pretrain eval
Browse files
README.md
CHANGED
|
@@ -44,44 +44,317 @@ This model **isn't** designed for immediate use but rather for Continued Pretrai
|
|
| 44 |
|
| 45 |
The objective is to streamline the cognitive or reasoning core, eliminating any redundant knowledge from the model.
|
| 46 |
|
| 47 |
-
[loss, val_loss]()
|
| 48 |
|
| 49 |
-
[val_ppl]()
|
| 50 |
|
| 51 |
-
[epoch]()
|
| 52 |
|
| 53 |
-
[learning_rate]()
|
| 54 |
|
| 55 |
-
##
|
|
|
|
|
|
|
| 56 |
|
| 57 |
```bash
|
| 58 |
litgpt evaluate --tasks 'hellaswag,gsm8k,truthfulqa_mc2,mmlu,winogrande,arc_challenge' --out_dir 'evaluate-quick/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 59 |
```
|
| 60 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
```bash
|
| 62 |
litgpt evaluate --tasks 'leaderboard' --out_dir 'evaluate-leaderboard/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 63 |
```
|
| 64 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
```bash
|
| 66 |
-
litgpt evaluate --tasks '
|
| 67 |
```
|
| 68 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
```bash
|
| 70 |
litgpt evaluate --tasks 'mmlu,mmlu_pro' --out_dir 'evaluate-mmlu/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 71 |
```
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 80 |
|
| 81 |
```bash
|
| 82 |
-
litgpt evaluate --tasks '
|
| 83 |
```
|
| 84 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 85 |
```bash
|
| 86 |
litgpt evaluate --tasks 'wikitext,qasper' --out_dir 'evaluate-long/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 87 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
|
| 45 |
The objective is to streamline the cognitive or reasoning core, eliminating any redundant knowledge from the model.
|
| 46 |
|
| 47 |
+
[loss, val_loss](https://api.wandb.ai/links/mtasic85/x4dxkpy7)
|
| 48 |
|
| 49 |
+
[val_ppl](https://api.wandb.ai/links/mtasic85/hr03vzo2)
|
| 50 |
|
| 51 |
+
[epoch](https://api.wandb.ai/links/mtasic85/e6ai0066)
|
| 52 |
|
| 53 |
+
[learning_rate](https://api.wandb.ai/links/mtasic85/lap8xl2w)
|
| 54 |
|
| 55 |
+
## Pretrain Evaluation
|
| 56 |
+
|
| 57 |
+
### lm-evaluation-harness
|
| 58 |
|
| 59 |
```bash
|
| 60 |
litgpt evaluate --tasks 'hellaswag,gsm8k,truthfulqa_mc2,mmlu,winogrande,arc_challenge' --out_dir 'evaluate-quick/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 61 |
```
|
| 62 |
|
| 63 |
+
| Tasks |Version| Filter |n-shot| Metric | |Value | |Stderr|
|
| 64 |
+
|---------------------------------------|------:|----------------|-----:|-----------|---|-----:|---|-----:|
|
| 65 |
+
|arc_challenge | 1|none | 0|acc |↑ |0.1852|± |0.0114|
|
| 66 |
+
| | |none | 0|acc_norm |↑ |0.2201|± |0.0121|
|
| 67 |
+
|gsm8k | 3|flexible-extract| 5|exact_match|↑ |0.0205|± |0.0039|
|
| 68 |
+
| | |strict-match | 5|exact_match|↑ |0.0000|± |0.0000|
|
| 69 |
+
|hellaswag | 1|none | 0|acc |↑ |0.2628|± |0.0044|
|
| 70 |
+
| | |none | 0|acc_norm |↑ |0.2705|± |0.0044|
|
| 71 |
+
|mmlu | 2|none | |acc |↑ |0.2468|± |0.0036|
|
| 72 |
+
| - humanities | 2|none | |acc |↑ |0.2459|± |0.0063|
|
| 73 |
+
| - formal_logic | 1|none | 0|acc |↑ |0.3175|± |0.0416|
|
| 74 |
+
| - high_school_european_history | 1|none | 0|acc |↑ |0.2364|± |0.0332|
|
| 75 |
+
| - high_school_us_history | 1|none | 0|acc |↑ |0.2304|± |0.0296|
|
| 76 |
+
| - high_school_world_history | 1|none | 0|acc |↑ |0.2194|± |0.0269|
|
| 77 |
+
| - international_law | 1|none | 0|acc |↑ |0.2479|± |0.0394|
|
| 78 |
+
| - jurisprudence | 1|none | 0|acc |↑ |0.2315|± |0.0408|
|
| 79 |
+
| - logical_fallacies | 1|none | 0|acc |↑ |0.2147|± |0.0323|
|
| 80 |
+
| - moral_disputes | 1|none | 0|acc |↑ |0.2168|± |0.0222|
|
| 81 |
+
| - moral_scenarios | 1|none | 0|acc |↑ |0.2726|± |0.0149|
|
| 82 |
+
| - philosophy | 1|none | 0|acc |↑ |0.1865|± |0.0221|
|
| 83 |
+
| - prehistory | 1|none | 0|acc |↑ |0.2191|± |0.0230|
|
| 84 |
+
| - professional_law | 1|none | 0|acc |↑ |0.2490|± |0.0110|
|
| 85 |
+
| - world_religions | 1|none | 0|acc |↑ |0.3450|± |0.0365|
|
| 86 |
+
| - other | 2|none | |acc |↑ |0.2385|± |0.0076|
|
| 87 |
+
| - business_ethics | 1|none | 0|acc |↑ |0.2200|± |0.0416|
|
| 88 |
+
| - clinical_knowledge | 1|none | 0|acc |↑ |0.2264|± |0.0258|
|
| 89 |
+
| - college_medicine | 1|none | 0|acc |↑ |0.2601|± |0.0335|
|
| 90 |
+
| - global_facts | 1|none | 0|acc |↑ |0.1900|± |0.0394|
|
| 91 |
+
| - human_aging | 1|none | 0|acc |↑ |0.2422|± |0.0288|
|
| 92 |
+
| - management | 1|none | 0|acc |↑ |0.2330|± |0.0419|
|
| 93 |
+
| - marketing | 1|none | 0|acc |↑ |0.2821|± |0.0295|
|
| 94 |
+
| - medical_genetics | 1|none | 0|acc |↑ |0.2900|± |0.0456|
|
| 95 |
+
| - miscellaneous | 1|none | 0|acc |↑ |0.2388|± |0.0152|
|
| 96 |
+
| - nutrition | 1|none | 0|acc |↑ |0.1993|± |0.0229|
|
| 97 |
+
| - professional_accounting | 1|none | 0|acc |↑ |0.2270|± |0.0250|
|
| 98 |
+
| - professional_medicine | 1|none | 0|acc |↑ |0.2610|± |0.0267|
|
| 99 |
+
| - virology | 1|none | 0|acc |↑ |0.2349|± |0.0330|
|
| 100 |
+
| - social sciences | 2|none | |acc |↑ |0.2632|± |0.0079|
|
| 101 |
+
| - econometrics | 1|none | 0|acc |↑ |0.2544|± |0.0410|
|
| 102 |
+
| - high_school_geography | 1|none | 0|acc |↑ |0.1869|± |0.0278|
|
| 103 |
+
| - high_school_government_and_politics| 1|none | 0|acc |↑ |0.2850|± |0.0326|
|
| 104 |
+
| - high_school_macroeconomics | 1|none | 0|acc |↑ |0.3128|± |0.0235|
|
| 105 |
+
| - high_school_microeconomics | 1|none | 0|acc |↑ |0.2773|± |0.0291|
|
| 106 |
+
| - high_school_psychology | 1|none | 0|acc |↑ |0.2422|± |0.0184|
|
| 107 |
+
| - human_sexuality | 1|none | 0|acc |↑ |0.2595|± |0.0384|
|
| 108 |
+
| - professional_psychology | 1|none | 0|acc |↑ |0.2435|± |0.0174|
|
| 109 |
+
| - public_relations | 1|none | 0|acc |↑ |0.2273|± |0.0401|
|
| 110 |
+
| - security_studies | 1|none | 0|acc |↑ |0.3265|± |0.0300|
|
| 111 |
+
| - sociology | 1|none | 0|acc |↑ |0.2537|± |0.0308|
|
| 112 |
+
| - us_foreign_policy | 1|none | 0|acc |↑ |0.3000|± |0.0461|
|
| 113 |
+
| - stem | 2|none | |acc |↑ |0.2404|± |0.0076|
|
| 114 |
+
| - abstract_algebra | 1|none | 0|acc |↑ |0.1700|± |0.0378|
|
| 115 |
+
| - anatomy | 1|none | 0|acc |↑ |0.2074|± |0.0350|
|
| 116 |
+
| - astronomy | 1|none | 0|acc |↑ |0.2105|± |0.0332|
|
| 117 |
+
| - college_biology | 1|none | 0|acc |↑ |0.2153|± |0.0344|
|
| 118 |
+
| - college_chemistry | 1|none | 0|acc |↑ |0.2000|± |0.0402|
|
| 119 |
+
| - college_computer_science | 1|none | 0|acc |↑ |0.2300|± |0.0423|
|
| 120 |
+
| - college_mathematics | 1|none | 0|acc |↑ |0.1700|± |0.0378|
|
| 121 |
+
| - college_physics | 1|none | 0|acc |↑ |0.2647|± |0.0439|
|
| 122 |
+
| - computer_security | 1|none | 0|acc |↑ |0.2700|± |0.0446|
|
| 123 |
+
| - conceptual_physics | 1|none | 0|acc |↑ |0.2766|± |0.0292|
|
| 124 |
+
| - electrical_engineering | 1|none | 0|acc |↑ |0.2552|± |0.0363|
|
| 125 |
+
| - elementary_mathematics | 1|none | 0|acc |↑ |0.2566|± |0.0225|
|
| 126 |
+
| - high_school_biology | 1|none | 0|acc |↑ |0.2097|± |0.0232|
|
| 127 |
+
| - high_school_chemistry | 1|none | 0|acc |↑ |0.2611|± |0.0309|
|
| 128 |
+
| - high_school_computer_science | 1|none | 0|acc |↑ |0.2600|± |0.0441|
|
| 129 |
+
| - high_school_mathematics | 1|none | 0|acc |↑ |0.2111|± |0.0249|
|
| 130 |
+
| - high_school_physics | 1|none | 0|acc |↑ |0.2517|± |0.0354|
|
| 131 |
+
| - high_school_statistics | 1|none | 0|acc |↑ |0.3056|± |0.0314|
|
| 132 |
+
| - machine_learning | 1|none | 0|acc |↑ |0.2857|± |0.0429|
|
| 133 |
+
|truthfulqa_mc2 | 2|none | 0|acc |↑ |0.5010|± |0.0159|
|
| 134 |
+
|winogrande | 1|none | 0|acc |↑ |0.5130|± |0.0140|
|
| 135 |
+
|
| 136 |
+
| Groups |Version|Filter|n-shot|Metric| |Value | |Stderr|
|
| 137 |
+
|------------------|------:|------|------|------|---|-----:|---|-----:|
|
| 138 |
+
|mmlu | 2|none | |acc |↑ |0.2468|± |0.0036|
|
| 139 |
+
| - humanities | 2|none | |acc |↑ |0.2459|± |0.0063|
|
| 140 |
+
| - other | 2|none | |acc |↑ |0.2385|± |0.0076|
|
| 141 |
+
| - social sciences| 2|none | |acc |↑ |0.2632|± |0.0079|
|
| 142 |
+
| - stem | 2|none | |acc |↑ |0.2404|± |0.0076|
|
| 143 |
+
|
| 144 |
```bash
|
| 145 |
litgpt evaluate --tasks 'leaderboard' --out_dir 'evaluate-leaderboard/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 146 |
```
|
| 147 |
|
| 148 |
+
| Tasks |Version|Filter|n-shot| Metric | |Value | |Stderr|
|
| 149 |
+
|-----------------------------------------------------------|-------|------|-----:|-----------------------|---|-----:|---|------|
|
| 150 |
+
|leaderboard | N/A| | | | | | | |
|
| 151 |
+
| - leaderboard_bbh | N/A| | | | | | | |
|
| 152 |
+
| - leaderboard_bbh_boolean_expressions | 1|none | 3|acc_norm |↑ |0.4680|± |0.0316|
|
| 153 |
+
| - leaderboard_bbh_causal_judgement | 1|none | 3|acc_norm |↑ |0.5187|± |0.0366|
|
| 154 |
+
| - leaderboard_bbh_date_understanding | 1|none | 3|acc_norm |↑ |0.1880|± |0.0248|
|
| 155 |
+
| - leaderboard_bbh_disambiguation_qa | 1|none | 3|acc_norm |↑ |0.3440|± |0.0301|
|
| 156 |
+
| - leaderboard_bbh_formal_fallacies | 1|none | 3|acc_norm |↑ |0.4720|± |0.0316|
|
| 157 |
+
| - leaderboard_bbh_geometric_shapes | 1|none | 3|acc_norm |↑ |0.1200|± |0.0206|
|
| 158 |
+
| - leaderboard_bbh_hyperbaton | 1|none | 3|acc_norm |↑ |0.5240|± |0.0316|
|
| 159 |
+
| - leaderboard_bbh_logical_deduction_five_objects | 1|none | 3|acc_norm |↑ |0.2160|± |0.0261|
|
| 160 |
+
| - leaderboard_bbh_logical_deduction_seven_objects | 1|none | 3|acc_norm |↑ |0.1400|± |0.0220|
|
| 161 |
+
| - leaderboard_bbh_logical_deduction_three_objects | 1|none | 3|acc_norm |↑ |0.3200|± |0.0296|
|
| 162 |
+
| - leaderboard_bbh_movie_recommendation | 1|none | 3|acc_norm |↑ |0.2360|± |0.0269|
|
| 163 |
+
| - leaderboard_bbh_navigate | 1|none | 3|acc_norm |↑ |0.4200|± |0.0313|
|
| 164 |
+
| - leaderboard_bbh_object_counting | 1|none | 3|acc_norm |↑ |0.1000|± |0.0190|
|
| 165 |
+
| - leaderboard_bbh_penguins_in_a_table | 1|none | 3|acc_norm |↑ |0.1575|± |0.0303|
|
| 166 |
+
| - leaderboard_bbh_reasoning_about_colored_objects | 1|none | 3|acc_norm |↑ |0.0920|± |0.0183|
|
| 167 |
+
| - leaderboard_bbh_ruin_names | 1|none | 3|acc_norm |↑ |0.2480|± |0.0274|
|
| 168 |
+
| - leaderboard_bbh_salient_translation_error_detection | 1|none | 3|acc_norm |↑ |0.1200|± |0.0206|
|
| 169 |
+
| - leaderboard_bbh_snarks | 1|none | 3|acc_norm |↑ |0.4888|± |0.0376|
|
| 170 |
+
| - leaderboard_bbh_sports_understanding | 1|none | 3|acc_norm |↑ |0.4600|± |0.0316|
|
| 171 |
+
| - leaderboard_bbh_temporal_sequences | 1|none | 3|acc_norm |↑ |0.2440|± |0.0272|
|
| 172 |
+
| - leaderboard_bbh_tracking_shuffled_objects_five_objects | 1|none | 3|acc_norm |↑ |0.1560|± |0.0230|
|
| 173 |
+
| - leaderboard_bbh_tracking_shuffled_objects_seven_objects| 1|none | 3|acc_norm |↑ |0.0960|± |0.0187|
|
| 174 |
+
| - leaderboard_bbh_tracking_shuffled_objects_three_objects| 1|none | 3|acc_norm |↑ |0.3800|± |0.0308|
|
| 175 |
+
| - leaderboard_bbh_web_of_lies | 1|none | 3|acc_norm |↑ |0.4720|± |0.0316|
|
| 176 |
+
| - leaderboard_gpqa | N/A| | | | | | | |
|
| 177 |
+
| - leaderboard_gpqa_diamond | 1|none | 0|acc_norm |↑ |0.1970|± |0.0283|
|
| 178 |
+
| - leaderboard_gpqa_extended | 1|none | 0|acc_norm |↑ |0.2509|± |0.0186|
|
| 179 |
+
| - leaderboard_gpqa_main | 1|none | 0|acc_norm |↑ |0.2589|± |0.0207|
|
| 180 |
+
| - leaderboard_ifeval | 3|none | 0|inst_level_loose_acc |↑ |0.2650|± | N/A|
|
| 181 |
+
| | |none | 0|inst_level_strict_acc |↑ |0.2530|± | N/A|
|
| 182 |
+
| | |none | 0|prompt_level_loose_acc |↑ |0.1590|± |0.0157|
|
| 183 |
+
| | |none | 0|prompt_level_strict_acc|↑ |0.1553|± |0.0156|
|
| 184 |
+
| - leaderboard_math_hard | N/A| | | | | | | |
|
| 185 |
+
| - leaderboard_math_algebra_hard | 1|none | 4|exact_match |↑ |0.0000|± | 0|
|
| 186 |
+
| - leaderboard_math_counting_and_prob_hard | 1|none | 4|exact_match |↑ |0.0000|± | 0|
|
| 187 |
+
| - leaderboard_math_geometry_hard | 1|none | 4|exact_match |↑ |0.0000|± | 0|
|
| 188 |
+
| - leaderboard_math_intermediate_algebra_hard | 1|none | 4|exact_match |↑ |0.0000|± | 0|
|
| 189 |
+
| - leaderboard_math_num_theory_hard | 1|none | 4|exact_match |↑ |0.0000|± | 0|
|
| 190 |
+
| - leaderboard_math_prealgebra_hard | 1|none | 4|exact_match |↑ |0.0000|± | 0|
|
| 191 |
+
| - leaderboard_math_precalculus_hard | 1|none | 4|exact_match |↑ |0.0000|± | 0|
|
| 192 |
+
| - leaderboard_mmlu_pro | 0.1|none | 5|acc |↑ |0.1174|± |0.0029|
|
| 193 |
+
| - leaderboard_musr | N/A| | | | | | | |
|
| 194 |
+
| - leaderboard_musr_murder_mysteries | 1|none | 0|acc_norm |↑ |0.5160|± |0.0317|
|
| 195 |
+
| - leaderboard_musr_object_placements | 1|none | 0|acc_norm |↑ |0.2695|± |0.0278|
|
| 196 |
+
| - leaderboard_musr_team_allocation | 1|none | 0|acc_norm |↑ |0.3480|± |0.0302|
|
| 197 |
+
|
| 198 |
```bash
|
| 199 |
+
litgpt evaluate --tasks 'gsm8k,mathqa' --out_dir 'evaluate-math/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 200 |
```
|
| 201 |
|
| 202 |
+
|Tasks |Version| Filter |n-shot| Metric | |Value | |Stderr|
|
| 203 |
+
|------|------:|----------------|-----:|-----------|---|-----:|---|-----:|
|
| 204 |
+
|gsm8k | 3|flexible-extract| 5|exact_match|↑ |0.0205|± |0.0039|
|
| 205 |
+
| | |strict-match | 5|exact_match|↑ |0.0000|± |0.0000|
|
| 206 |
+
|mathqa| 1|none | 0|acc |↑ |0.2010|± |0.0073|
|
| 207 |
+
| | |none | 0|acc_norm |↑ |0.2077|± |0.0074|
|
| 208 |
+
|
| 209 |
```bash
|
| 210 |
litgpt evaluate --tasks 'mmlu,mmlu_pro' --out_dir 'evaluate-mmlu/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 211 |
```
|
| 212 |
|
| 213 |
+
| Tasks |Version| Filter |n-shot| Metric | |Value | |Stderr|
|
| 214 |
+
|---------------------------------------|------:|--------------|-----:|-----------|---|-----:|---|-----:|
|
| 215 |
+
|mmlu | 2|none | |acc |↑ |0.2468|± |0.0036|
|
| 216 |
+
| - humanities | 2|none | |acc |↑ |0.2459|± |0.0063|
|
| 217 |
+
| - formal_logic | 1|none | 0|acc |↑ |0.3175|± |0.0416|
|
| 218 |
+
| - high_school_european_history | 1|none | 0|acc |↑ |0.2364|± |0.0332|
|
| 219 |
+
| - high_school_us_history | 1|none | 0|acc |↑ |0.2304|± |0.0296|
|
| 220 |
+
| - high_school_world_history | 1|none | 0|acc |↑ |0.2194|± |0.0269|
|
| 221 |
+
| - international_law | 1|none | 0|acc |↑ |0.2479|± |0.0394|
|
| 222 |
+
| - jurisprudence | 1|none | 0|acc |↑ |0.2315|± |0.0408|
|
| 223 |
+
| - logical_fallacies | 1|none | 0|acc |↑ |0.2147|± |0.0323|
|
| 224 |
+
| - moral_disputes | 1|none | 0|acc |↑ |0.2168|± |0.0222|
|
| 225 |
+
| - moral_scenarios | 1|none | 0|acc |↑ |0.2726|± |0.0149|
|
| 226 |
+
| - philosophy | 1|none | 0|acc |↑ |0.1865|± |0.0221|
|
| 227 |
+
| - prehistory | 1|none | 0|acc |↑ |0.2191|± |0.0230|
|
| 228 |
+
| - professional_law | 1|none | 0|acc |↑ |0.2490|± |0.0110|
|
| 229 |
+
| - world_religions | 1|none | 0|acc |↑ |0.3450|± |0.0365|
|
| 230 |
+
| - other | 2|none | |acc |↑ |0.2385|± |0.0076|
|
| 231 |
+
| - business_ethics | 1|none | 0|acc |↑ |0.2200|± |0.0416|
|
| 232 |
+
| - clinical_knowledge | 1|none | 0|acc |↑ |0.2264|± |0.0258|
|
| 233 |
+
| - college_medicine | 1|none | 0|acc |↑ |0.2601|± |0.0335|
|
| 234 |
+
| - global_facts | 1|none | 0|acc |↑ |0.1900|± |0.0394|
|
| 235 |
+
| - human_aging | 1|none | 0|acc |↑ |0.2422|± |0.0288|
|
| 236 |
+
| - management | 1|none | 0|acc |↑ |0.2330|± |0.0419|
|
| 237 |
+
| - marketing | 1|none | 0|acc |↑ |0.2821|± |0.0295|
|
| 238 |
+
| - medical_genetics | 1|none | 0|acc |↑ |0.2900|± |0.0456|
|
| 239 |
+
| - miscellaneous | 1|none | 0|acc |↑ |0.2388|± |0.0152|
|
| 240 |
+
| - nutrition | 1|none | 0|acc |↑ |0.1993|± |0.0229|
|
| 241 |
+
| - professional_accounting | 1|none | 0|acc |↑ |0.2270|± |0.0250|
|
| 242 |
+
| - professional_medicine | 1|none | 0|acc |↑ |0.2610|± |0.0267|
|
| 243 |
+
| - virology | 1|none | 0|acc |↑ |0.2349|± |0.0330|
|
| 244 |
+
| - social sciences | 2|none | |acc |↑ |0.2632|± |0.0079|
|
| 245 |
+
| - econometrics | 1|none | 0|acc |↑ |0.2544|± |0.0410|
|
| 246 |
+
| - high_school_geography | 1|none | 0|acc |↑ |0.1869|± |0.0278|
|
| 247 |
+
| - high_school_government_and_politics| 1|none | 0|acc |↑ |0.2850|± |0.0326|
|
| 248 |
+
| - high_school_macroeconomics | 1|none | 0|acc |↑ |0.3128|± |0.0235|
|
| 249 |
+
| - high_school_microeconomics | 1|none | 0|acc |↑ |0.2773|± |0.0291|
|
| 250 |
+
| - high_school_psychology | 1|none | 0|acc |↑ |0.2422|± |0.0184|
|
| 251 |
+
| - human_sexuality | 1|none | 0|acc |↑ |0.2595|± |0.0384|
|
| 252 |
+
| - professional_psychology | 1|none | 0|acc |↑ |0.2435|± |0.0174|
|
| 253 |
+
| - public_relations | 1|none | 0|acc |↑ |0.2273|± |0.0401|
|
| 254 |
+
| - security_studies | 1|none | 0|acc |↑ |0.3265|± |0.0300|
|
| 255 |
+
| - sociology | 1|none | 0|acc |↑ |0.2537|± |0.0308|
|
| 256 |
+
| - us_foreign_policy | 1|none | 0|acc |↑ |0.3000|± |0.0461|
|
| 257 |
+
| - stem | 2|none | |acc |↑ |0.2404|± |0.0076|
|
| 258 |
+
| - abstract_algebra | 1|none | 0|acc |↑ |0.1700|± |0.0378|
|
| 259 |
+
| - anatomy | 1|none | 0|acc |↑ |0.2074|± |0.0350|
|
| 260 |
+
| - astronomy | 1|none | 0|acc |↑ |0.2105|± |0.0332|
|
| 261 |
+
| - college_biology | 1|none | 0|acc |↑ |0.2153|± |0.0344|
|
| 262 |
+
| - college_chemistry | 1|none | 0|acc |↑ |0.2000|± |0.0402|
|
| 263 |
+
| - college_computer_science | 1|none | 0|acc |↑ |0.2300|± |0.0423|
|
| 264 |
+
| - college_mathematics | 1|none | 0|acc |↑ |0.1700|± |0.0378|
|
| 265 |
+
| - college_physics | 1|none | 0|acc |↑ |0.2647|± |0.0439|
|
| 266 |
+
| - computer_security | 1|none | 0|acc |↑ |0.2700|± |0.0446|
|
| 267 |
+
| - conceptual_physics | 1|none | 0|acc |↑ |0.2766|± |0.0292|
|
| 268 |
+
| - electrical_engineering | 1|none | 0|acc |↑ |0.2552|± |0.0363|
|
| 269 |
+
| - elementary_mathematics | 1|none | 0|acc |↑ |0.2566|± |0.0225|
|
| 270 |
+
| - high_school_biology | 1|none | 0|acc |↑ |0.2097|± |0.0232|
|
| 271 |
+
| - high_school_chemistry | 1|none | 0|acc |↑ |0.2611|± |0.0309|
|
| 272 |
+
| - high_school_computer_science | 1|none | 0|acc |↑ |0.2600|± |0.0441|
|
| 273 |
+
| - high_school_mathematics | 1|none | 0|acc |↑ |0.2111|± |0.0249|
|
| 274 |
+
| - high_school_physics | 1|none | 0|acc |↑ |0.2517|± |0.0354|
|
| 275 |
+
| - high_school_statistics | 1|none | 0|acc |↑ |0.3056|± |0.0314|
|
| 276 |
+
| - machine_learning | 1|none | 0|acc |↑ |0.2857|± |0.0429|
|
| 277 |
+
|mmlu_pro | 2|custom-extract| |exact_match|↑ |0.0000|± |0.0000|
|
| 278 |
+
| - biology | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 279 |
+
| - business | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 280 |
+
| - chemistry | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 281 |
+
| - computer_science | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 282 |
+
| - economics | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 283 |
+
| - engineering | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 284 |
+
| - health | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 285 |
+
| - history | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 286 |
+
| - law | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 287 |
+
| - math | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 288 |
+
| - other | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 289 |
+
| - philosophy | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 290 |
+
| - physics | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 291 |
+
| - psychology | 1|custom-extract| 5|exact_match|↑ |0.0000|± |0.0000|
|
| 292 |
|
| 293 |
+
| Groups |Version| Filter |n-shot| Metric | |Value | |Stderr|
|
| 294 |
+
|------------------|------:|--------------|------|-----------|---|-----:|---|-----:|
|
| 295 |
+
|mmlu | 2|none | |acc |↑ |0.2468|± |0.0036|
|
| 296 |
+
| - humanities | 2|none | |acc |↑ |0.2459|± |0.0063|
|
| 297 |
+
| - other | 2|none | |acc |↑ |0.2385|± |0.0076|
|
| 298 |
+
| - social sciences| 2|none | |acc |↑ |0.2632|± |0.0079|
|
| 299 |
+
| - stem | 2|none | |acc |↑ |0.2404|± |0.0076|
|
| 300 |
+
|mmlu_pro | 2|custom-extract| |exact_match|↑ |0.0000|± |0.0000|
|
| 301 |
|
| 302 |
```bash
|
| 303 |
+
litgpt evaluate --tasks 'arc_challenge,boolq,gpqa,hellaswag,openbookqa,piqa,truthfulqa_mc2,winogrande' --out_dir 'evaluate-reasoning/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 304 |
```
|
| 305 |
|
| 306 |
+
| Tasks |Version| Filter |n-shot| Metric | |Value | |Stderr|
|
| 307 |
+
|-------------------------------|------:|----------------|-----:|-----------|---|-----:|---|-----:|
|
| 308 |
+
|arc_challenge | 1|none | 0|acc |↑ |0.1852|± |0.0114|
|
| 309 |
+
| | |none | 0|acc_norm |↑ |0.2201|± |0.0121|
|
| 310 |
+
|boolq | 2|none | 0|acc |↑ |0.4446|± |0.0087|
|
| 311 |
+
|gpqa_diamond_cot_n_shot | 2|flexible-extract| 0|exact_match|↑ |0.0859|± |0.0200|
|
| 312 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 313 |
+
|gpqa_diamond_cot_zeroshot | 1|flexible-extract| 0|exact_match|↑ |0.0606|± |0.0170|
|
| 314 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 315 |
+
|gpqa_diamond_generative_n_shot | 2|flexible-extract| 0|exact_match|↑ |0.1717|± |0.0269|
|
| 316 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 317 |
+
|gpqa_diamond_n_shot | 2|none | 0|acc |↑ |0.2677|± |0.0315|
|
| 318 |
+
| | |none | 0|acc_norm |↑ |0.2677|± |0.0315|
|
| 319 |
+
|gpqa_diamond_zeroshot | 1|none | 0|acc |↑ |0.1970|± |0.0283|
|
| 320 |
+
| | |none | 0|acc_norm |↑ |0.1970|± |0.0283|
|
| 321 |
+
|gpqa_extended_cot_n_shot | 2|flexible-extract| 0|exact_match|↑ |0.0971|± |0.0127|
|
| 322 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 323 |
+
|gpqa_extended_cot_zeroshot | 1|flexible-extract| 0|exact_match|↑ |0.0696|± |0.0109|
|
| 324 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 325 |
+
|gpqa_extended_generative_n_shot| 2|flexible-extract| 0|exact_match|↑ |0.1502|± |0.0153|
|
| 326 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 327 |
+
|gpqa_extended_n_shot | 2|none | 0|acc |↑ |0.2399|± |0.0183|
|
| 328 |
+
| | |none | 0|acc_norm |↑ |0.2399|± |0.0183|
|
| 329 |
+
|gpqa_extended_zeroshot | 1|none | 0|acc |↑ |0.2473|± |0.0185|
|
| 330 |
+
| | |none | 0|acc_norm |↑ |0.2473|± |0.0185|
|
| 331 |
+
|gpqa_main_cot_n_shot | 2|flexible-extract| 0|exact_match|↑ |0.1116|± |0.0149|
|
| 332 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 333 |
+
|gpqa_main_cot_zeroshot | 1|flexible-extract| 0|exact_match|↑ |0.0625|± |0.0114|
|
| 334 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 335 |
+
|gpqa_main_generative_n_shot | 2|flexible-extract| 0|exact_match|↑ |0.1384|± |0.0163|
|
| 336 |
+
| | |strict-match | 0|exact_match|↑ |0.0000|± |0.0000|
|
| 337 |
+
|gpqa_main_n_shot | 2|none | 0|acc |↑ |0.2388|± |0.0202|
|
| 338 |
+
| | |none | 0|acc_norm |↑ |0.2388|± |0.0202|
|
| 339 |
+
|gpqa_main_zeroshot | 1|none | 0|acc |↑ |0.2500|± |0.0205|
|
| 340 |
+
| | |none | 0|acc_norm |↑ |0.2500|± |0.0205|
|
| 341 |
+
|hellaswag | 1|none | 0|acc |↑ |0.2628|± |0.0044|
|
| 342 |
+
| | |none | 0|acc_norm |↑ |0.2705|± |0.0044|
|
| 343 |
+
|openbookqa | 1|none | 0|acc |↑ |0.1360|± |0.0153|
|
| 344 |
+
| | |none | 0|acc_norm |↑ |0.2620|± |0.0197|
|
| 345 |
+
|piqa | 1|none | 0|acc |↑ |0.5550|± |0.0116|
|
| 346 |
+
| | |none | 0|acc_norm |↑ |0.5528|± |0.0116|
|
| 347 |
+
|truthfulqa_mc2 | 2|none | 0|acc |↑ |0.5010|± |0.0159|
|
| 348 |
+
|winogrande | 1|none | 0|acc |↑ |0.5130|± |0.0140|
|
| 349 |
+
|
| 350 |
```bash
|
| 351 |
litgpt evaluate --tasks 'wikitext,qasper' --out_dir 'evaluate-long/' --batch_size 4 --dtype 'bfloat16' out/pretrain/final/
|
| 352 |
```
|
| 353 |
+
|
| 354 |
+
| Tasks |Version|Filter|n-shot| Metric | | Value | |Stderr|
|
| 355 |
+
|---------------|------:|------|-----:|---------------|---|--------:|---|------|
|
| 356 |
+
|qasper_bool | 1|none | 0|f1 |↑ | 0.8966|± |0.0166|
|
| 357 |
+
|qasper_freeform| 2|none | 0|f1_abstractive |↑ | 0.0597|± |0.0052|
|
| 358 |
+
|wikitext | 2|none | 0|bits_per_byte |↓ | 2.2154|± | N/A|
|
| 359 |
+
| | |none | 0|byte_perplexity|↓ | 4.6441|± | N/A|
|
| 360 |
+
| | |none | 0|word_perplexity|↓ |3683.1019|± | N/A|
|