{"id":22409839,"url":"https://github.com/squeezeailab/llm2llm","last_synced_at":"2025-07-31T20:31:19.822Z","repository":{"id":229567326,"uuid":"775366361","full_name":"SqueezeAILab/LLM2LLM","owner":"SqueezeAILab","description":"[ACL 2024] LLM2LLM: Boosting LLMs with Novel Iterative Data Enhancement","archived":false,"fork":false,"pushed_at":"2024-03-25T05:33:34.000Z","size":214,"stargazers_count":151,"open_issues_count":2,"forks_count":10,"subscribers_count":6,"default_branch":"main","last_synced_at":"2024-10-18T23:12:41.052Z","etag":null,"topics":["data-augmentation","llama","llama2","llm","llms","natural-language-processing","nlp","synthetic-dataset-generation","transformer"],"latest_commit_sha":null,"homepage":"https://arxiv.org/abs/2403.15042","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/SqueezeAILab.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2024-03-21T08:59:17.000Z","updated_at":"2024-10-14T15:35:12.000Z","dependencies_parsed_at":null,"dependency_job_id":"ab9f3519-2634-40e7-afa2-ba9b2dc85149","html_url":"https://github.com/SqueezeAILab/LLM2LLM","commit_stats":{"total_commits":3,"total_committers":2,"mean_commits":1.5,"dds":"0.33333333333333337","last_synced_commit":"bcbcb1ae767b9d2f4565d44295e08d10670a8c1d"},"previous_names":["squeezeailab/llm2llm"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SqueezeAILab%2FLLM2LLM","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SqueezeAILab%2FLLM2LLM/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SqueezeAILab%2FLLM2LLM/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SqueezeAILab%2FLLM2LLM/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/SqueezeAILab","download_url":"https://codeload.github.com/SqueezeAILab/LLM2LLM/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":228295577,"owners_count":17897596,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["data-augmentation","llama","llama2","llm","llms","natural-language-processing","nlp","synthetic-dataset-generation","transformer"],"created_at":"2024-12-05T12:10:05.770Z","updated_at":"2024-12-05T12:10:06.327Z","avatar_url":"https://github.com/SqueezeAILab.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# LLM2LLM: Boosting LLMs with Novel Iterative Data Enhancement [[Paper](https://arxiv.org/abs/2403.15042)]\n\n![Thumbnail](figs/llm2llm.png)\n\nThis is the code for the LLM2LLM paper.\n\n## Reproducing Main Experiments\n\nWe have provided code required to reproduce our main experiments for GSM8K. Instructions for other datasets will be uploaded soon.\n\n1. Download a copy of LLaMA-2-7B, and the appropriate dataset\n2. Clone the GSM8K dataset by running\n```\ncd GSM8K\ngit clone https://github.com/openai/grade-school-math.git\n```\n3. Run `generate_seed_data.py` and adjust `SUBSAMPLE_SPLIT` to get seed data.\n4. Ensure that all settings in `config.yaml` are accurate\n5. Run `python GSM8K/generator_data.py GSM8K/config.yaml`\n6. `cd` into your experiment folder and run `./run_all.sh`\n7. After all of the iterations have finished, run \n```\npython report_results.py --results_file_name test_0.jsonl GSM8K/grade-school-math/grade_school_math/data/test.jsonl $EXP_FOLDER\n```\nto get a detailed breakdown of the performance of the model at each iteration.\n\nThis will produce an output folder that contains all the data and model checkpoints.\n\n## Roadmap\n\nWe are planning on adding the code required to reproduce our experiments on other datasets.\n\n## Citation\n\nLLM2LLM has been developed as part of the following paper. We would appreciate if you would please cite this paper if you found this library useful for your work:\n\n```\n@article{lee2024llm2llm,\n      title={LLM2LLM: Boosting LLMs with Novel Iterative Data Enhancement}, \n      author={Lee, Nicholas and Wattanawong, Thanakul and Kim, Sehoon and Mangalam, Karttikeya and Shen, Sheng and Anumanchipali, Gopala and Mahoney, Michael W and Keutzer, Kurt and Gholami, Amir},\n      journel={arXiv},\n      year={2024},\n}\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsqueezeailab%2Fllm2llm","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsqueezeailab%2Fllm2llm","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsqueezeailab%2Fllm2llm/lists"}