SMALLTOLARGE (S2L): Scalable Data Selection for Fine-tuning Large Language Models by Summarizing Training Loss Trajectories of Small Models. Yang, Y., Mishra, S., Chiang, J., & Mirzasoleiman, B. In The 38th Conference on Neural Information Processing Systems (NeurIPS 2024), Vancouver, Canada, December, 2024. doi abstract bibtex Despite the effectiveness of data selection for pretraining and instruction fine-tuninglarge language models (LLMs), improving data efficiency in supervised fine-tuning(SFT) for specialized domains poses significant challenges due to the complexityof fine-tuning data. To bridge this gap, we introduce an effective and scalabledata selection method for SFT, SmallToLarge (S2L), which trains a smallmodel, clusters loss trajectories of the examples, and samples from these clusters toguide data selection for larger models. We prove that during fine-tuning, sampleswithin the same loss trajectory cluster exhibit similar gradients. Then, we showthat S2L subsets have a bounded gradient error w.r.t. the full data, hence guaranteeconvergence to the neighborhood of the optimal solution. We demonstrate throughextensive experiments that S2L significantly improves data efficiency in SFT formathematical problem-solving, reducing the training data requirement to just 11%of the original MathInstruct dataset to match full dataset performance whileoutperforming state-of-the-art data selection algorithms by an average of 4.7%across 6 in- and out-domain evaluation datasets. Remarkably, selecting only 50Kdata for SFT, S2L achieves a 32.7% accuracy on the challenging MATHbenchmark, improving Phi-2 by 16.6%. In clinical text summarization on theMIMIC-III dataset, S2L again outperforms training on the full dataset usingonly 50% of the data. Notably, S2L can perform scalable data selection using areference model 100× smaller than the target model, proportionally reducing thecomputational cost.
@inproceedings{yang_smalltolarge_2024,
address = {Vancouver, Canada},
title = {{SMALLTOLARGE} ({S2L}): {Scalable} {Data} {Selection} for {Fine}-tuning {Large} {Language} {Models} by {Summarizing} {Training} {Loss} {Trajectories} of {Small} {Models}},
doi = {https://doi.org/10.52202/079017-2655},
abstract = {Despite the effectiveness of data selection for pretraining and instruction fine-tuninglarge language models (LLMs), improving data efficiency in supervised fine-tuning(SFT) for specialized domains poses significant challenges due to the complexityof fine-tuning data. To bridge this gap, we introduce an effective and scalabledata selection method for SFT, SmallToLarge (S2L), which trains a smallmodel, clusters loss trajectories of the examples, and samples from these clusters toguide data selection for larger models. We prove that during fine-tuning, sampleswithin the same loss trajectory cluster exhibit similar gradients. Then, we showthat S2L subsets have a bounded gradient error w.r.t. the full data, hence guaranteeconvergence to the neighborhood of the optimal solution. We demonstrate throughextensive experiments that S2L significantly improves data efficiency in SFT formathematical problem-solving, reducing the training data requirement to just 11\%of the original MathInstruct dataset to match full dataset performance whileoutperforming state-of-the-art data selection algorithms by an average of 4.7\%across 6 in- and out-domain evaluation datasets. Remarkably, selecting only 50Kdata for SFT, S2L achieves a 32.7\% accuracy on the challenging MATHbenchmark, improving Phi-2 by 16.6\%. In clinical text summarization on theMIMIC-III dataset, S2L again outperforms training on the full dataset usingonly 50\% of the data. Notably, S2L can perform scalable data selection using areference model 100× smaller than the target model, proportionally reducing thecomputational cost.},
language = {en},
booktitle = {The 38th {Conference} on {Neural} {Information} {Processing} {Systems} ({NeurIPS} 2024)},
author = {Yang, Yu and Mishra, Siddhartha and Chiang, Jeffrey and Mirzasoleiman, Baharan},
month = dec,
year = {2024},
keywords = {Foundational, SYS: CosmicAI Contact Author, WG: Explorable},
}
Downloads: 0
{"_id":"CYB2KpAec78wX2dun","bibbaseid":"yang-mishra-chiang-mirzasoleiman-smalltolarges2lscalabledataselectionforfinetuninglargelanguagemodelsbysummarizingtraininglosstrajectoriesofsmallmodels-2024","author_short":["Yang, Y.","Mishra, S.","Chiang, J.","Mirzasoleiman, B."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","address":"Vancouver, Canada","title":"SMALLTOLARGE (S2L): Scalable Data Selection for Fine-tuning Large Language Models by Summarizing Training Loss Trajectories of Small Models","doi":"https://doi.org/10.52202/079017-2655","abstract":"Despite the effectiveness of data selection for pretraining and instruction fine-tuninglarge language models (LLMs), improving data efficiency in supervised fine-tuning(SFT) for specialized domains poses significant challenges due to the complexityof fine-tuning data. To bridge this gap, we introduce an effective and scalabledata selection method for SFT, SmallToLarge (S2L), which trains a smallmodel, clusters loss trajectories of the examples, and samples from these clusters toguide data selection for larger models. We prove that during fine-tuning, sampleswithin the same loss trajectory cluster exhibit similar gradients. Then, we showthat S2L subsets have a bounded gradient error w.r.t. the full data, hence guaranteeconvergence to the neighborhood of the optimal solution. We demonstrate throughextensive experiments that S2L significantly improves data efficiency in SFT formathematical problem-solving, reducing the training data requirement to just 11%of the original MathInstruct dataset to match full dataset performance whileoutperforming state-of-the-art data selection algorithms by an average of 4.7%across 6 in- and out-domain evaluation datasets. Remarkably, selecting only 50Kdata for SFT, S2L achieves a 32.7% accuracy on the challenging MATHbenchmark, improving Phi-2 by 16.6%. In clinical text summarization on theMIMIC-III dataset, S2L again outperforms training on the full dataset usingonly 50% of the data. Notably, S2L can perform scalable data selection using areference model 100× smaller than the target model, proportionally reducing thecomputational cost.","language":"en","booktitle":"The 38th Conference on Neural Information Processing Systems (NeurIPS 2024)","author":[{"propositions":[],"lastnames":["Yang"],"firstnames":["Yu"],"suffixes":[]},{"propositions":[],"lastnames":["Mishra"],"firstnames":["Siddhartha"],"suffixes":[]},{"propositions":[],"lastnames":["Chiang"],"firstnames":["Jeffrey"],"suffixes":[]},{"propositions":[],"lastnames":["Mirzasoleiman"],"firstnames":["Baharan"],"suffixes":[]}],"month":"December","year":"2024","keywords":"Foundational, SYS: CosmicAI Contact Author, WG: Explorable","bibtex":"@inproceedings{yang_smalltolarge_2024,\n\taddress = {Vancouver, Canada},\n\ttitle = {{SMALLTOLARGE} ({S2L}): {Scalable} {Data} {Selection} for {Fine}-tuning {Large} {Language} {Models} by {Summarizing} {Training} {Loss} {Trajectories} of {Small} {Models}},\n\tdoi = {https://doi.org/10.52202/079017-2655},\n\tabstract = {Despite the effectiveness of data selection for pretraining and instruction fine-tuninglarge language models (LLMs), improving data efficiency in supervised fine-tuning(SFT) for specialized domains poses significant challenges due to the complexityof fine-tuning data. To bridge this gap, we introduce an effective and scalabledata selection method for SFT, SmallToLarge (S2L), which trains a smallmodel, clusters loss trajectories of the examples, and samples from these clusters toguide data selection for larger models. We prove that during fine-tuning, sampleswithin the same loss trajectory cluster exhibit similar gradients. Then, we showthat S2L subsets have a bounded gradient error w.r.t. the full data, hence guaranteeconvergence to the neighborhood of the optimal solution. We demonstrate throughextensive experiments that S2L significantly improves data efficiency in SFT formathematical problem-solving, reducing the training data requirement to just 11\\%of the original MathInstruct dataset to match full dataset performance whileoutperforming state-of-the-art data selection algorithms by an average of 4.7\\%across 6 in- and out-domain evaluation datasets. Remarkably, selecting only 50Kdata for SFT, S2L achieves a 32.7\\% accuracy on the challenging MATHbenchmark, improving Phi-2 by 16.6\\%. In clinical text summarization on theMIMIC-III dataset, S2L again outperforms training on the full dataset usingonly 50\\% of the data. Notably, S2L can perform scalable data selection using areference model 100× smaller than the target model, proportionally reducing thecomputational cost.},\n\tlanguage = {en},\n\tbooktitle = {The 38th {Conference} on {Neural} {Information} {Processing} {Systems} ({NeurIPS} 2024)},\n\tauthor = {Yang, Yu and Mishra, Siddhartha and Chiang, Jeffrey and Mirzasoleiman, Baharan},\n\tmonth = dec,\n\tyear = {2024},\n\tkeywords = {Foundational, SYS: CosmicAI Contact Author, WG: Explorable},\n}\n\n\n\n","author_short":["Yang, Y.","Mishra, S.","Chiang, J.","Mirzasoleiman, B."],"key":"yang_smalltolarge_2024","id":"yang_smalltolarge_2024","bibbaseid":"yang-mishra-chiang-mirzasoleiman-smalltolarges2lscalabledataselectionforfinetuninglargelanguagemodelsbysummarizingtraininglosstrajectoriesofsmallmodels-2024","role":"author","urls":{},"keyword":["Foundational","SYS: CosmicAI Contact Author","WG: Explorable"],"metadata":{"authorlinks":{}}},"bibtype":"inproceedings","biburl":"https://bibbase.org/zotero-group/pratikmhatre/5933976","dataSources":["yJr5AAtJ5Sz3Q4WT4"],"keywords":["foundational","sys: cosmicai contact author","wg: explorable"],"search_terms":["smalltolarge","s2l","scalable","data","selection","fine","tuning","large","language","models","summarizing","training","loss","trajectories","small","models","yang","mishra","chiang","mirzasoleiman"],"title":"SMALLTOLARGE (S2L): Scalable Data Selection for Fine-tuning Large Language Models by Summarizing Training Loss Trajectories of Small Models","year":2024}