mirror of
https://github.com/open-compass/opencompass.git
synced 2025-05-30 16:03:24 +08:00
![]() * add ceval, gsm8k modelscope surpport * update race, mmlu, arc, cmmlu, commonsenseqa, humaneval and unittest * update bbh, flores, obqa, siqa, storycloze, summedits, winogrande, xsum datasets * format file * format file * update dataset format * support ms_dataset * udpate dataset for modelscope support * merge myl_dev and update test_ms_dataset * udpate dataset for modelscope support * update readme * update eval_api_zhipu_v2 * remove unused code * add get_data_path function * update readme * remove tydiqa japanese subset * add ceval, gsm8k modelscope surpport * update race, mmlu, arc, cmmlu, commonsenseqa, humaneval and unittest * update bbh, flores, obqa, siqa, storycloze, summedits, winogrande, xsum datasets * format file * format file * update dataset format * support ms_dataset * udpate dataset for modelscope support * merge myl_dev and update test_ms_dataset * update readme * udpate dataset for modelscope support * update eval_api_zhipu_v2 * remove unused code * add get_data_path function * remove tydiqa japanese subset * update util * remove .DS_Store * fix md format * move util into package * update docs/get_started.md * restore eval_api_zhipu_v2.py, add environment setting * Update dataset * Update * Update * Update * Update --------- Co-authored-by: Yun lin <yunlin@U-Q9X2K4QV-1904.local> Co-authored-by: Yunnglin <mao.looper@qq.com> Co-authored-by: Yun lin <yunlin@laptop.local> Co-authored-by: Yunnglin <maoyl@smail.nju.edu.cn> Co-authored-by: zhangsongyang <zhangsongyang@pjlab.org.cn> |
||
---|---|---|
.. | ||
deprecated_math_agent_evaluatorv2_gen_861b4f.py | ||
deprecated_math_evaluatorv2_gen_265cce.py | ||
math_0shot_gen_393424.py | ||
math_4shot_base_gen_db136b.py | ||
math_4shot_example_from_google_research.py | ||
math_agent_evaluatorv2_gen_0c1b4e.py | ||
math_agent_gen_0c1b4e.py | ||
math_agent_gen_861b4f.py | ||
math_agent_gen_af2293.py | ||
math_evaluatorv2_gen_2f4a71.py | ||
math_evaluatorv2_gen_cecb31.py | ||
math_gen_1ed9c2.py | ||
math_gen_5e8458.py | ||
math_gen_78ced2.py | ||
math_gen_265cce.py | ||
math_gen_943d32.py | ||
math_gen_0957ff.py | ||
math_gen_559593.py | ||
math_gen_736506.py | ||
math_gen.py | ||
math_intern_evaluator_gen_265cce.py | ||
math_llm_judge.py | ||
README.md |
MATH
python3 run.py --models hf_internlm2_7b --datasets math_4shot_base_gen_db136b --debug
python3 run.py --models hf_internlm2_chat_7b --datasets math_0shot_gen_393424 --debug
Base Models
model | math |
---|---|
llama-7b-turbomind | 2.94 |
llama-13b-turbomind | 3.84 |
llama-30b-turbomind | 6.54 |
llama-65b-turbomind | 10.66 |
llama-2-7b-turbomind | 3.58 |
llama-2-13b-turbomind | 5.30 |
llama-2-70b-turbomind | 13.26 |
llama-3-8b-turbomind | 16.42 |
llama-3-70b-turbomind | 39.64 |
internlm2-1.8b-turbomind | 9.42 |
internlm2-7b-turbomind | 25.16 |
internlm2-20b-turbomind | 32.24 |
qwen-1.8b-turbomind | 6.30 |
qwen-7b-turbomind | 15.56 |
qwen-14b-turbomind | 30.38 |
qwen-72b-turbomind | 44.18 |
qwen1.5-0.5b-hf | 4.16 |
qwen1.5-1.8b-hf | 11.32 |
qwen1.5-4b-hf | 17.50 |
qwen1.5-7b-hf | 17.34 |
qwen1.5-14b-hf | 36.18 |
qwen1.5-32b-hf | 45.74 |
qwen1.5-72b-hf | 41.56 |
qwen1.5-moe-a2-7b-hf | 27.96 |
mistral-7b-v0.1-hf | 13.44 |
mistral-7b-v0.2-hf | 12.74 |
mixtral-8x7b-v0.1-hf | 29.46 |
mixtral-8x22b-v0.1-hf | 41.82 |
yi-6b-hf | 6.60 |
yi-34b-hf | 18.80 |
deepseek-7b-base-hf | 4.66 |
deepseek-67b-base-hf | 18.76 |
Chat Models
model | math |
---|---|
qwen1.5-0.5b-chat-hf | 0.56 |
qwen1.5-1.8b-chat-hf | 4.94 |
qwen1.5-4b-chat-hf | 7.34 |
qwen1.5-7b-chat-hf | 22.14 |
qwen1.5-14b-chat-hf | 32.22 |
qwen1.5-32b-chat-hf | 41.80 |
qwen1.5-72b-chat-hf | 45.22 |
qwen1.5-110b-chat-hf | 54.38 |
internlm2-chat-1.8b-hf | 14.06 |
internlm2-chat-1.8b-sft-hf | 13.10 |
internlm2-chat-7b-hf | 28.08 |
internlm2-chat-7b-sft-hf | 27.60 |
internlm2-chat-20b-hf | 34.68 |
internlm2-chat-20b-sft-hf | 32.54 |
llama-3-8b-instruct-hf | 27.50 |
llama-3-70b-instruct-hf | 47.52 |
llama-3-8b-instruct-lmdeploy | 27.42 |
llama-3-70b-instruct-lmdeploy | 46.90 |
mistral-7b-instruct-v0.1-hf | 8.48 |
mistral-7b-instruct-v0.2-hf | 10.82 |
mixtral-8x7b-instruct-v0.1-hf | 27.02 |