Update README.md
Browse files
README.md
CHANGED
|
@@ -35,6 +35,87 @@ score of 64.81, but only
|
|
| 35 |
|GSM8k (5-shot) |62.70|
|
| 36 |
|
| 37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
## 🧩 Configuration
|
| 39 |
|
| 40 |
```yaml
|
|
|
|
| 35 |
|GSM8k (5-shot) |62.70|
|
| 36 |
|
| 37 |
|
| 38 |
+
| Model |AGIEval|GPT4All|TruthfulQA|Bigbench|Average|
|
| 39 |
+
|--------------------------------------------------------------|------:|------:|---------:|-------:|------:|
|
| 40 |
+
|[Spaetzle-v12-7b](https://huggingface.co/cstr/Spaetzle-v12-7b)| 42.64| 74.3| 58.44| 44.44| 54.95|
|
| 41 |
+
|
| 42 |
+
### AGIEval
|
| 43 |
+
| Task |Version| Metric |Value| |Stderr|
|
| 44 |
+
|------------------------------|------:|--------|----:|---|-----:|
|
| 45 |
+
|agieval_aqua_rat | 0|acc |24.02|± | 2.69|
|
| 46 |
+
| | |acc_norm|21.65|± | 2.59|
|
| 47 |
+
|agieval_logiqa_en | 0|acc |36.10|± | 1.88|
|
| 48 |
+
| | |acc_norm|37.63|± | 1.90|
|
| 49 |
+
|agieval_lsat_ar | 0|acc |24.35|± | 2.84|
|
| 50 |
+
| | |acc_norm|23.04|± | 2.78|
|
| 51 |
+
|agieval_lsat_lr | 0|acc |48.82|± | 2.22|
|
| 52 |
+
| | |acc_norm|47.25|± | 2.21|
|
| 53 |
+
|agieval_lsat_rc | 0|acc |60.59|± | 2.98|
|
| 54 |
+
| | |acc_norm|57.99|± | 3.01|
|
| 55 |
+
|agieval_sat_en | 0|acc |76.21|± | 2.97|
|
| 56 |
+
| | |acc_norm|74.76|± | 3.03|
|
| 57 |
+
|agieval_sat_en_without_passage| 0|acc |46.60|± | 3.48|
|
| 58 |
+
| | |acc_norm|45.63|± | 3.48|
|
| 59 |
+
|agieval_sat_math | 0|acc |37.27|± | 3.27|
|
| 60 |
+
| | |acc_norm|33.18|± | 3.18|
|
| 61 |
+
|
| 62 |
+
Average: 42.64%
|
| 63 |
+
|
| 64 |
+
### GPT4All
|
| 65 |
+
| Task |Version| Metric |Value| |Stderr|
|
| 66 |
+
|-------------|------:|--------|----:|---|-----:|
|
| 67 |
+
|arc_challenge| 0|acc |59.13|± | 1.44|
|
| 68 |
+
| | |acc_norm|61.26|± | 1.42|
|
| 69 |
+
|arc_easy | 0|acc |83.67|± | 0.76|
|
| 70 |
+
| | |acc_norm|80.89|± | 0.81|
|
| 71 |
+
|boolq | 1|acc |87.83|± | 0.57|
|
| 72 |
+
|hellaswag | 0|acc |66.45|± | 0.47|
|
| 73 |
+
| | |acc_norm|84.63|± | 0.36|
|
| 74 |
+
|openbookqa | 0|acc |37.40|± | 2.17|
|
| 75 |
+
| | |acc_norm|45.80|± | 2.23|
|
| 76 |
+
|piqa | 0|acc |82.15|± | 0.89|
|
| 77 |
+
| | |acc_norm|83.13|± | 0.87|
|
| 78 |
+
|winogrande | 0|acc |76.56|± | 1.19|
|
| 79 |
+
|
| 80 |
+
Average: 74.3%
|
| 81 |
+
|
| 82 |
+
### TruthfulQA
|
| 83 |
+
| Task |Version|Metric|Value| |Stderr|
|
| 84 |
+
|-------------|------:|------|----:|---|-----:|
|
| 85 |
+
|truthfulqa_mc| 1|mc1 |42.59|± | 1.73|
|
| 86 |
+
| | |mc2 |58.44|± | 1.58|
|
| 87 |
+
|
| 88 |
+
Average: 58.44%
|
| 89 |
+
|
| 90 |
+
### Bigbench
|
| 91 |
+
| Task |Version| Metric |Value| |Stderr|
|
| 92 |
+
|------------------------------------------------|------:|---------------------|----:|---|-----:|
|
| 93 |
+
|bigbench_causal_judgement | 0|multiple_choice_grade|55.26|± | 3.62|
|
| 94 |
+
|bigbench_date_understanding | 0|multiple_choice_grade|64.77|± | 2.49|
|
| 95 |
+
|bigbench_disambiguation_qa | 0|multiple_choice_grade|37.60|± | 3.02|
|
| 96 |
+
|bigbench_geometric_shapes | 0|multiple_choice_grade|32.31|± | 2.47|
|
| 97 |
+
| | |exact_str_match |21.45|± | 2.17|
|
| 98 |
+
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|31.00|± | 2.07|
|
| 99 |
+
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|22.43|± | 1.58|
|
| 100 |
+
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|53.00|± | 2.89|
|
| 101 |
+
|bigbench_movie_recommendation | 0|multiple_choice_grade|40.40|± | 2.20|
|
| 102 |
+
|bigbench_navigate | 0|multiple_choice_grade|51.30|± | 1.58|
|
| 103 |
+
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|68.50|± | 1.04|
|
| 104 |
+
|bigbench_ruin_names | 0|multiple_choice_grade|48.66|± | 2.36|
|
| 105 |
+
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|30.36|± | 1.46|
|
| 106 |
+
|bigbench_snarks | 0|multiple_choice_grade|70.17|± | 3.41|
|
| 107 |
+
|bigbench_sports_understanding | 0|multiple_choice_grade|70.39|± | 1.45|
|
| 108 |
+
|bigbench_temporal_sequences | 0|multiple_choice_grade|31.00|± | 1.46|
|
| 109 |
+
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|21.44|± | 1.16|
|
| 110 |
+
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|18.29|± | 0.92|
|
| 111 |
+
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|53.00|± | 2.89|
|
| 112 |
+
|
| 113 |
+
Average: 44.44%
|
| 114 |
+
|
| 115 |
+
Average score: 54.95%
|
| 116 |
+
|
| 117 |
+
Elapsed time: 02:50:51
|
| 118 |
+
|
| 119 |
## 🧩 Configuration
|
| 120 |
|
| 121 |
```yaml
|