Papers › VeLO: Training Versatile Learned Optimizers by Scaling Up

VeLO: Training Versatile Learned Optimizers by Scaling Up

17 Nov 2022arXiv:2211.09760archive 2025-07-28

Luke Metz, James Harrison, C. Daniel Freeman, Amil Merchant, Lucas Beyer, James Bradbury, Naman Agrawal, Ben Poole, Igor Mordatch, Adam Roberts, Jascha Sohl-Dickstein

While deep learning models have replaced hand-designed features across many domains, these models are still trained with hand-designed optimizers. In this work, we leverage the same scaling approach behind the success of deep learning to learn versatile optimizers. We train an optimizer for deep learning which is itself a small neural network that ingests gradients and outputs parameter updates. Meta-trained with approximately four thousand TPU-months of compute on a wide variety of optimization tasks, our optimizer not only exhibits compelling performance, but optimizes in interesting and unexpected ways. It requires no hyperparameter tuning, instead automatically adapting to the specifics of the problem being optimized. We open source our learned optimizer, meta-training code, the associated train and test data, and an extensive optimizer benchmark suite with baselines at velo-code.github.io.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

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

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 20 code samples harvested from 2 repositories linked to this paper: 15 ran (of which 0 constructed an object rather than computing a result; 15 with no instrument failure: 1 honoured, 1 violated, 13 with no contract checked; 0 where Syntology's instrument failed) · 5 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code; here there are none.

By repository: official repository: 11 samples from 1 repository, 8 ran (8 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 9 samples from 1 repository, 7 ran (7 with no instrument failure; 0 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-24 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

google/learned_optimization officialjaxSyntology: 11 samples, 8 ran (0 instrument-failed) · 3 unverified report
amoudgl/celo mentioned on GitHubjaxMITSyntology: 9 samples, 7 ran (0 instrument-failed) · 2 unverified 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

20 samples harvested: 15 ran (of which 0 constructed an object rather than computing a result; 15 with no instrument failure: 1 honoured, 1 violated, 13 with no contract checked; 0 where Syntology's instrument failed) · 5 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
1ran · violated contract
13ran
5unverified

Licence: 0 of the 20 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 2 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “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. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (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 are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. 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 six 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; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. 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.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

google/learned_optimization official repository 11 samples, 8 ran (0 instrument-failed) · 3 unverified
colors_for_num learned_optimization/notebook_utils.py ran Apache-2.0 (permissive) · dc03aa48546537e4 · report fixture: signature num:int/0/int, rep:int/0/int, stride:int/0/int, categorical:bool/0/bool · returned list · level recorded 2026-09-25
ema in 15 papers
census: 4 implementations from 7 papers, 2 distinct outputs across 2 buckets, one shared input per bucket
learned_optimization/notebook_utils.py ran Apache-2.0 (permissive) · e6c862d84c69f94c · report fixture: signature data:ndarray/1/float, alpha:float/0/float, ignore_nan:bool/0/bool · returned ndarray · level recorded 2026-09-25
file_open learned_optimization/filesystem.py ran Apache-2.0 (permissive) · 95b8c213917f58e2 · report fixture: signature path:str/0/str, mode:str/0/str · returned TextIOWrapper · level recorded 2026-09-25
maybe_do learned_optimization/jax_utils.py ran Apache-2.0 (permissive) · 84bfa52883247bc1 · report fixture: signature pred:int/0/int, do_fn:callable/0/?, operand:ArrayImpl/1/float · returned ArrayImpl · level recorded 2026-09-25
maybe_static_cond learned_optimization/jax_utils.py ran Apache-2.0 (permissive) · d2feb266da7afc46 · report fixture: signature pred:bool/0/bool, true_fn:callable/0/?, false_fn:callable/0/?, val:ArrayImpl/1/float · returned ArrayImpl · level recorded 2026-09-25
threaded_tqdm_map learned_optimization/notebook_utils.py ran Apache-2.0 (permissive) · 56f6fcb69aa5d0e8 · report fixture: signature threads:int/0/int, func:callable/0/?, data:list/1/Tensor · returned list · level recorded 2026-09-25
threaded_tqdm_map learned_optimization/py_utils.py ran Apache-2.0 (permissive) · 2a3363443caf8537 · report fixture: signature threads:int/0/int, func:callable/0/?, data:list/1/dict · returned list · level recorded 2026-09-25
uniquify_server_name learned_optimization/distributed.py ran fingerprinted Apache-2.0 (permissive) · c5f32a000fc8d76e · report fixture: signature server_name:str/0/str, experiment_name:str/0/str · returned str · level recorded 2026-09-25
cached_jit learned_optimization/jax_utils.py unverified Apache-2.0 (permissive) · 4a6a734cd7a980fb · report
copy learned_optimization/filesystem.py unverified Apache-2.0 (permissive) · 7b798c49a45c7bd4 · report
make_dirs learned_optimization/filesystem.py unverified Apache-2.0 (permissive) · acfa17a3ae480c74 · report
amoudgl/celo community (archive-listed) 9 samples, 7 ran (0 instrument-failed) · 2 unverified
decay_to_param celo/optimizers/adafac_mlp_lopt.py ran · violated contract fingerprinted MIT (permissive) · 519c4737cd569640 · report fixture: signature x:ArrayImpl/1/float · returned ArrayImpl · level recorded 2026-09-24
factored_dims celo/optimizers/celo.py ran fingerprinted MIT (permissive) · 048bb21eeedc76ed · report fixture: signature shape:tuple/1/int · returned tuple · level recorded 2026-09-25
factored_dims celo/optimizers/celo2_optax.py ran fingerprinted MIT (permissive) · f1a9b2bb9feb2fa3 · report fixture: signature shape:tuple/1/int · returned tuple · level recorded 2026-09-25
load_state in 19 papers celo/utils.py ran MIT (permissive) · d69eb755b5da1e5c · report fixture: signature path:str/0/str, state:dict/0/? · returned dict · level recorded 2026-09-25
orthogonalize_via_newton_schulz celo/optimizers/celo2_optax.py ran MIT (permissive) · 66d025f06569a161 · report fixture: signature x:ArrayImpl/2/float, ns_coeffs:ArrayImpl/1/float, ns_steps:int/0/int, eps:float/0/float · returned ArrayImpl · level recorded 2026-09-25
second_moment_normalizer celo/optimizers/adafac_mlp_lopt.py ran MIT (permissive) · d18fac6308869679 · report fixture: signature x:ArrayImpl/2/float, axis:int/0/int, eps:float/0/float · returned ArrayImpl · level recorded 2026-09-25
tanh_embedding celo/optimizers/adafac_mlp_lopt.py ran · honoured contract MIT (permissive) · 888ee6dfa73adb32 · report fixture: signature x:ArrayImpl/0/float · returned ArrayImpl · level recorded 2026-09-24
load_checkpoint in 277 papers celo/optimizers/celo2_optax.py unverified MIT (permissive) · ffce6a5291c41f40 · report
print_param_routing celo/optimizers/celo2.py unverified MIT (permissive) · cab026933643448f · report

Tasks

Deep Learning

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Test

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