Update README.md
Browse files
README.md
CHANGED
|
@@ -37,6 +37,64 @@ Detailed results can be found [here](https://huggingface.co/datasets/open-llm-le
|
|
| 37 |
|MuSR (0-shot) |11.15|
|
| 38 |
|MMLU-PRO (5-shot) |30.34|
|
| 39 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
|
| 41 |
# merge tree
|
| 42 |
|
|
|
|
| 37 |
|MuSR (0-shot) |11.15|
|
| 38 |
|MMLU-PRO (5-shot) |30.34|
|
| 39 |
|
| 40 |
+
| Model |AGIEval|TruthfulQA|Bigbench|
|
| 41 |
+
|--------------------------------------------------------------------------------|------:|---------:|-------:|
|
| 42 |
+
|[llama3.1-8b-spaetzle-v90](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v90)| 42.05| 57.2| 44.75|
|
| 43 |
+
|
| 44 |
+
### AGIEval
|
| 45 |
+
| Task |Version| Metric |Value| |Stderr|
|
| 46 |
+
|------------------------------|------:|--------|----:|---|-----:|
|
| 47 |
+
|agieval_aqua_rat | 0|acc |24.02|± | 2.69|
|
| 48 |
+
| | |acc_norm|23.62|± | 2.67|
|
| 49 |
+
|agieval_logiqa_en | 0|acc |40.09|± | 1.92|
|
| 50 |
+
| | |acc_norm|39.78|± | 1.92|
|
| 51 |
+
|agieval_lsat_ar | 0|acc |22.17|± | 2.75|
|
| 52 |
+
| | |acc_norm|21.74|± | 2.73|
|
| 53 |
+
|agieval_lsat_lr | 0|acc |50.39|± | 2.22|
|
| 54 |
+
| | |acc_norm|45.29|± | 2.21|
|
| 55 |
+
|agieval_lsat_rc | 0|acc |64.31|± | 2.93|
|
| 56 |
+
| | |acc_norm|58.36|± | 3.01|
|
| 57 |
+
|agieval_sat_en | 0|acc |81.07|± | 2.74|
|
| 58 |
+
| | |acc_norm|73.79|± | 3.07|
|
| 59 |
+
|agieval_sat_en_without_passage| 0|acc |45.15|± | 3.48|
|
| 60 |
+
| | |acc_norm|38.83|± | 3.40|
|
| 61 |
+
|agieval_sat_math | 0|acc |40.91|± | 3.32|
|
| 62 |
+
| | |acc_norm|35.00|± | 3.22|
|
| 63 |
+
|
| 64 |
+
Average: 42.05%
|
| 65 |
+
|
| 66 |
+
### TruthfulQA
|
| 67 |
+
| Task |Version|Metric|Value| |Stderr|
|
| 68 |
+
|-------------|------:|------|----:|---|-----:|
|
| 69 |
+
|truthfulqa_mc| 1|mc1 |39.66|± | 1.71|
|
| 70 |
+
| | |mc2 |57.20|± | 1.51|
|
| 71 |
+
|
| 72 |
+
Average: 57.2%
|
| 73 |
+
|
| 74 |
+
### Bigbench
|
| 75 |
+
| Task |Version| Metric |Value| |Stderr|
|
| 76 |
+
|------------------------------------------------|------:|---------------------|----:|---|-----:|
|
| 77 |
+
|bigbench_causal_judgement | 0|multiple_choice_grade|58.42|± | 3.59|
|
| 78 |
+
|bigbench_date_understanding | 0|multiple_choice_grade|70.46|± | 2.38|
|
| 79 |
+
|bigbench_disambiguation_qa | 0|multiple_choice_grade|31.40|± | 2.89|
|
| 80 |
+
|bigbench_geometric_shapes | 0|multiple_choice_grade|33.43|± | 2.49|
|
| 81 |
+
| | |exact_str_match | 0.00|± | 0.00|
|
| 82 |
+
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|30.00|± | 2.05|
|
| 83 |
+
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|24.29|± | 1.62|
|
| 84 |
+
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|56.00|± | 2.87|
|
| 85 |
+
|bigbench_movie_recommendation | 0|multiple_choice_grade|38.20|± | 2.18|
|
| 86 |
+
|bigbench_navigate | 0|multiple_choice_grade|50.20|± | 1.58|
|
| 87 |
+
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|69.50|± | 1.03|
|
| 88 |
+
|bigbench_ruin_names | 0|multiple_choice_grade|54.46|± | 2.36|
|
| 89 |
+
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|32.77|± | 1.49|
|
| 90 |
+
|bigbench_snarks | 0|multiple_choice_grade|65.19|± | 3.55|
|
| 91 |
+
|bigbench_sports_understanding | 0|multiple_choice_grade|50.30|± | 1.59|
|
| 92 |
+
|bigbench_temporal_sequences | 0|multiple_choice_grade|45.70|± | 1.58|
|
| 93 |
+
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|22.08|± | 1.17|
|
| 94 |
+
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|17.03|± | 0.90|
|
| 95 |
+
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|56.00|± | 2.87|
|
| 96 |
+
|
| 97 |
+
Average: 44.75%
|
| 98 |
|
| 99 |
# merge tree
|
| 100 |
|