{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/compute-constrained-data-selection","title":"Compute-Constrained Data Selection","arxiv_id":"2410.16208","date":"2024-10-21","proceeding":null,"authors":["Junjie Oscar Yin","Alexander M. Rush"],"abstract":"Data selection can reduce the amount of training data needed to finetune LLMs; however, the efficacy of data selection scales directly with its compute. Motivated by the practical challenge of compute-constrained finetuning, we consider the setting in which both the cost of selecting data and training are budgeted for. We first formalize the problem of data selection with a cost-aware utility function, and model the data selection problem as trading off initial-selection cost for training gain. We run a comprehensive sweep of experiments across multiple tasks, varying compute budget by scaling finetuning tokens, model sizes, and data selection compute. Interestingly we find that many powerful data selection methods are almost never compute-optimal, and that cheaper data selection alternatives dominate both from a theoretical and empirical perspective. For compute-optimal training, we find that perplexity and gradient data selection require training-to-selection model size ratios of 5x and 10x, respectively.","url_abs":"https://arxiv.org/abs/2410.16208v3","url_pdf":"https://arxiv.org/pdf/2410.16208v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"compute-constrained-data-selection","repo_url":"https://github.com/oseyosey/ccds","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"compute-constrained-data-selection","repo_url":"https://github.com/hamishivi/automated-instruction-selection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2410.16208","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2410.16208"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/oseyosey/CCDS","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/oseyosey/ccds","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hamishivi/automated-instruction-selection","reach":null}],"summary":{"ran_draft_wrong":1,"unverified":16},"by_repo_kind":{"official":{"samples":17,"ran":1,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"97229c97d4273595","entry":"encode_with_messages_format","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/finetune.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/finetune.py","link_basis":"plan_row","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"97229c97d4273595"}},{"code_sha256_prefix":"c50f39ac5b3486bb","entry":"concatenated_forward","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/dpo_utils.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/dpo_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"c50f39ac5b3486bb"}},{"code_sha256_prefix":"da4eb3f18cb89ad4","entry":"concatenated_inputs","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/dpo_utils.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/dpo_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"da4eb3f18cb89ad4"}},{"code_sha256_prefix":"e51c9713fadad16f","entry":"create_prompt_with_llama2_chat_format","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/eval/templates.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/eval/templates.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e51c9713fadad16f"}},{"code_sha256_prefix":"697d92a5d51eff3e","entry":"create_prompt_with_tulu_chat_format","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/eval/templates.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/eval/templates.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"697d92a5d51eff3e"}},{"code_sha256_prefix":"dfe736d72f7fb77f","entry":"dispatch_openai_chat_requesets","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/eval/dispatch_openai_requests.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/eval/dispatch_openai_requests.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"dfe736d72f7fb77f"}},{"code_sha256_prefix":"394a6cf6a2b821be","entry":"dispatch_openai_prompt_requesets","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/eval/dispatch_openai_requests.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/eval/dispatch_openai_requests.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"394a6cf6a2b821be"}},{"code_sha256_prefix":"48ff6fdc0e5dae2e","entry":"dpo_loss","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/dpo_utils.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/dpo_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"48ff6fdc0e5dae2e"}},{"code_sha256_prefix":"96351a65183bb428","entry":"encode_few_shot_example","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/reformat_datasets.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/reformat_datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"96351a65183bb428"}},{"code_sha256_prefix":"f839fd4c6f96a26f","entry":"encode_instruction_example","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/reformat_datasets.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/reformat_datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"f839fd4c6f96a26f"}},{"code_sha256_prefix":"254ab4972fe2c4e2","entry":"encode_with_messages_format","repo":"oseyosey/ccds","repo_kind":"official","path":"ccds/evaluation/open_instruct/dpo_tune.py","file_url":"https://github.com/oseyosey/ccds/blob/HEAD/ccds/evaluation/open_instruct/dpo_tune.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"254ab4972fe2c4e2"}},{"code_sha256_prefix":"5966cc3120c46d61","entry":"encode_with_prompt_completion_format","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/finetune.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/finetune.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"5966cc3120c46d61"}},{"code_sha256_prefix":"352416cd121404a7","entry":"generate_completions","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/eval/utils.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/eval/utils.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"352416cd121404a7"}},{"code_sha256_prefix":"d134c88af0236263","entry":"get_next_word_predictions","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/eval/utils.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/eval/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d134c88af0236263"}},{"code_sha256_prefix":"976ed9d65974a290","entry":"get_statistics_for_messages_data","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/get_statistics.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/get_statistics.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"976ed9d65974a290"}},{"code_sha256_prefix":"c6e64d6945c8512f","entry":"get_statistics_for_prompt_completion_data","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/get_statistics.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/get_statistics.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"c6e64d6945c8512f"}},{"code_sha256_prefix":"08f4816632095bc7","entry":"should_be_filtered","repo":"oseyosey/CCDS","repo_kind":"official","path":"ccds/evaluation/open_instruct/reformat_datasets.py","file_url":"https://github.com/oseyosey/CCDS/blob/HEAD/ccds/evaluation/open_instruct/reformat_datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"08f4816632095bc7"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}