wavlm-large
microsoft Not declared out 5 years ago
Text in, embeddings out.
Verified 2026-10-25 Rank #64 in Embeddings · score 85.8 method v1 Sources 3 Downloads 30d 854k · #125 of 619 tracked
Where it's used
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 2 quantized builds.
What people built on it
41 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
41 derivatives on the Hub, most of them fine-tunes
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Reference
Everything else, folded: the full tables, every source.
Use-case evidence
1 use case
What it is good at 1 use case
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Facts
11 fields
Listed in Embeddings
Facts
- License: Not declared (friction: review). No license declared on the model card. (license file, observed —)
- Created 2022-03-02, last modified 2022-02-02. (HF API, observed —)
| Pipeline tag | feature-extraction |
|---|---|
| Library | transformers |
| Architectures | WavLMModel |
| Model type | wavlm |
| MoE | no |
| Papers | arXiv:1912.07875, arXiv:2106.06909, arXiv:2101.00390, arXiv:2110.13900 |
| License | Not declared |
| Hub snapshot | 2026-10-25 |
All variants
40 variants
Lineage
finetune: 39quantized: 2
Top variants (40 shown, ranked by downloads)
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| Orange/Speaker-wavLM-tbr | finetune | — | 317M | S | cc-by-sa-3.0 | Orange | 16 | 279k |
| Orange/Speaker-wavLM-pro | finetune | — | 317M | S | cc-by-sa-3.0 | Orange | 4 | 174.9k |
| tiantiaf/wavlm-large-age-sex | finetune | — | 318M | S | openrail | tiantiaf | 11 | 64.5k |
| tiantiaf/wavlm-large-categorical-emotion | finetune | — | 316M | S | openrail | tiantiaf | 4 | 56.6k |
| tiantiaf/wavlm-large-voice-quality | finetune | — | 318M | S | openrail | tiantiaf | 4 | 54.1k |
| lucadellalib/focalcodec_50hz | finetune | — | 142M | S | apache-2.0 | lucadellalib | 2 | 49.3k |
| huper29/huper_recognizer | finetune | — | 316M | S | mit | huper29 | 8 | 37.6k |
| Orange/Speaker-wavLM-id | finetune | — | 317M | S | cc-by-sa-3.0 | Orange | 1 | 16.3k |
| tiantiaf/wavlm-large-msp-podcast-emotion-dim | finetune | — | 316M | S | openrail | tiantiaf | 5 | 10.3k |
| lucadellalib/focalcodec_12_5hz | finetune | — | 145M | S | apache-2.0 | lucadellalib | 3 | 9.1k |
| tiantiaf/wavlm-large-narrow-accent | finetune | — | 318M | S | openrail | tiantiaf | 1 | 8.1k |
| tiantiaf/wavlm-large-speech-flow | finetune | — | 318M | S | openrail | tiantiaf | 1 | 7.4k |
| tiantiaf/wavlm-large-broader-accent | finetune | — | 318M | S | openrail | tiantiaf | 2 | 6.6k |
| lucadellalib/focalcodec_50hz_4k_causal | finetune | — | 249M | S | apache-2.0 | lucadellalib | 1 | 6.4k |
| lucadellalib/focalcodec_50hz_2k_causal | finetune | — | 249M | S | apache-2.0 | lucadellalib | 0 | 5.5k |
| lucadellalib/focalcodec_25hz | finetune | — | 144M | S | apache-2.0 | lucadellalib | 1 | 4.5k |
| lucadellalib/focalcodec_50hz_65k_causal | finetune | — | 249M | S | apache-2.0 | lucadellalib | 0 | 1.4k |
| Xenova/wavlm-large | quantized | — | — | — | — | Xenova | 2 | 1.2k |
| eliya/forensics_0.3B_base_deepfake_classifier | finetune | — | — | — | cc-by-nc-4.0 | eliya | 8 | 770 |
| nyralabs/nyra_forced_aligner_en | finetune | — | — | — | other | nyralabs | 6 | 283 |
| imenba6/wavlm-large-fr-phoneme-recognition | finetune | — | 315M | S | mit | imenba6 | 0 | 537 |
| brdhaker3/TunASR | finetune | — | — | — | mit | brdhaker3 | 0 | 464 |
| lucadellalib/zipcodec | finetune | — | 842M | S | apache-2.0 | lucadellalib | 0 | 287 |
| drivendata/pasketti-phonetic | finetune | — | 316M | S | bigcode-openrail-m | drivendata | 2 | 153 |
| eliya/forensics_0.3B_wavlm_oc_softmax_deepfake_classifier | finetune | — | — | — | cc-by-nc-4.0 | eliya | 1 | 133 |
| Nikhil0097/wavlm-large-emotion-vad | finetune | — | 315M | S | cc-by-nc-sa-4.0 | Nikhil0097 | 1 | 122 |
| nyralabs/nyra_forced_aligner_en_pro | finetune | — | — | — | other | nyralabs | 3 | 0 |
| eliya/forensics_0.3B_v2_deepfake_age_gender_classifier | finetune | — | — | — | cc-by-nc-4.0 | eliya | 2 | 2 |
| Jianshu001/wavlm-phoneme-scorer | finetune | — | — | — | mit | Jianshu001 | 2 | 0 |
| pat229988/spoof-pad-wavlm-nes2net-tflite | finetune | — | — | — | other | pat229988 | 0 | 73 |
| onnx-community/wavlm-large-ONNX | quantized | — | — | — | — | onnx-community | 0 | 68 |
| Haopeng/CTC_for_IF-MDD | finetune | — | — | — | apache-2.0 | Haopeng | 1 | 0 |
| anonym-coling2027/E2E-GOP-SpeechOcean762 | finetune | — | 326M | S | mit | anonym-coling2027 | 0 | 43 |
| Haopeng/wavlm-large-s3t2ss-libri960-ce-segmentation | finetune | — | 317M | S | — | Haopeng | 0 | 29 |
| Allinsama/tiantiaf__wavlm-large-voice-quality | finetune | — | 318M | S | openrail | Allinsama | 0 | 17 |
| RogerTzeng/Sparse-MERIT | finetune | — | — | — | — | RogerTzeng | 0 | 0 |
| gwenn-ha-dev/wavlm-large-mlx | finetune | mlx | — | — | mit | gwenn-ha-dev | 0 | 0 |
| Jianshu001/wavlm-word-v13c2 | finetune | — | — | — | apache-2.0 | Jianshu001 | 0 | 0 |
| cyanwingsbird/EmpathyEval | finetune | — | — | — | mit | cyanwingsbird | 0 | 0 |
| arijitraj01/SER_WavLM_Fine-tuning | finetune | — | — | — | apache-2.0 | arijitraj01 | 0 | 0 |
Repository changelog
1 change
Articles and coverage
2 pieces
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency.
Analysis2
Papers
4 papers
Papers 4
arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).
- WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing this model peer-reviewed · measuredSanyuan Chen, Chengyi Wang, Zhengyang Chen +16 · 2021-10-26 · IEEE Journal of Selected Topics in Signal Processing · PDF · HF paper page (3 upvotes) · evidence
- GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audio cited in card peer-reviewed · measuredGuoguo Chen, Shuzhou Chai, Guanbo Wang +18 · 2021-06-13 · Interspeech (USB) · PDF · HF paper page (1 upvotes) · evidence
- Libri-Light: A Benchmark for ASR with Limited or No Supervision cited in card peer-reviewed · measuredJacob Kahn, Morgane Rivière, Weiyi Zheng +12 · 2019-12-17 · IEEE International Conference on Acoustics Speech and Signal Processing · PDF · HF paper page · evidence
Similar models
3 models
Similar models
- bge-large-zh-v1.5 same task (feature-extraction), more providers, permissive license
- codebert-base same task (feature-extraction), more providers
- bart-large same task (feature-extraction), more providers, permissive license
Sources
3 sources
Sources
Hub snapshot: 2026-10-25 · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:45 UTC · Summary: 2026-10-14 14:07 UTC · Last verified 2026-10-25
Follow wavlm-large
Watch this model
- /feeds/models/microsoft--wavlm-large.xml(RSS 2.0) · every change to wavlm-large
- /feeds/families/microsoft--wavlm-large.xml(RSS 2.0) · every change across the wavlm-large family
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.