We lifted 11 functions out of this paper's own repositories and ran 9 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| cyzhh/MMOS | canonical | 9 of 11 |
| Function | Status | Where it lives |
|---|---|---|
| encode_with_messages_format | Ran | cyzhh/MMOS/train/finetune.py pointer only (licence: NONE) · get_code("bc1a96ec28a11a5f") |
| encode_with_prompt_completion_format | Ran | cyzhh/MMOS/train/finetune.py pointer only (licence: NONE) · get_code("1906977026be9a86") |
| is_digit | Ran | cyzhh/MMOS/src/grader.py pointer only (licence: NONE) · get_code("8ae8c583b21633b3") |
| majority_vote | Ran | cyzhh/MMOS/eval/evaluate.py pointer only (licence: NONE) · get_code("c6427c534ee50108") |
| math_equal | Ran | cyzhh/MMOS/src/grader.py pointer only (licence: NONE) · get_code("65c738adb08e913c") |
| math_equal_process | Ran | cyzhh/MMOS/src/grader.py pointer only (licence: NONE) · get_code("f63c83de2a26f67c") |
| move_key_to_first | Ran | cyzhh/MMOS/src/rerank.py pointer only (licence: NONE) · get_code("38ab3477602c199f") |
| replace_var_names_in_code_blocks | Ran | cyzhh/MMOS/src/nodup.py pointer only (licence: NONE) · get_code("adff4d5b0a7ae14f") |
| show_example | Ran | cyzhh/MMOS/src/rerank.py pointer only (licence: NONE) · get_code("2c87889db2b36bfb") |
| build_paths | Not yet run | cyzhh/MMOS/src/utils.py pointer only (licence: NONE) · get_code("65de7160ff3afee9") |
| replace_var_names | Not yet run | cyzhh/MMOS/src/nodup.py pointer only (licence: NONE) · get_code("74f23325e7b85336") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Large language models (LLMs) are displaying emergent abilities for math reasoning tasks,and there is a growing attention on enhancing the ability of open-source LLMs through supervised fine-tuning (SFT).In this paper, we aim to explore a general data strategy for supervised data to help optimize and expand math reasoning ability.Firstly, we determine the ability boundary of reasoning paths augmentation by identifying these paths' minimal optimal set.Secondly, we validate that different abilities of the model can be cumulatively enhanced by Mix of Minimal Optimal Sets of corresponding types of data, while our models MMOS achieve SOTA performance on series base models under much lower construction costs.Besides, we point out GSM-HARD is not really hard and today's LLMs no longer lack numerical robustness.Also, we provide an Auto Problem Generator for robustness testing and educational applications.Our code and data are publicly available at https://github.com/cyzhh/MMOS.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2403.00799")
get_code_for_paper("2403.00799")
have("2403.00799")
Connect an agent — have() is free.