Papers › No Train No Gain: Revisiting Efficient Training Algorithms For Transformer-based...

No Train No Gain: Revisiting Efficient Training Algorithms For Transformer-based Language Models

12 Jul 2023NeurIPS 2023 11arXiv:2307.06440archive 2025-07-28

Jean Kaddour, Oscar Key, Piotr Nawrot, Pasquale Minervini, Matt J. Kusner

The computation necessary for training Transformer-based language models has skyrocketed in recent years. This trend has motivated research on efficient training algorithms designed to improve training, validation, and downstream performance faster than standard training. In this work, we revisit three categories of such algorithms: dynamic architectures (layer stacking, layer dropping), batch selection (selective backprop, RHO loss), and efficient optimizers (Lion, Sophia). When pre-training BERT and T5 with a fixed computation budget using such methods, we find that their training, validation, and downstream gains vanish compared to a baseline with a fully-decayed learning rate. We define an evaluation protocol that enables computation to be done on arbitrary machines by mapping all computation time to a reference machine which we call reference system time. We discuss the limitations of our proposed protocol and release our code to encourage rigorous research in efficient training procedures: https://github.com/JeanKaddour/NoTrainNoGain.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2307.06440")

Code

Syntology Ran 10 of 13 code samples harvested from 1 repository linked to this paper; 3 have no recorded run. Of those that ran: 3 ran · violated contract; 7 ran with no contract checked.

By repository: official repository: 12 samples from 1 repository, 9 ran; 1 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

jeankaddour/notrainnogain officialmentioned in paperpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

13 samples harvested; 10 ran; 0 honoured the contract we drafted; 3 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

3ran · violated contract
7ran
3unverified

Licence: 13 of the 13 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from JeanKaddour/NoTrainNoGain. Some samples are identical code Syntology first harvested from another repository; for those, this paper's copy is not located and its licence is not recorded. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: 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. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

check_deadline JeanKaddour/NoTrainNoGain/bert/pretrain_bert_rho_loss.py official repository ran · violated contract no licence file found · pointer only · a7137af5385dbe09 · report
check_early_termination JeanKaddour/NoTrainNoGain/bert/pretrain_bert_rho_loss.py official repository ran · violated contract no licence file found · pointer only · 5bf52ae7f2f2317a · report
get_example_ids_from_batch JeanKaddour/NoTrainNoGain/bert/efficient_training/extract_il_losses.py official repository ran no licence file found · pointer only · 2840d349f4c8b0c1 · report
get_layer_fn JeanKaddour/NoTrainNoGain/bert/cramming/architectures/fused_layers.py official repository ran no licence file found · pointer only · 4681043cc92ad4c3 · report
layer_structure JeanKaddour/NoTrainNoGain/bert/cramming/architectures/fused_layers.py official repository ran no licence file found · pointer only · 98dec28a22e6c58a · report
load_chunks JeanKaddour/NoTrainNoGain/bert/pretrain_bert_rho_loss.py official repository ran no licence file found · pointer only · f048db3c79eb5ee8 · report
sample_active_layers JeanKaddour/NoTrainNoGain/bert/efficient_training/layer_drop.py official repository ran no licence file found · pointer only · 276a79af4663120e · report
scaled_layer_training JeanKaddour/NoTrainNoGain/bert/cramming/architectures/fused_layers.py official repository ran no licence file found · pointer only · 79efd868ef38d1ad · report
within_budget JeanKaddour/NoTrainNoGain/bert/pretrain_bert.py official repository ran no licence file found · pointer only · d782f0fce6dc8a6b · report
find_pretrained_checkpoint JeanKaddour/NoTrainNoGain/bert/cramming/utils.py official repository unverified no licence file found · pointer only · a054d17634dd4dfd · report
save_summary JeanKaddour/NoTrainNoGain/bert/cramming/utils.py official repository unverified no licence file found · pointer only · 02ca444386b6a0b5 · report
system_startup JeanKaddour/NoTrainNoGain/bert/cramming/utils.py official repository unverified no licence file found · pointer only · 871963e556958789 · report
exists identical code first harvested elsewhere ran · violated contract licence of this copy not recorded · aa5486a3650902d8 · report

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

AdafactorAdamAttentionAttention DropoutBERTBPEDense ConnectionsDropoutGated Linear UnitInverse Square Root ScheduleLayer NormalizationLinear LayerLinear Warmup With Linear DecayMulti-Head AttentionResidual ConnectionSentencePieceSoftmaxT5Weight DecayWordPiece

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections