade_stderr": 0.013663187134877651
+    },
+    "bigbench_tracking_shuffled_objects_seven_objects": {
+      "multiple_choice_grade": 0.14,
+      "multiple_choice_grade_stderr": 0.00829694743648913
+    },
+    "bigbench_ruin_names": {
+      "multiple_choice_grade": 0.34375,
+      "multiple_choice_grade_stderr": 0.02246478414865448
+    },
+    "bigbench_reasoning_about_colored_objects": {
+      "multiple_choice_grade": 0.2485,
+      "multiple_choice_grade_stderr": 0.009665432493822852
+    },
+    "bigbench_dyck_languages": {
+      "multiple_choice_grade": 0.144,
+      "multiple_choice_grade_stderr": 0.01110798754893915
+    },
+    "bigbench_logical_deduction_five_objects": {
+      "multiple_choice_grade": 0.26,
+      "multiple_choice_grade_stderr": 0.019635965529725512
+    },
+    "bigbench_sports_understanding": {
+      "multiple_choice_grade": 0.5030425963488844,
+      "multiple_choice_grade_stderr": 0.015931029729145698
+    },
+    "bigbench_tracking_shuffled_objects_three_objects": {
+      "multiple_choice_grade": 0.37,
+      "multiple_choice_grade_stderr": 0.027921294063982
+    },
+    "bigbench_geometric_shapes": {
+      "multiple_choice_grade": 0.20055710306406685,
+      "multiple_choice_grade_stderr": 0.021162707757982353,
+      "exact_str_match": 0.0,
+      "exact_str_match_stderr": 0.0
+    },
+    "bigbench_hyperbaton": {
+      "multiple_choice_grade": 0.48618,
+      "multiple_choice_grade_stderr": 0.0022352360227943418
+    },
+    "bigbench_logical_deduction_seven_objects": {
+      "multiple_choice_grade": 0.19142857142857142,
+      "multiple_choice_grade_stderr": 0.014880721436998012
+    },
+    "bigbench_snarks": {
+      "multiple_choice_grade": 0.4972375690607735,
+      "multiple_choice_grade_stderr": 0.037267230837657574
+    },
+    "bigbench_formal_fallacies_syllogisms_negation": {
+      "multiple_choice_grade": 0.5005633802816901,
+      "multiple_choice_grade_stderr": 0.004196051878850066
+    },
+    "bigbench_tracking_shuffled_objects_five_objects": {
+      "multiple_choice_grade": 0.184,
+      "multiple_choice_grade_stderr": 0.010964094540602657
+    },
+    "bigbench_movie_recommendation": {
+      "multiple_choice_grade": 0.264,
+      "multiple_choice_grade_stderr": 0.019732885585922087
+    }
+  },
+  "versions": {
+    "bigbench_disambiguation_qa": 0,
+    "bigbench_logical_deduction_three_objects": 0,
+    "bigbench_causal_judgement": 0,
+    "bigbench_date_understanding": 0,
+    "bigbench_navigate": 0,
+    "bigbench_salient_translation_error_detection": 0,
+    "bigbench_temporal_sequences": 0,
+    "bigbench_tracking_shuffled_objects_seven_objects": 0,
+    "bigbench_ruin_names": 0,
+    "bigbench_reasoning_about_colored_objects": 0,
+    "bigbench_dyck_languages": 0,
+    "bigbench_logical_deduction_five_objects": 0,
+    "bigbench_sports_understanding": 0,
+    "bigbench_tracking_shuffled_objects_three_objects": 0,
+    "bigbench_geometric_shapes": 0,
+    "bigbench_hyperbaton": 0,
+    "bigbench_logical_deduction_seven_objects": 0,
+    "bigbench_snarks": 0,
+    "bigbench_formal_fallacies_syllogisms_negation": 0,
+    "bigbench_tracking_shuffled_objects_five_objects": 0,
+    "bigbench_movie_recommendation": 0
+  },
+  "config": {
+    "model": "hf-causal-experimental",
+    "model_args": "pretrained=bigscience/bloom-7b1,use_accelerate=True",
+    "num_fewshot": 3,
+    "batch_size": "auto",
+    "device": "cuda",
+    "no_cache": true,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "description_dict": {}
+  }
+}

results/bloom/bloom-7b1/bloom-7b1_common_sense_reasoning_0-shot.json

+{
+  "results": {
+    "copa": {
+      "acc": 0.72,
+      "acc_stderr": 0.04512608598542127
+    },
+    "winogrande": {
+      "acc": 0.6432517758484609,
+      "acc_stderr": 0.013463393958028726
+    },
+    "piqa": {
+      "acc": 0.7274211099020674,
+      "acc_stderr": 0.010389256803296021,
+      "acc_norm": 0.7366702937976061,
+      "acc_norm_stderr": 0.010276185322196764
+    },
+    "arc_challenge": {
+      "acc": 0.3037542662116041,
+      "acc_stderr": 0.013438909184778757,
+      "acc_norm": 0.33532423208191126,
+      "acc_norm_stderr": 0.013796182947785564
+    },
+    "arc_easy": {
+      "acc": 0.6494107744107744,
+      "acc_stderr": 0.009791003829831557,
+      "acc_norm": 0.5732323232323232,
+      "acc_norm_stderr": 0.010149141043955626
+    },
+    "boolq": {
+      "acc": 0.6287461773700306,
+      "acc_stderr": 0.008450174658715903
+    },
+    "wsc273": {
+      "acc": 0.8131868131868132,
+      "acc_stderr": 0.023632761722644544
+    },
+    "openbookqa": {
+      "acc": 0.252,
+      "acc_stderr": 0.019435727282249536,
+      "acc_norm": 0.358,
+      "acc_norm_stderr": 0.021461434862859122
+    },
+    "prost": {
+      "acc": 0.26184884713919726,
+      "acc_stderr": 0.003211967450351038,
+      "acc_norm": 0.30572160546541416,
+      "acc_norm_stderr": 0.003365914208405272
+    },
+    "mc_taco": {
+      "em": 0.13588588588588588,
+      "f1": 0.5052611696967991
+    },
+    "hellaswag": {
+      "acc": 0.4623580959968134,
+      "acc_stderr": 0.0049756211474061025,
+      "acc_norm": 0.5967934674367655,
+      "acc_norm_stderr": 0.0048953903414456264
+    },
+    "swag": {
+      "acc": 0.5024992502249325,
+      "acc_stderr": 0.0035350478846161142,
+      "acc_norm": 0.6825952214335699,
+      "acc_norm_stderr": 0.0032909332559412758
+    }
+  },
+  "versions": {
+    "copa": 0,
+    "winogrande": 0,
+    "piqa": 0,
+    "arc_challenge": 0,
+    "arc_easy": 0,
+    "boolq": 1,
+    "wsc273": 0,
+    "openbookqa": 0,
+    "prost": 0,
+    "mc_taco": 0,
+    "hellaswag": 0,
+    "swag": 0
+  },
+  "config": {
+    "model": "hf-causal-experimental",
+    "model_args": "pretrained=bigscience/bloom-7b1,use_accelerate=True",
+    "num_fewshot": 0,
+    "batch_size": "auto",
+    "device": "cuda:0",
+    "no_cache": true,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "description_dict": {}
+  }
+}

results/bloom/bloom-7b1/bloom-7b1_gsm8k_8-shot.json

+{
+  "results": {
+    "gsm8k": {
+      "acc": 0.018953752843062926,
+      "acc_stderr": 0.0037560783410314704
+    }
+  },
+  "versions": {
+    "gsm8k": 0
+  },
+  "config": {
+    "model": "hf-causal-experimental",
+    "model_args": "pretrained=bigscience/bloom-7b1,use_accelerate=True",
+    "num_fewshot": 8,
+    "batch_size": "auto",
+    "device": "cuda",
+    "no_cache": true,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "description_dict": {}
+  }
+}

results/bloom/bloom-7b1/bloom-7b1_pawsx_0-shot.json

+{
+  "results": {
+    "pawsx_zh": {
+      "acc": 0.4735,
+      "acc_stderr": 0.011167418260963935
+    },
+    "pawsx_de": {
+      "acc": 0.5285,
+      "acc_stderr": 0.011164954236428803
+    },
+    "pawsx_en": {
+      "acc": 0.613,
+      "acc_stderr": 0.010893798117218195
+    },
+    "pawsx_ko": {
+      "acc": 0.451,
+      "acc_stderr": 0.01112930504188632
+    },
+    "pawsx_fr": {
+      "acc": 0.509,
+      "acc_stderr": 0.011181324206260283
+    },
+    "pawsx_es": {
+      "acc": 0.5935,
+      "acc_stderr": 0.010985864536294245
+    },
+    "pawsx_ja": {
+      "acc": 0.4545,
+      "acc_stderr": 0.01113673598700373
+    }
+  },
+  "versions": {
+    "pawsx_zh": 0,
+    "pawsx_de": 0,
+    "pawsx_en": 0,
+    "pawsx_ko": 0,
+    "pawsx_fr": 0,
+    "pawsx_es": 0,
+    "pawsx_ja": 0
+  },
+  "config": {
+    "model": "hf-causal-experimental",
+    "model_args": "pretrained=bigscience/bloom-7b1",
+    "num_fewshot": 0,
+    "batch_size": "auto",
+    "device": "cuda",
+    "no_cache": true,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "description_dict": {}
+  }
+}