Pith. sign in

REVIEW 2 major objections 6 minor 2 cited by

Parameter Efficient Instruction Tuning: An Empirical Study

T0 review · 2 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper argues that only LoRA and adapters can substitute for full finetuning in instruction tuning, and then only with an ideal hyperparameter recipe.

desk verdict Useful empirical sweep supporting the 'LoRA/adapter can nearly match full fine-tuning' claim, but the LLAMA-2/TÜLU negative findings rest on transferred hyperparameters and an unsupported long-form claim. read the letter →

arxiv 2411.16775 v1 pith:ZMSVW6DH submitted 2024-11-25 cs.CL cs.AIcs.LG

Pengfei He This is my paper · ORCID
classification cs.CLcs.AIcs.LG
keywords parameter-efficientfinetuninginstructiontuningLoRAadaptershyperparametersearchcross-taskgeneralizationmemorizationlargelanguagemodels
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks when parameter-efficient finetuning (PEFT) can stand in for full finetuning when teaching large language models to follow instructions. The authors run a systematic comparison of five PEFT families under a wide grid of learning rates, LoRA ranks, adapter sizes, model sizes, and numbers of training tasks. Their central claim is that only LoRA and adapters come close to full finetuning, and they do so only under an ideal training setting: a matching learning rate, the largest permissible rank or bottleneck width, and a diverse set of instruction tasks. With that recipe, LoRA reaches 47.1 RougeL and the adapter 46.7 on SuperNI with T5-3B, against 47.8 for full finetuning. The paper also establishes that these methods are unstable away from the ideal setting, that LoRA is a slower learner needing more tasks to generalize, that LoRA has weaker task-level memorization, and that both fall short on complex reasoning, coding, and long-form generation in open instruction settings.

What carries the argument

The argument is carried by the contrast between two parameter-efficient mechanisms. LoRA freezes the pretrained weights and trains a low-rank update of rank $r$ applied additively to the query and key projections; the adapter inserts a trainable bottleneck layer of width $s$ between frozen transformer blocks. The study's method is a grid search that cross-products the learning rate $\eta$ with these PEFT sizes, alongside controlled variation of model size, number of training tasks, and train/test task overlap, scored by RougeL on SuperNI and by a broader suite on TÜLU. The key observed mechanism is a stability boundary: as $r$ or $s$ grows, the optimal learning rate narrows, and too high a learning rate drives training instability that can collapse RougeL to near zero.

What would settle it

Re-run the open instruction tuning setup with a hyperparameter search over learning rates from $10^{-5}$ to $3\times10^{-4}$ and LoRA ranks from 128 to 1024 on the TÜLU dataset; if a tuned LoRA still lags full finetuning on GSM8K, BBH, and Codex-Eval by the same margins, the paper's boundary holds, whereas if it matches, the boundary was set by the reused hyperparameters.

Watch

Extended reading notes

Core claim

The central discovery is an empirical boundary on when parameter-efficient tuning pays for instruction tuning. Among the methods tested, only LoRA and adapters come close to full finetuning, and they do so only when the learning rate is appropriate, the LoRA rank or adapter size is set to the largest value allowed, and the training tasks are diverse. Under those conditions, LoRA trails full finetuning by 0.7 RougeL points and the adapter by 1.1 points on the first benchmark. Away from the ideal setting, both methods show training instability, occasionally collapsing to near-zero RougeL, and the instability grows worse as rank or adapter size increases. The paper further reports that LoRA needs more tasks than adapter or full finetuning to generalize to unseen tasks, that LoRA memorizes task-level patterns less effectively, and that in open instruction tuning on a 7-billion-parameter model, both LoRA and adapter underperform full finetuning on reasoning, coding, and long-form generation, with LoRA ahead of adapter overall.

Load-bearing premise

The study assumes that the hyperparameters that work best on the smaller model and the first dataset transfer unchanged to the larger model and the second dataset; if they are suboptimal there, the reported shortfalls in reasoning and coding could be a tuning artifact rather than a real limitation of LoRA and adapters.

Editorial extensions

If this is right

  • Under the ideal setting, practitioners can substitute LoRA or an adapter for full finetuning while training only about 10 percent or 7 percent of the parameters, cutting memory and storage costs.
  • Finding the ideal setting requires searching the learning rate, because a high learning rate combined with a large rank or bottleneck width can make training collapse; the reported gains should not be read as rank-only improvements.
  • When only a handful of instruction tasks are available, full finetuning remains the better choice, with adapter as the safer PEFT fallback, while LoRA needs a large task pool to show its generalization strength.
  • For deployments that emphasize complex reasoning, coding, or long-form generation, full finetuning still outperforms both PEFT methods, and among the two, LoRA is the stronger choice overall.
  • LoRA's weaker task-level memorization means applications that expect to reuse a task type seen in training may lose accuracy with LoRA compared with adapter or full finetuning.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the hyperparameters from the first setup were re-tuned for the larger model instead of transferred unchanged, the reported reasoning and coding gaps might narrow; the paper's own limitations flag that its optimal settings may not reflect the latest architectures and datasets.
  • Because performance kept rising with rank up to the tested maximum of 512, ranks beyond that could close more of the remaining gap, although the observed instability and growing storage cost set a practical ceiling.
  • LoRA's weaker task-level memorization could be reframed as a privacy-friendly property in deployments where retaining task-specific training patterns is undesirable, complementing earlier work on instance-level memorization.
  • A direct next experiment would be a per-model learning-rate schedule that anneals from high to low during training to avoid the found instability while preserving the expressiveness of large ranks.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 6 minor

Summary. This paper presents an empirical study of parameter-efficient fine-tuning (PEFT) methods for instruction tuning. The authors evaluate five PEFT families (LoRA, adapter, prefix tuning, prompt tuning, BitFit) and full fine-tuning on T5-3B with the SuperNI dataset, searching over learning rates and PEFT-specific hyperparameters. They find that only LoRA and adapter approach full fine-tuning performance under what they call ideal training settings (large rank/adapter size, appropriate learning rate, diverse tasks), while also reporting training instability at high learning rates/ranks, weaker low-data generalization for LoRA, weaker task-level memorization, and underperformance on reasoning/coding/long-form generation in a second setup on LLAMA-2 7B with the TÜLU dataset.

Significance. If the central claim holds, the paper provides practical guidance for practitioners choosing between PEFT and full fine-tuning for instruction tuning, with the headline result that LoRA and adapter can nearly match full fine-tuning with favorable hyperparameters. The study has notable strengths: a broad hyperparameter grid, coverage of several PEFT families, two model/dataset settings, public benchmarks, and detailed appendix tables. The authors also honestly disclose limitations and reproducibility issues. However, the strength of the conclusions is limited by the lack of statistical uncertainty quantification and by the untested transfer of hyperparameters from the first setup to the second.

major comments (2)
  1. [Sec. 2.2, Table 3, Appendix A] The negative finding that LoRA and adapter underperform in complex reasoning, coding, and long-form generation (Table 3: FT 43.4 vs LoRA 38.78 vs adapter 35.26) rests on hyperparameters transferred unchanged from the T5/SuperNI setup: learning rate 1e-4, LoRA rank 512, adapter size 512, with no tuning or sensitivity analysis on LLAMA-2/TÜLU. Given that Table 4 shows LoRA's RougeL collapses from 47.1 at lr=1e-4 to 23.2 at 5e-4 and 0.0 at 1e-3 on T5, the gap in Table 3 could reflect a misplaced learning rate or rank for the larger model rather than an intrinsic limitation of PEFT. The abstract's conclusion that LoRA/adapter 'fall short' in these capabilities is load-bearing and needs support from either a hyperparameter search on Setup 2 or a small sensitivity analysis (e.g., a few learning rates and ranks) demonstrating that the gap persists across reasonable configurations.
  2. [Tables 1-3, Fig. 1a/3a, Appendix D] The quantitative claims lack statistical support. Table 1, Table 2, and Table 3 report single values without confidence intervals or error bars, and while Fig. 1a and Fig. 3a state that results are averaged over three seeds, no variance is shown. Appendix D further states that runs with unstable training that produce near-zero RougeL are included in the reported averages. This conflation of training divergence with typical performance directly affects Findings 3 and 5 (training instability and low-data generalization) and makes it difficult to judge whether the gap between LoRA/adapter and full fine-tuning in Table 1 is significant. Please report the number of runs per setting, provide variance or error bars for averaged results, and either exclude diverged runs or report them separately as a measure of instability.
minor comments (6)
  1. [Sec. 3.6] The text refers to '(Fig.??)' when discussing memorization results; the figure number is missing and should be Table 2 or a proper figure reference.
  2. [Sec. 4 (Related Work)] The statement 'we have selected seven PEFT method' does not match the five methods evaluated in the experiments; additionally, LoRA is cited as reference [10] (Lester et al., prompt tuning) rather than [8] (Hu et al., LoRA), and the citation for prompt tuning should be checked.
  3. [Fig. 2 caption] The caption says 'Each point represents a run with different random seeds,' which is ambiguous: does each point correspond to a single run (not averaged) or an average over seeds? Please clarify for Figs. 1b, 2a, 2b, and 3b.
  4. [Table 3] The 'Average' column averages five metrics with different scales (EM, F1, P@10) across heterogeneous capabilities; this aggregation is difficult to interpret and should be justified or replaced with a per-category comparison.
  5. [Sec. 2.1] The description says the data comprises 707 training tasks and 50 validation tasks 'all randomly selected from SuperNI original training tasks'; it would be clearer to state whether the 50 validation tasks are disjoint from the 707 training tasks and how they were selected.
  6. [Appendix C] The hyperparameter lists for Setup 1 do not mention LoRA alpha, which interacts with rank and learning rate; please state whether alpha was fixed or tuned, especially since alpha=512 is used in Setup 2.

Circularity Check

1 steps flagged · score 2.0 of 10

Mild post-hoc selection of 'ideal' hyperparameters on the same benchmark makes the headline closeness partly definitional, but the study is otherwise an empirical report without load-bearing self-citations.

  1. fitted input called prediction [Abstract; Sec 3.1, Table 1]
    "Our empirical study shows that only LoRA and adapter can get close to full finetuning with ideal training settings. ... Our extensive hyperparameter search reveals that among the five PEFTs, only LoRA and adapter are proved close to full finetuning in instruction tuning settings (See Table 1). ... For each PEFT method, best performance is reported as the best RougeL score on SuperNI across all hyperparameter combinations."

    The 'ideal training settings' are not independently defined; they are the hyperparameters that maximize RougeL on the same SuperNI test set used to report the headline numbers. The claim that LoRA and adapter 'can get close to full finetuning' is therefore a restatement of the best grid point. If any searched configuration is close, the reported maximum will be close by construction, so the conclusion is partially forced by the selection procedure. This is a mild selection-on-the-test-set issue rather than a derivation collapse, and it does not affect the comparative rankings across methods.

full rationale

The paper is an empirical study rather than a formal derivation, so most circularity patterns do not apply. There are no load-bearing self-citations and no imported uniqueness theorems. The central comparison between PEFT methods and full finetuning is measured directly on SuperNI and TÜLU, with full results reported in tables. The only notable circularity-adjacent issue is that the headline 'ideal training settings' are selected as the best-performing hyperparameters on the same SuperNI evaluation set, making the top-level claim conditional on post-hoc selection. This is standard practice in empirical tuning studies, but it means the specific numbers (e.g., LoRA 47.1 vs FT 47.8) are maxima over a grid rather than predictions from a fitted model. The transfer of these hyperparameters to LLAMA-2/TÜLU without retuning (Sec 2.2) is a correctness and generalizability risk, not a circularity, and the paper's own Limitations section acknowledges that the optimal hyperparameters 'might not reflect latest model architecture's performance.' Overall, the findings retain independent empirical content, so the circularity score is low.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new entities, particles, forces, or conserved quantities; it is an empirical comparison of existing PEFT methods.

free parameters (4)
  • LoRA rank = r=512
    Selected as best among {8,32,64,128,256,512} on SuperNI; used for T5 and LLAMA-2 experiments.
  • Adapter bottleneck size = s=512
    Selected as best among {8,32,64,128,256,512} on SuperNI; used for T5 and LLAMA-2 experiments.
  • Learning rate = 1e-4 for LoRA and adapter
    Selected as best from grid {1e-5,5e-5,1e-4,5e-4,1e-3} on SuperNI; applied to TÜLU experiments without re-tuning.
  • LoRA alpha = 512
    Set equal to LoRA rank for TÜLU experiments; not grid-searched, affects scaling of LoRA update.
assumptions (6)
  • domain assumption T5-3B and LLAMA-2 7B are representative base models for instruction tuning
    Used as the only base models in the two setups (Sec 2.1, 2.2); findings may not transfer to other architectures.
  • domain assumption SuperNI and TÜLU are representative instruction tuning benchmarks
    All conclusions are drawn from these two datasets (Sec 2.1, 2.2).
  • domain assumption RougeL is a valid metric for instruction following on SuperNI
    Evaluation uses RougeL only for the first setup (Sec 2.1).
  • ad hoc to paper Optimal hyperparameters transfer from T5/SuperNI to LLAMA-2/TÜLU
    Sec 2.2 uses the best hyperparameters from Setup 1 without re-tuning; the paper's Limitations note this may not hold.
  • domain assumption Task-level memorization is measured by held-out instances of training tasks
    Sec 3.6 defines memorization this way; different definitions could yield different results.
  • ad hoc to paper Failed runs with near-zero scores are included in averaged results
    Appendix D says unstable runs led to near-zero RougeL scores and reduced average scores; this choice affects low-data and instability findings.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Parameter Efficient Instruction Tuning: An Empirical Study." pith.science (2026). https://pith.science/paper/ZMSVW6DH

@misc{pith2026241116775,
  author       = {Pith},
  title        = {Pith review of: Parameter Efficient Instruction Tuning: An Empirical Study},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZMSVW6DH}},
  note         = {Machine review of arXiv:2411.16775}
}
read the original abstract

Instruction tuning has become an important step for finetuning pretrained language models to better follow human instructions and generalize on various tasks. Nowadays, pretrained language models become increasingly larger, and full parameter finetuning is overwhelmingly costly. Therefore, Parameter Efficient Finetuning (PEFT) has arisen as a cost-effective practice for instruction tuning because of significantly smaller computational, memory, and storage cost compared to full finetuning. Despite their widespread adaptations, the vast hyperparameter spaces, the number of PEFT methods, the different focus of instruction tuning capabilities make disentangling the impact of each aspect difficult. This study systematically investigates several representative PEFT methods, surveying the effect of hyperparameter choices including training hyperparameters and PEFT-specific hyperparameters, how different models sizes and the number of instruction tasks affect the performance, in-task-distribution memorization and open instruction following capability. Our empirical study shows that only LoRA and adapter can get close to full finetuning with ideal training settings. The ideal training setting includes an appropriate learning rate, largest LoRA rank or adapter size allowed and diverse training tasks. On the other hand, LoRA and adapter suffer from training instability if such an ideal training condition is not met. Additionally, LoRA requires a greater number of tasks for effective unseen task generalization, exhibit slower learning speed. Moreover, LoRA has weaker task-level memorization. Lastly, LoRA and adapter fall short in complex reasoning, coding and long-form generation compared to finetuning in open instruction tuning settings but it shows stronger capabilities compared to adapter.

Figures

Figures reproduced from arXiv: 2411.16775 by the authors.

Figure 1
Figure 1. The impact of LoRA ranks and adapter size on LLM performance [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Model size impact on performance. 3.5 Finding 5: LoRA underperforms in low-data setting Despite fully trained LoRA is close to or on par with full finetuning, larger data size always benefits LoRA across ranks as shown in Fig. 3b. On the other hand, in many industrial adaptations, there could only be a limited number of instruction tasks available. The important question is how many different instruction tasks requi… view at source ↗
Figure 3
Figure 3. Data size impact on LLM performance. 3.6 Finding 6: LoRA has worse task-level memorization Under the ideal downstream tuning setting, it is advantageous to provide training instruction tasks same as testing instruction tasks, and this resemble traditional training and test dataset split on a instance level. This raises a natural research question: "How do LoRA and adapter perform for test data which is in-distributi… view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. A Language-Guided Bayesian Optimization for Efficient LoRA Hyperparameter Search

    cs.CL 2026-01 conditional novelty 5.0 of 10

    LLM embeddings plus Bayesian optimization find better LoRA hyperparameters in ~30 proxy trials than standard published settings.

  2. Look Within or Look Beyond? A Theoretical Comparison Between Parameter-Efficient and Full Fine-Tuning

    cs.LG 2025-05 reject novelty 4.0 of 10

    The paper claims PEFT is a strict, less robust, lower-capacity subset of full fine-tuning, but the mathematical proofs contain load-bearing errors and the experiments, while suggestive, cannot repair them.

Reference graph

Works this paper leans on

30 extracted references · 23 canonical work pages · cited by 2 Pith papers

  1. [1]

    Biderman, J

    D. Biderman, J. G. Ortiz, J. Portes, M. Paul, P. Greengard, C. Jennings, D. King, S. Havens, V. Chiley, J. Frankle, C. Blakeney, and J. P. Cunningham. Lora learns less and forgets less, 2024. URL https://arxiv.org/abs/2405.09673

    arXiv 2024

  2. [2]

    Bommasani, D

    R. Bommasani, D. A. Hudson, E. Adeli, R. Altman, S. Arora, S. von Arx, M. S. Bernstein, J. Bohg, A. Bosselut, E. Brunskill, E. Brynjolfsson, S. Buch, D. Card, R. Castellon, N. Chat- terji, A. Chen, K. Creel, J. Q. Davis, D. Demszky, C. Donahue, M. Doumbouya, E. Durmus, S. Ermon, J. Etchemendy, K. Ethayarajh, L. Fei-Fei, C. Finn, T. Gale, L. Gillespie, K. ...

    work page 2022

  3. [3]

    Zhang, X.Shi, M

    J.Chen, A. Zhang, X.Shi, M. Li, A. Smola, andD. Yang. Parameter-efficientfine-tuning design spaces. In The Eleventh International Conference on Learning Representations, 2023. URL https://openreview.net/forum?id=XSRSWxyJIC

    work page 2023

  4. [4]

    D. Dai, L. Dong, Y. Hao, Z. Sui, B. Chang, and F. Wei. Knowledge neurons in pretrained transformers, 2022

    work page 2022

  5. [5]

    Dettmers, A

    T. Dettmers, A. Pagnoni, A. Holtzman, and L. Zettlemoyer. Qlora: Efficient finetuning of quantized llms, 2023

    work page 2023

  6. [6]

    Towardsaunifiedviewofparameter- efficient transfer learning, 2022

    J.He,C.Zhou,X.Ma,T.Berg-Kirkpatrick,andG.Neubig. Towardsaunifiedviewofparameter- efficient transfer learning, 2022

    work page 2022

  7. [7]

    Houlsby, A

    N. Houlsby, A. Giurgiu, S. Jastrzebski, B. Morrone, Q. de Laroussilhe, A. Gesmundo, M. At- tariyan, and S. Gelly. Parameter-efficient transfer learning for nlp, 2019

    work page 2019

  8. [8]

    E. J. Hu, Y. Shen, P. Wallis, Z. Allen-Zhu, Y. Li, S. Wang, L. Wang, and W. Chen. Lora: Low-rank adaptation of large language models, 2021

    2021

Show all 30 references
  1. [9]

    Ivison, Y

    H. Ivison, Y. Wang, V. Pyatkin, N. Lambert, M. Peters, P. Dasigi, J. Jang, D. Wadden, N. A. Smith, I. Beltagy, and H. Hajishirzi. Camels in a changing climate: Enhancing lm adaptation with tulu 2, 2023

    2023

  2. [10]

    Lester, R

    B. Lester, R. Al-Rfou, and N. Constant. The power of scale for parameter-efficient prompt tuning. In Conference on Empirical Methods in Natural Language Processing (EMNLP), 2021. URL https://arxiv.org/abs/2104.08691

    2021 arXiv

  3. [11]

    X. L. Li and P. Liang. Prefix-tuning: Optimizing continuous prompts for generation, 2021

    2021

  4. [12]

    Lialin, V

    V. Lialin, V. Deshpande, and A. Rumshisky. Scaling down to scale up: A guide to parameter- efficient fine-tuning. arXiv preprint arXiv:2303.15647, 2023

    2023 arXiv

  5. [13]

    C.-Y. Lin. ROUGE: A package for automatic evaluation of summaries. InText Summariza- tion Branches Out, pages 74–81, Barcelona, Spain, July 2004. Association for Computational Linguistics. URL https://aclanthology.org/W04-1013

    2004

  6. [14]

    Few-shotparameter- efficient fine-tuning is better and cheaper than in-context learning, 2022

    H.Liu,D.Tam,M.Muqeeth,J.Mohta,T.Huang,M.Bansal,andC.Raffel. Few-shotparameter- efficient fine-tuning is better and cheaper than in-context learning, 2022

    2022

  7. [15]

    Mireshghallah, A

    F. Mireshghallah, A. Uniyal, T. Wang, D. Evans, and T. Berg-Kirkpatrick. Memorization in nlp fine-tuning methods, 2022. 8

    2022

  8. [16]

    Ouyang, J

    L. Ouyang, J. Wu, X. Jiang, D. Almeida, C. L. Wainwright, P. Mishkin, C. Zhang, S. Agarwal, K.Slama,A.Ray,J.Schulman,J.Hilton,F.Kelton,L.Miller,M.Simens,A.Askell,P.Welinder, P. Christiano, J. Leike, and R. Lowe. Training language models to follow instructions with human feedback, 2022

    2022

  9. [17]

    Petrov, P

    A. Petrov, P. H. S. Torr, and A. Bibi. When do prompting and prefix-tuning work? a theory of capabilities and limitations, 2023

    2023

  10. [18]

    Raffel, N

    C. Raffel, N. Shazeer, A. Roberts, K. Lee, S. Narang, M. Matena, Y. Zhou, W. Li, and P. J. Liu. Exploring the limits of transfer learning with a unified text-to-text transformer.Journal of Machine Learning Research (JMLR), 2020. URLhttps://arxiv.org/abs/1910.10683

    2020 arXiv

  11. [19]

    Taori, I

    R. Taori, I. Gulrajani, T. Zhang, Y. Dubois, X. Li, C. Guestrin, P. Liang, and T. B. Hashimoto. Stanford alpaca: An instruction-following llama model. GitHub repository, 2023. URLhttps: //github.com/tatsu-lab/stanford_alpaca

    2023

  12. [20]

    Touvron, L

    H. Touvron, L. Martin, K. Stone, P. Albert, A. Almahairi, Y. Babaei, N. Bashlykov, S. Batra, P. Bhargava, S. Bhosale, D. Bikel, L. Blecher, C. C. Ferrer, M. Chen, G. Cucurull, D. Esiobu, J. Fernandes, J. Fu, W. Fu, B. Fuller, C. Gao, V. Goswami, N. Goyal, A. Hartshorn, S. Hoss...

    2023

  13. [21]

    Y. Wang, Y. Kordi, S. Mishra, A. Liu, N. A. Smith, D. Khashabi, and H. Hajishirzi. Self- instruct: Aligning language models with self-generated instructions. InAnnual Meeting of the Association for Computational Linguistics, 2022. URLhttps://api.semanticscholar. org/CorpusID:254877310

    2022

  14. [22]

    Y. Wang, S. Mishra, P. Alipoormolabashi, Y. Kordi, A. Mirzaei, A. Arunkumar, A. Ashok, A. S. Dhanasekaran, A. Naik, D. Stap, et al. Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ Tasks. InEMNLP, 2022

    2022

  15. [23]

    Y. Wang, H. Ivison, P. Dasigi, J. Hessel, T. Khot, K. R. Chandu, D. Wadden, K. MacMillan, N. A. Smith, I. Beltagy, and H. Hajishirzi. How far can camels go? exploring the state of instruction tuning on open resources, 2023

    2023

  16. [24]

    Self-instruct: Aligning language models with self-generated instructions, 2023

    Y.Wang,Y.Kordi,S.Mishra,A.Liu,N.A.Smith,D.Khashabi,andH.Hajishirzi. Self-instruct: Aligning language models with self-generated instructions, 2023. URLhttps://arxiv.org/ abs/2212.10560

    2023 arXiv

  17. [25]

    J. Wei, M. Bosma, V. Y. Zhao, K. Guu, A. W. Yu, B. Lester, N. Du, A. M. Dai, and Q. V. Le. Finetuned language models are zero-shot learners, 2022

    2022

  18. [26]

    E. B. Zaken, S. Ravfogel, and Y. Goldberg. Bitfit: Simple parameter-efficient fine-tuning for transformer-based masked language-models, 2022

    2022

  19. [27]

    Zhang, L

    S. Zhang, L. Dong, X. Li, S. Zhang, X. Sun, S. Wang, J. Li, R. Hu, T. Zhang, F. Wu, and G. Wang. Instruction tuning for large language models: A survey, 2023

    2023

  20. [28]

    Lewis, L

    C.Zhou,P.Liu,P.Xu,S.Iyer,J.Sun,Y.Mao,X.Ma,A.Efrat,P.Yu,L.Yu,S.Zhang,G.Ghosh, M. Lewis, L. Zettlemoyer, and O. Levy. Lima: Less is more for alignment, 2023

    2023

  21. [29]

    Astraios: Parameter-efficient instruction tuning code large language models, 2024

    T.Y.Zhuo,A.Zebaze,N.Suppattarachai,L.vonWerra,H.deVries,Q.Liu,andN.Muennighoff. Astraios: Parameter-efficient instruction tuning code large language models, 2024. URLhttps: //arxiv.org/abs/2401.00788

    2024 arXiv

  22. [30]

    D. M. Ziegler, N. Stiennon, J. Wu, T. B. Brown, A. Radford, D. Amodei, P. Christiano, and G. Irving. Fine-tuning language models from human preferences, 2020. 9 Supplementary Material Acknowledgments and Disclosure of Funding We thank our founding advisor Jieyu Zhang at AdaBit...

    2020

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.