Papers › AudioLDM 2: Learning Holistic Audio Generation with Self-supervised Pretraining

AudioLDM 2: Learning Holistic Audio Generation with Self-supervised Pretraining

10 Aug 2023arXiv:2308.05734archive 2025-07-28

Haohe Liu, Yi Yuan, Xubo Liu, Xinhao Mei, Qiuqiang Kong, Qiao Tian, Yuping Wang, Wenwu Wang, Yuxuan Wang, Mark D. Plumbley

Although audio generation shares commonalities across different types of audio, such as speech, music, and sound effects, designing models for each type requires careful consideration of specific objectives and biases that can significantly differ from those of other types. To bring us closer to a unified perspective of audio generation, this paper proposes a framework that utilizes the same learning method for speech, music, and sound effect generation. Our framework introduces a general representation of audio, called "language of audio" (LOA). Any audio can be translated into LOA based on AudioMAE, a self-supervised pre-trained representation learning model. In the generation process, we translate any modalities into LOA by using a GPT-2 model, and we perform self-supervised audio generation learning with a latent diffusion model conditioned on LOA. The proposed framework naturally brings advantages such as in-context learning abilities and reusable self-supervised pretrained AudioMAE and latent diffusion models. Experiments on the major benchmarks of text-to-audio, text-to-music, and text-to-speech demonstrate state-of-the-art or competitive performance against previous approaches. Our code, pretrained model, and demo are available at https://audioldm.github.io/audioldm2.

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 27 code samples harvested from 2 repositories linked to this paper: 16 ran (of which 0 constructed an object rather than computing a result; 13 with no instrument failure: 4 honoured, 4 violated, 5 with no contract checked; 3 where Syntology's instrument failed) · 11 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 13 samples from 1 repository, 8 ran (6 with no instrument failure; 2 where Syntology's instrument failed); community (archive-listed): 14 samples from 1 repository, 8 ran (7 with no instrument failure; 1 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 2026-09-22 and 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 16 samples that ran has no recorded date.

haoheliu/AudioLDM2 officialmentioned on GitHubpytorchNOASSERTIONSyntology: 13 samples, 8 ran (2 instrument-failed) · 5 unverified report
haoheliu/audioldm-training-finetuning mentioned on GitHubpytorchMITSyntology: 14 samples, 8 ran (1 instrument-failed) · 6 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

27 samples harvested: 16 ran (of which 0 constructed an object rather than computing a result; 13 with no instrument failure: 4 honoured, 4 violated, 5 with no contract checked; 3 where Syntology's instrument failed) · 11 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.

4ran · honoured contract
4ran · violated contract
3ran · our draft was wrong
5ran
11unverified

Licence: 19 of the 27 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 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.

haoheliu/AudioLDM2 official repository 13 samples, 8 ran (2 instrument-failed) · 5 unverified
get_bit_depth in 5 papers audioldm2/utils.py ran · honoured contract no licence file found · pointer only · 33e4f7a8bf80ee3a · report fixture: signature fname:str/0/str · returned int · level recorded 2026-09-24
get_duration in 6 papers audioldm2/utils.py ran · honoured contract no licence file found · pointer only · 159fc130dc64d332 · report fixture: signature fname:str/0/str · returned float · level recorded 2026-09-24
get_map in 5 papers audioldm2/clap/open_clip/loss.py ran · our draft was wrong fingerprinted no licence file found · pointer only · 8f1fbba0d901fc52 · report fixture: signature pred:Tensor/2/float, target:Tensor/2/float · returned float64 · level recorded 2026-09-24
get_padding in 84 papers audioldm2/hifigan/models.py ran · honoured contract fingerprinted no licence file found · pointer only · a26f85d7c72ef39a · report fixture: signature kernel_size:int/0/int, dilation:int/0/int · returned int · level recorded 2026-09-22
get_param_num in 10 papers audioldm2/utilities/model.py ran · honoured contract no licence file found · pointer only · f1cba67eb6d6e4ad · report fixture: signature model:callable/0/? · returned int · level recorded, date not recorded
read_list in 14 papers audioldm2/utils.py ran licence not identified · pointer only · b18f6b953126d4e6 · report fixture: signature fname:str/0/str · returned list · level recorded 2026-09-24
set_cond_text audioldm2/pipeline.py ran licence not identified · pointer only · 360696965b9c6f2a · report fixture: signature latent_diffusion:callable/0/? · returned MockLatentDiffusion · level recorded 2026-09-24
torch_version_orig_mod_remove audioldm2/utilities/model.py ran · our draft was wrong no licence file found · pointer only · 4dab43e0c6c8c170 · report fixture: signature state_dict:dict/0/? · returned dict · level recorded 2026-09-24
build_model_from_openai_state_dict in 42 papers audioldm2/clap/open_clip/model.py unverified licence not identified · pointer only · fde7529206b8c46f · report
gather_features in 78 papers audioldm2/clap/open_clip/loss.py unverified no licence file found · pointer only · 9b31c784540df0d7 · report
get_available_checkpoint_keys in 5 papers audioldm2/utilities/model.py unverified no licence file found · pointer only · 89bb9be7815a9834 · report
lp_gather_features audioldm2/clap/open_clip/loss.py unverified no licence file found · pointer only · 473061649d09d30c · report
trace_model in 12 papers audioldm2/clap/open_clip/model.py unverified no licence file found · pointer only · 6c018f9e8c31090c · report
haoheliu/audioldm-training-finetuning community (archive-listed) 14 samples, 8 ran (1 instrument-failed) · 6 unverified
calculate_relative_bandwidth audioldm_train/dataset_plugin.py ran MIT (permissive) · 22310b73059fa8cb · report fixture: signature config:dict/0/?, dl_output:dict/0/?, metadata:dict/0/? · returned dict · level recorded 2026-09-24
cosine_lr in 31 papers audioldm_train/modules/clap/training/scheduler.py ran MIT (permissive) here; the stored copy is pointer only · 932d44bbf8bd0e3f · report fixture: signature optimizer:SGD/0/?, base_lr:float/0/float, warmup_length:int/0/int, steps:int/0/int · returned function · level recorded 2026-09-26
get_default_params in 12 papers audioldm_train/modules/clap/training/params.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 575ba28efe232215 · report fixture: signature model_name:str/0/str · returned dict · level recorded 2026-09-25
int16_to_float32 in 8 papers audioldm_train/modules/clap/training/data.py ran · violated contract fingerprinted MIT (permissive) here; the stored copy is pointer only · 47bbdd629908bce3 · report fixture: signature x:ndarray/1/int · returned ndarray · level recorded 2026-09-25
is_global_master in 11 papers audioldm_train/modules/clap/training/distributed.py ran · violated contract MIT (permissive) here; the stored copy is pointer only · 7c29b374198bddcd · report fixture: signature args:Args/0/? · returned bool · level recorded 2026-09-22
is_local_master in 9 papers audioldm_train/modules/clap/training/distributed.py ran · violated contract MIT (permissive) here; the stored copy is pointer only · 0377e309e95cfc3e · report fixture: signature args:Args/0/? · returned bool · level recorded 2026-09-22
is_master in 43 papers audioldm_train/modules/clap/training/distributed.py ran · violated contract MIT (permissive) here; the stored copy is pointer only · b092791002b40e8a · report fixture: signature args:Args/0/? · returned bool · level recorded 2026-09-22
is_pretrained_params audioldm_train/modules/clap/training/lp_main.py ran MIT (permissive) · c7e729e9f89167e4 · report fixture: signature n:str/0/str · returned bool · level recorded 2026-09-24
get_vits_phoneme_ids audioldm_train/dataset_plugin.py unverified MIT (permissive) · 0186ae56bb268a36 · report
get_vits_phoneme_ids_no_padding audioldm_train/dataset_plugin.py unverified MIT (permissive) · ad984e4532566626 · report
run in 143 papers audioldm_train/modules/clap/training/zero_shot.py unverified MIT (permissive) · 7ffcc636c121640f · report
unwrap_model in 41 papers audioldm_train/modules/clap/training/lp_train.py unverified MIT (permissive) · 2bf62c0287050bdd · report
update_top_k_performance audioldm_train/modules/clap/training/lp_main.py unverified MIT (permissive) · 4ccd5f0108a2c3f2 · report

Tasks

Audio GenerationIn-Context LearningRepresentation LearningText to SpeechText-to-Music Generationtext-to-speech

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Audio Generation AudioCaps AudioLDM2-large CLAP_LAION 0.53 #8 of 23 Archive leaderboard report
Audio Generation AudioCaps AudioLDM2-large CLAP_MS 0.37 #8 of 23 Archive leaderboard report
Audio Generation AudioCaps AudioLDM2-large FAD 2.02 #8 of 23 Archive leaderboard report
Audio Generation AudioCaps AudioLDM2-large FD 26.18 #8 of 23 Archive leaderboard report
Audio Generation AudioCaps AudioLDM2-large FD_openl3 158.04 #8 of 23 Archive leaderboard report
Audio Generation AudioCaps AudioLDM2-large IS 8.55 #8 of 23 Archive leaderboard report
Audio Generation AudioCaps AudioLDM2-large KL_passt 1.68 #8 of 23 Archive leaderboard report
Audio Generation AudioCaps AudioLDM 2-AC-Large CLAP_LAION 0.243 #13 of 23 Archive leaderboard report
Audio Generation AudioCaps AudioLDM 2-AC-Large FAD 1.42 #13 of 23 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-large CLAP_LAION 0.48 #8 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-large CLAP_MS 0.47 #8 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-large FAD 2.93 #8 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-large FD 16.34 #8 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-large FD_openl3 190.16 #8 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-large IS 2.59 #8 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-large KL_passt 1.00 #8 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM 2-Full FAD 3.13 #9 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM 2-Full KL_passt 1.20 #9 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-music FD_openl3 354.05 #21 of 21 Archive leaderboard report
Text-to-Music Generation MusicCaps AudioLDM2-music KL_passt 1.53 #21 of 21 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

AdamAttentionAttention DropoutBPECosine AnnealingDense ConnectionsDiffusionDiscriminative Fine-TuningDropoutGPT-2Latent Diffusion ModelLayer NormalizationLinear LayerLinear Warmup With Cosine AnnealingMulti-Head AttentionResidual 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