Papers › HybridFlow: A Flexible and Efficient RLHF Framework

HybridFlow: A Flexible and Efficient RLHF Framework

28 Sep 2024arXiv:2409.19256archive 2025-07-28

Guangming Sheng, Chi Zhang, Zilingfeng Ye, Xibin Wu, Wang Zhang, Ru Zhang, Yanghua Peng, Haibin Lin, Chuan Wu

Reinforcement Learning from Human Feedback (RLHF) is widely used in Large Language Model (LLM) alignment. Traditional RL can be modeled as a dataflow, where each node represents computation of a neural network (NN) and each edge denotes data dependencies between the NNs. RLHF complicates the dataflow by expanding each node into a distributed LLM training or generation program, and each edge into a many-to-many multicast. Traditional RL frameworks execute the dataflow using a single controller to instruct both intra-node computation and inter-node communication, which can be inefficient in RLHF due to large control dispatch overhead for distributed intra-node computation. Existing RLHF systems adopt a multi-controller paradigm, which can be inflexible due to nesting distributed computation and data communication. We propose HybridFlow, which combines single-controller and multi-controller paradigms in a hybrid manner to enable flexible representation and efficient execution of the RLHF dataflow. We carefully design a set of hierarchical APIs that decouple and encapsulate computation and data dependencies in the complex RLHF dataflow, allowing efficient operation orchestration to implement RLHF algorithms and flexible mapping of the computation onto various devices. We further design a 3D-HybridEngine for efficient actor model resharding between training and generation phases, with zero memory redundancy and significantly reduced communication overhead. Our experimental results demonstrate 1.53×~20.57× throughput improvement when running various RLHF algorithms using HybridFlow, as compared with state-of-the-art baselines. HybridFlow source code will be available at https://github.com/volcengine/verl.

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 16 code samples harvested from 1 repository linked to this paper: 10 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 1 violated, 3 with no contract checked; 6 where Syntology's instrument failed) · 6 unverified. Where Syntology's instrument failed: 6 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 16 samples from 1 repository, 10 ran (4 with no instrument failure; 6 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 5 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

agentica-project/verl-pipeline mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Self-rewarding correction for mathematical reasoning (arXiv:2502.19613; that paper's own run record: 10 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 1 violated, 5 with no contract checked; 4 where Syntology's instrument failed) · 6 unverified); nothing checks that they implement this paper's method. report
du-nlp-lab/lengthreward mentioned on GitHubpytorchApache-2.0Syntology: 16 samples, 10 ran (6 instrument-failed) · 6 unverified report
hiyouga/easyr1 mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Scaling RL to Long Videos (arXiv:2507.07966; that paper's own run record: 18 ran (of which 1 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 1 violated, 6 with no contract checked; 11 where Syntology's instrument failed) · 6 unverified) and SaFeR-ToolKit: Structured Reasoning via Virtual Tool Calling for Multimodal Safety (arXiv:2603.02635; that paper's own run record: 9 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 7 where Syntology's instrument failed) · 6 unverified); nothing checks that they implement this paper's method. report
intelligent-internet/ii_verl mentioned on GitHubpytorchnot reachable when probed 2026-09-17 — repositories for recent papers often appear after camera-readySyntology: not harvested report
uscnlp-lime/verl mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Efficient Reinforcement Finetuning via Adaptive Curriculum Learning (arXiv:2504.05520; that paper's own run record: 9 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 1 violated, 3 with no contract checked; 4 where Syntology's instrument failed) · 6 unverified); nothing checks that they implement this paper's method. report
volcengine/verl mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Training Verifiers to Solve Math Word Problems (arXiv:2110.14168; that paper's own run record: 5 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified) and Stable Adaptive Thinking via Advantage Shaping and Length-Aware Gradient Regulation (arXiv:2602.22556; that paper's own run record: 2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
yaof20/real mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Flaming-hot Initiation with Regular Execution Sampling for Large Language Models (arXiv:2410.21236; that paper's own run record: 7 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 1 violated, 0 with no contract checked; 5 where Syntology's instrument failed) · 4 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

16 samples harvested: 10 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 1 violated, 3 with no contract checked; 6 where Syntology's instrument failed) · 6 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 · violated contract
6ran · our draft was wrong
3ran
6unverified

Licence: 6 of the 16 samples are pointer only, meaning Syntology does not serve that copy's text. For 6 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 du-nlp-lab/lengthreward. “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.

du-nlp-lab/lengthreward community (archive-listed) 16 samples, 10 ran (6 instrument-failed) · 6 unverified
compute_ce_dpo_loss_rm in 48 papers recipe/prime/prime_core_algos.py ran Apache-2.0 (permissive) here; the stored copy is pointer only · f220595adfc5e83c · report fixture: signature token_level_scores:Tensor/2/float, acc:Tensor/1/float, response_mask:Tensor/2/float, beta:float/0/float · returned Tensor · level recorded 2026-09-24
compute_detach_dpo_loss_rm in 49 papers recipe/prime/prime_core_algos.py ran · our draft was wrong Apache-2.0 (permissive) here; the stored copy is pointer only · 4d91e4383716e42c · report fixture: signature token_level_scores:Tensor/2/float, acc:Tensor/1/float, Q_bc:list/1/Tensor, acc_bc:Tensor/2/float, response_mask:Tensor/2/float, beta:float/0/float, bon_mode:str/0/str · returned Tensor · level recorded 2026-09-24
compute_online_dpo_loss in 39 papers recipe/spin/core_algos.py ran Apache-2.0 (permissive) here; the stored copy is pointer only · f19825983a523003 · report fixture: signature policy_chosen_logps:Tensor/1/float, policy_rejected_logps:Tensor/1/float, reference_chosen_logps:Tensor/1/float, reference_rejected_logps:Tensor/1/float, beta:float/0/float, label_smoothing:float/0/float, loss_type:str/0/str, reference_free:bool/0/bool · returned Tensor · level recorded 2026-09-24
compute_onlinedpo_pref in 39 papers recipe/spin/core_algos.py ran · our draft was wrong Apache-2.0 (permissive) here; the stored copy is pointer only · 41b6e81dc3ecba79 · report fixture: signature token_level_rewards:Tensor/2/float, response_mask:Tensor/2/float · returned Tensor · level recorded 2026-09-24
create_prompt_response recipe/char_count/create_dataset.py ran · our draft was wrong Apache-2.0 (permissive) · e685ec4d0fa29f37 · report fixture: signature min_length:int/0/int, max_length:int/0/int · returned tuple · level recorded 2026-09-25
get_kl_controller in 53 papers recipe/spin/core_algos.py ran · our draft was wrong Apache-2.0 (permissive) here; the stored copy is pointer only · 757d109d90121142 · report fixture: signature kl_ctrl:KLControlConfig/0/? · returned AdaptiveKLController · level recorded 2026-09-24
is_ipv4 in 42 papers verl/utils/net_utils.py ran · violated contract Apache-2.0 (permissive) · 735e93af30b15fa9 · report fixture: signature ip_str:str/0/str · returned bool · level recorded 2026-09-24
reward_func in 49 papers recipe/r1/reward_score.py ran Apache-2.0 (permissive) · c30b2a96103abd2c · report fixture: signature data_source:str/0/str, solution_str:str/0/str, ground_truth:str/0/str, extra_info:NoneType/0/? · returned MagicMock · level recorded 2026-09-24
union_tensor_dict in 69 papers verl/protocol.py ran · our draft was wrong Apache-2.0 (permissive) · 21331a58f93375e4 · report fixture: signature tensor_dict1:TensorDict/1/float, tensor_dict2:TensorDict/1/float · returned TensorDict · level recorded 2026-09-22
unpad_dataproto in 102 papers verl/protocol.py ran · our draft was wrong Apache-2.0 (permissive) · 83b1686d7ac03c21 · report fixture: signature data:DataProto/0/?, pad_size:int/0/int · returned DataProto · level recorded 2026-09-24
get_generation_config in 66 papers verl/utils/model.py unverified Apache-2.0 (permissive) · c9d58036da0286f2 · report
get_huggingface_actor_config in 83 papers verl/utils/model.py unverified Apache-2.0 (permissive) · 59904a8955b34252 · report
get_weight_loader in 93 papers verl/models/weight_loader_registry.py unverified Apache-2.0 (permissive) · acf240079d434746 · report
get_weight_saver in 69 papers verl/models/weight_loader_registry.py unverified Apache-2.0 (permissive) here; the stored copy is pointer only · cd6b1e0d6afb2762 · report
is_ipv6 in 42 papers verl/utils/net_utils.py unverified Apache-2.0 (permissive) · 705afe175ad8e404 · report
squeeze in 92 papers verl/utils/model.py unverified Apache-2.0 (permissive) · 3b15e2ac7497c441 · report

Tasks

Large Language Model

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

SET

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