Papers › OPT: Open Pre-trained Transformer Language Models

OPT: Open Pre-trained Transformer Language Models

2 May 2022arXiv:2205.01068archive 2025-07-28

Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen, Christopher Dewan, Mona Diab, Xian Li, Xi Victoria Lin, Todor Mihaylov, Myle Ott, Sam Shleifer, Kurt Shuster, Daniel Simig, Punit Singh Koura, Anjali Sridhar, Tianlu Wang, Luke Zettlemoyer

Large language models, which are often trained for hundreds of thousands of compute days, have shown remarkable capabilities for zero- and few-shot learning. Given their computational cost, these models are difficult to replicate without significant capital. For the few that are available through APIs, no access is granted to the full model weights, making them difficult to study. We present Open Pre-trained Transformers (OPT), a suite of decoder-only pre-trained transformers ranging from 125M to 175B parameters, which we aim to fully and responsibly share with interested researchers. We show that OPT-175B is comparable to GPT-3, while requiring only 1/7th the carbon footprint to develop. We are also releasing our logbook detailing the infrastructure challenges we faced, along with code for experimenting with all of the released models.

PaperPDFCodeCode Syntology ranMethod as code

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 24 code samples harvested from 5 repositories linked to this paper: 14 ran (of which 2 constructed an object rather than computing a result; 13 with no instrument failure: 0 honoured, 0 violated, 13 with no contract checked; 1 where Syntology's instrument failed) · 10 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 12 samples from 1 repository, 9 ran (8 with no instrument failure; 1 where Syntology's instrument failed); community (archive-listed): 12 samples from 4 repositories, 5 ran (5 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.

No sample is linked to this paper from 3 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

facebookresearch/metaseq officialmentioned in paperpytorchMITSyntology: 12 samples, 9 ran (1 instrument-failed) · 3 unverified report
ecolab-postech/owq mentioned on GitHubpytorchSyntology: 2 samples, 2 ran (0 instrument-failed) · 0 unverified report
liangyuwang/zo2 mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Fine-Tuning Language Models with Just Forward Passes (arXiv:2305.17333; that paper's own run record: 10 ran (of which 2 constructed an object rather than computing a result; 6 with no instrument failure: 1 honoured, 0 violated, 5 with no contract checked; 4 where Syntology's instrument failed) · 7 unverified); nothing checks that they implement this paper's method. report
mthcom/hscore-dataset-pruning mentioned on GitHubpytorchApache-2.0Syntology: 8 samples, 2 ran (0 instrument-failed) · 6 unverified report
pku-alignment/safe-rlhf mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Safe RLHF: Safe Reinforcement Learning from Human Feedback (arXiv:2310.12773; that paper's own run record: 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified) and ACL Anthology id 2024.acl-long.140 (that paper's own run record: 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
xvyaward/owq mentioned on GitHubpytorchSyntology: 1 sample, 0 ran · 1 unverified report
znhy1024/protoco mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
2023-MindSpore-1/ms-code-218 mindsporeSyntology: not harvested report
MindCode-4/code-2 mindsporeSyntology: not harvested report
xiuyu0000/new_papers_codes mindsporeSyntology: not harvested report
yangyucheng000/University mindsporeSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution (arXiv:2409.12191; that paper's own run record: 12 ran (of which 1 constructed an object rather than computing a result; 10 with no instrument failure: 4 honoured, 1 violated, 5 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. 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.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

24 samples harvested: 14 ran (of which 2 constructed an object rather than computing a result; 13 with no instrument failure: 0 honoured, 0 violated, 13 with no contract checked; 1 where Syntology's instrument failed) · 10 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 · our draft was wrong
13ran
10unverified

Licence: 17 of the 24 samples are pointer only, meaning Syntology does not serve that copy's text. For 1 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 5 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.

facebookresearch/metaseq official repository 12 samples, 9 ran (1 instrument-failed) · 3 unverified
checkpoint_paths metaseq/checkpoint_utils.py ran MIT recorded; this copy not marked cleared · pointer only · c570c23ad25ca1ab · report fixture: signature path:str/0/str, pattern:str/0/str · returned list · level recorded 2026-09-24
find_offsets in 7 papers metaseq/file_chunker_utils.py ran MIT recorded; this copy not marked cleared · pointer only · 840d9070e4fd8ba6 · report fixture: signature filename:str/0/str, num_chunks:int/0/int · returned list · level recorded 2026-09-24
gelu in 308 papers
census: 43 implementations from 74 papers, 6 distinct outputs on one shared input
metaseq/modules/activation_functions.py ran MIT recorded; this copy not marked cleared · pointer only · 04b65716e8591ed3 · report fixture: returned Tensor · level recorded 2026-09-24
gelu_back metaseq/modules/activation_functions.py ran MIT recorded; this copy not marked cleared · pointer only · 81272627e6280106 · report fixture: returned Tensor · level recorded 2026-09-24
get_next_token in 5 papers metaseq/hub_utils.py ran MIT recorded; this copy not marked cleared · pointer only · 70652fe4ded4564b · report fixture: signature logits:Tensor/3/float, tokenizer:StandInTokenizer/0/? · returned str · level recorded 2026-09-24
relu_squared metaseq/modules/activation_functions.py ran MIT recorded; this copy not marked cleared · pointer only · 684b43f1cd6f7e56 · report fixture: returned Tensor · level recorded 2026-09-24
tensorize_input metaseq/hub_utils.py ran MIT recorded; this copy not marked cleared · pointer only · e7743da916a7857a · report fixture: signature tokenizer:MockTokenizer/0/?, prompt:str/0/str · returned Tensor · level recorded 2026-09-24
url_to_filename in 73 papers metaseq/file_utils.py ran · our draft was wrong fingerprinted MIT recorded; this copy not marked cleared · pointer only · 9418b43048e4e4ec · report fixture: signature url:str/0/str, etag:str/0/str · returned str · level recorded 2026-09-24
with_incremental_state in 12 papers metaseq/incremental_decoding_utils.py ran MIT recorded; this copy not marked cleared · pointer only · 1dcaa0202f989ada · report fixture: signature cls:callable/0/? · returned type · level recorded 2026-09-24
cached_path in 47 papers metaseq/file_utils.py unverified MIT recorded; this copy not marked cleared · pointer only · fb80a6deb3bee2c5 · report
load_archive_file in 14 papers metaseq/file_utils.py unverified MIT recorded; this copy not marked cleared · pointer only · 0d372aa1e290a613 · report
setup_vocab_and_merges metaseq/hub_utils.py unverified MIT recorded; this copy not marked cleared · pointer only · 54309a919eca36b7 · report
ecolab-postech/owq community (archive-listed) 2 samples, 2 ran (0 instrument-failed) · 0 unverified
QuantLinear in 6 papers owq/quant.py ran · metamorphic tier: deterministic no licence file found · pointer only · be9074477838726e · report fixture: returned QuantLinear · level recorded 2026-09-22
QuantMatMul owq/quant.py ran no licence file found · pointer only · f5d0291b559e1a6a · report fixture: returned QuantMatMul · level recorded 2026-09-22
mthcom/hscore-dataset-pruning community (archive-listed) 8 samples, 2 ran (0 instrument-failed) · 6 unverified
flatten_outputs create_scores.py ran Apache-2.0 (permissive) · 8ce19774aeb3fda0 · report fixture: signature outputs:list/1/MockOutput, nb_labels:int/0/int, nb_epochs:int/0/int · returned Tensor · level recorded 2026-09-24
flatten_outputs create_scores_squad.py ran Apache-2.0 (permissive) · be1104875f8a36ce · report fixture: signature outputs:list/1/tuple, nb_epochs:int/0/int, nb_labels:int/0/int · returned Tensor · level recorded 2026-09-24
add_to_rewards create_scores.py unverified Apache-2.0 (permissive) · 00c32fbe5c2af279 · report
flatten_outputs create_scores_race.py unverified Apache-2.0 (permissive) · 84c5f4182ed8c57d · report
load_labels in 29 papers create_scores.py unverified Apache-2.0 (permissive) · 559cff9b7463d528 · report
load_labels in 29 papers create_scores_race.py unverified Apache-2.0 (permissive) · 9387fa891879e614 · report
postprocess_qa_predictions in 10 papers utils_qa.py unverified Apache-2.0 (permissive) · 6eff2c131107723d · report
postprocess_qa_predictions_with_beam_search in 10 papers utils_qa.py unverified Apache-2.0 (permissive) here; the stored copy is pointer only · 8c51f5aaf2f659e2 · report
xvyaward/owq community (archive-listed) 1 sample, 0 ran · 1 unverified
Quantizer in 10 papers owq/quant.py unverified no licence file found · pointer only · 352d10b844458470 · report
znhy1024/protoco community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
LoRALinear models/lora.py ran fingerprinted no licence file found · pointer only · e32e446737bb12b7 · report fixture: signature input:Tensor/2/float · returned Tensor · level recorded 2026-09-24

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for OPT (graph method: OPT). 1 brick; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

  1. LoRALinear entry · level 1 · ran · e32e446737bb12b7

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2205.01068") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

DecoderHate Speech DetectionLanguage ModellingStereotypical Bias Analysis

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Hate Speech Detection Ethos Binary OPT-175B (few-shot) F1-score 0.759 #4 of 12 Archive leaderboard report
Hate Speech Detection Ethos Binary OPT-175B (one-shot) F1-score 0.713 #6 of 12 Archive leaderboard report
Hate Speech Detection Ethos Binary OPT-175B (zero-shot) F1-score 0.667 #7 of 12 Archive leaderboard report
Hate Speech Detection Ethos Binary Davinci (zero-shot) F1-score 0.628 #10 of 12 Archive leaderboard report
Hate Speech Detection Ethos Binary Davinci (one-shot) F1-score 0.616 #11 of 12 Archive leaderboard report
Hate Speech Detection Ethos Binary Davinci (few-shot) F1-score 0.354 #12 of 12 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Age 64.4 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Disability 76.7 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Gender 62.6 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Nationality 61.6 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Overall 67.2 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Physical Appearance 74.6 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Race/Color 64.7 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Religion 62.6 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Sexual Orientation 76.2 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs GPT-3 Socioeconomic status 73.8 #2 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Age 67.8 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Disability 76.7 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Gender 65.7 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Nationality 62.9 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Overall 69.5 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Physical Appearance 76.2 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Race/Color 68.6 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Religion 65.7 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Sexual Orientation 78.6 #3 of 4 Archive leaderboard report
Stereotypical Bias Analysis CrowS-Pairs OPT-175B Socioeconomic status 76.2 #3 of 4 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Introduced by this paper: OPT

AdamAttentionAttention DropoutBPECosine AnnealingDense ConnectionsDropoutGPT-3Layer NormalizationLinear LayerLinear Warmup With Cosine AnnealingMulti-Head AttentionOPTResidual ConnectionSoftmaxWeight Decay

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