# wav2vec2-indonesian-javanese-sundanese by Indonesian NLP Source: https://savrn.com/models/wav2vec2-indonesian-javanese-sundanese Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## SAVRN's Notes on wav2vec2-indonesian-javanese-sundanese Three languages, one checkpoint: Indonesian, Javanese and Sundanese ASR from the indonesian-nlp group, trained on Common Voice 7.0 alongside magic_data, OpenSLR and TITML corpora. Page-reported Common Voice numbers are strong — 4.056 WER on CV 6.1, 4.492 on CV 7 — for 3.5 GB of weights inside a 23-file repository that runs on any spare GPU. The honest tradeoff sits beside them: 48.94 WER on the Robust Speech Event dev set and 68.95 on its test data. That spread — clean-read numbers under five, noisy-read numbers near fifty and sixty-nine — marks where this model lives: matched recording conditions like broadcast or studio audio, not far-field or degraded sources. Apache 2.0 permits commercial transcription with the usual notices. No parameter count is published, files date to August 2022, and no host prices it per token. The Russian and Polish wav2vec2 fine-tunes on this list carry the same warning. SAVRN Research ## Model Card By Indonesian NLP, published under apache-2.0, revision e5e699fa5aa5. ### Multilingual Speech Recognition for Indonesian Languages This is the model built for the project [Multilingual Speech Recognition for Indonesian Languages](https://github.com/indonesian-nlp/multilingual-asr). It is a fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) model on the [Indonesian Common Voice dataset](https://huggingface.co/datasets/common_voice), [High-quality TTS data for Javanese - SLR41](https://huggingface.co/datasets/openslr), and [High-quality TTS data for Sundanese - SLR44](https://huggingface.co/datasets/openslr) datasets. We also provide a [live demo](https://huggingface.co/spaces/indonesian-nlp/multilingual-asr) to test the model. When using this model, make sure that your speech input is sampled at 16kHz. ### Usage The model can be used directly (without a language model) as follows: [Read the full model card (339 words)](https://savrn.com/models/wav2vec2-indonesian-javanese-sundanese/card) ## Configuration Architecture Wav2Vec2ForCTC Layers 24 Hidden size 1,024 Feed-forward size 4,096 Attention heads 16 Vocabulary size 30 Stored precision float32 Model type wav2vec2 ## Identity and Version Repository indonesian-nlp/wav2vec2-indonesian-javanese-sundanese Publisher Indonesian NLP Task Speech recognition Modality Audio Library transformers Parameters Not stated by the source Languages id, jv, sun Revision e5e699fa5aa5bdce999276a90a29589587b58ac9 First published 2022-03-02 Last updated 2022-08-19 ## Files and Weights 23 files, 3.5 GB in total. The weights are 2 files totalling 3.5 GB in bin. Weights2 files · 3.5 GB Configuration7 files · 7.3 KB Tokenizer2 files · 795 B Documentation1 file · 5.6 KB Other10 files · 5.3 MB Repository1 file · 1.2 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | language_model/4gram.bin | Weights | 2.2 GB | c2b756cefd29 | | pytorch_model.bin | Weights | 1.3 GB | ea90feff05f7 | | added_tokens.json | Configuration | 3 B | — | | alphabet.json | Configuration | 189 B | — | | config.json | Configuration | 1.9 KB | — | | eval.py | Configuration | 4.8 KB | — | | language_model/attrs.json | Configuration | 78 B | — | | preprocessor_config.json | Configuration | 262 B | — | | special_tokens_map.json | Configuration | 83 B | — | | README.md | Documentation | 5.6 KB | — | | common_voice_id_test_eval_results.txt | Other | 50 B | — | | language_model/unigrams.txt | Other | 4.6 MB | — | | log_common_voice_id_test_predictions.txt | Other | 76.9 KB | — | | log_common_voice_id_test_targets.txt | Other | 77.0 KB | — | | log_mozilla-foundation_common_voice_7_0_id_test_predictions.txt | Other | 129.3 KB | — | | log_mozilla-foundation_common_voice_7_0_id_test_targets.txt | Other | 129.5 KB | — | | log_mozilla-foundation_common_voice_8_0_id_test_predictions.txt | Other | 155.3 KB | — | | log_mozilla-foundation_common_voice_8_0_id_test_targets.txt | Other | 155.5 KB | — | | mozilla-foundation_common_voice_7_0_id_test_eval_results.txt | Other | 50 B | — | | mozilla-foundation_common_voice_8_0_id_test_eval_results.txt | Other | 50 B | — | | .gitattributes | Repository | 1.2 KB | — | | tokenizer_config.json | Tokenizer | 526 B | — | | vocab.json | Tokenizer | 269 B | — | ## License and Download License apache-2.0 Access Open weights, no gate Download size 3.5 GB [Download from Indonesian NLP](https://huggingface.co/indonesian-nlp/wav2vec2-indonesian-javanese-sundanese) Released by Indonesian NLP through its official repository on Hugging Face. [Read the license](https://www.apache.org/licenses/LICENSE-2.0). ## Built From - Trained on (disclosed) magic_data - Trained on (disclosed) mozilla-foundation/common_voice_7_0 - Trained on (disclosed) openslr - Trained on (disclosed) titml ## Evaluations Each result is shown as reported, with the conditions its reporter stated. None is a SAVRN measurement. A comparison lines two results up only when their configuration, unit and setup are all stated and identical. | Benchmark | Conditions | Result | Reported by | Revision | Date | | --- | --- | --- | --- | --- | --- | | Common Voice 6.1 | Task Automatic Speech RecognitionMetric Test CERComparison conditions not established | 1.472 | [indonesian-nlp](https://huggingface.co/indonesian-nlp/wav2vec2-indonesian-javanese-sundanese) Publisher reported | Evaluated revision not stated | — | | Common Voice 6.1 | Task Automatic Speech RecognitionMetric Test WERComparison conditions not established | 4.056 | [indonesian-nlp](https://huggingface.co/indonesian-nlp/wav2vec2-indonesian-javanese-sundanese) Publisher reported | Evaluated revision not stated | — | | Common Voice 7 | Task Automatic Speech RecognitionMetric Test CERComparison conditions not established | 1.577 | [indonesian-nlp](https://huggingface.co/indonesian-nlp/wav2vec2-indonesian-javanese-sundanese) Publisher reported | Evaluated revision not stated | — | | Common Voice 7 | Task Automatic Speech RecognitionMetric Test WERComparison conditions not established | 4.492 | [indonesian-nlp](https://huggingface.co/indonesian-nlp/wav2vec2-indonesian-javanese-sundanese) Publisher reported | Evaluated revision not stated | — | | Robust Speech Event - Dev Data | Task Automatic Speech RecognitionMetric Test WERComparison conditions not established | 48.94 | [indonesian-nlp](https://huggingface.co/indonesian-nlp/wav2vec2-indonesian-javanese-sundanese) Publisher reported | Evaluated revision not stated | — | | Robust Speech Event - Test Data | Task Automatic Speech RecognitionMetric Test WERComparison conditions not established | 68.95 | [indonesian-nlp](https://huggingface.co/indonesian-nlp/wav2vec2-indonesian-javanese-sundanese) Publisher reported | Evaluated revision not stated | — | ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 3.5 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About wav2vec2-indonesian-javanese-sundanese ### Can I use wav2vec2-indonesian-javanese-sundanese commercially? Yes. wav2vec2-indonesian-javanese-sundanese is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors. ## Similar Models Model · Speech recognition ### [wav2vec2-large-xlsr-53-japanese](https://savrn.com/models/wav2vec2-large-xlsr-53-japanese) [Jonatas Grosman](https://savrn.com/model-publishers/jonatasgrosman) Fine-tuned facebook/wav2vec2-large-xlsr-53 on Japanese using the train and validation splits of Common Voice 6.1, CSS10 and JSUT. When using this model, make sure that your speech input is sampled at 16kHz. This model has been fine-tuned thanks to the GPU credits generously given by the OVHcloud:) The script used for training can be found here: https://github.com/jonatasgrosman/wav2vec2-sprint The model can be used directly (without a language model) as follows... Using the HuggingSound library: The model can be evaluated as follows on the Japanese test data of Common Voice. In the table below I report the Word Error Rate (WER) and the Character Error Rate (CER) of the model. I ran the… Open weights apache-2.0 transformers [View model](https://savrn.com/models/wav2vec2-large-xlsr-53-japanese) Model · Speech recognition ### [whisperkit-coreml](https://savrn.com/models/whisperkit-coreml) [Argmax](https://savrn.com/model-publishers/argmaxinc) WhisperKit is part of Argmax OSS, an On-device Speech AI SDK for Apple Silicon: https://github.com/argmaxinc/argmax-oss-swift Check out the WhisperKit paper and presentation from ICML 2025: https://icml.cc/virtual/2025/47854 For real-time transcription with speakers and custom vocabulary, check out Argmax Pro SDK: https://www.argmaxinc.com/blog/argmax-sdk-2 Open weights mit whisperkit [View model](https://savrn.com/models/whisperkit-coreml) Model · Speech recognition ### [speaker-diarization-3.1](https://savrn.com/models/speaker-diarization-3-1) [Pyannote](https://savrn.com/model-publishers/pyannote) Using this open-source model in production? Consider switching to pyannoteAI for better and faster options. This pipeline is the same as pyannote/speaker-diarization-3.0 except it removes the problematic use of onnxruntime. Both speaker segmentation and embedding now run in pure PyTorch. This should ease deployment and possibly speed up inference. It requires pyannote.audio version 3.1 or higher. It ingests mono audio sampled at 16kHz and outputs speaker diarization as an Annotation instance: - stereo or multi-channel audio files are automatically downmixed to mono by averaging the channels. - audio files sampled at a different rate are resampled to 16kHz automatically upon loading. 1.… Access requested at publisher mit pyannote-audio [View model](https://savrn.com/models/speaker-diarization-3-1) Model · Speech recognition ### [speaker-diarization-community-1](https://savrn.com/models/speaker-diarization-community-1) [Pyannote](https://savrn.com/model-publishers/pyannote) This pipeline ingests mono audio sampled at 16kHz and outputs speaker diarization. - stereo or multi-channel audio files are automatically downmixed to mono by averaging the channels. - audio files sampled at a different rate are resampled to 16kHz automatically upon loading. The main improvements brought by Community-1 are: - improved speaker assignment and counting - simpler reconciliation with transcription timestamps with exclusive speaker diarization - easy offline use (i.e. without internet connection) - (optionally) hosted on pyannoteAI cloud 1. pip install pyannote.audio 3. Create access token at hf.co/settings/tokens. Out of the box, Community-1 is much better than… Access requested at publisher cc-by-4.0 pyannote-audio [View model](https://savrn.com/models/speaker-diarization-community-1) Model · Speech recognition ### [faster-whisper-small](https://savrn.com/models/faster-whisper-small) [Systran](https://savrn.com/model-publishers/systran) This repository contains the conversion of openai/whisper-small to the CTranslate2 model format. This model can be used in CTranslate2 or projects based on CTranslate2 such as faster-whisper. The original model was converted with the following command: Note that the model weights are saved in FP16. This type can be changed when the model is loaded using the computetype option in CTranslate2. For more information about the original model, see its model card. Open weights mit ctranslate2 [View model](https://savrn.com/models/faster-whisper-small) Model · Speech recognition ### [wav2vec2-large-xlsr-53-portuguese](https://savrn.com/models/wav2vec2-large-xlsr-53-portuguese) [Jonatas Grosman](https://savrn.com/model-publishers/jonatasgrosman) Fine-tuned facebook/wav2vec2-large-xlsr-53 on Portuguese using the train and validation splits of Common Voice 6.1. When using this model, make sure that your speech input is sampled at 16kHz. This model has been fine-tuned thanks to the GPU credits generously given by the OVHcloud:) The script used for training can be found here: https://github.com/jonatasgrosman/wav2vec2-sprint The model can be used directly (without a language model) as follows... Using the HuggingSound library: 1. To evaluate on mozilla-foundation/commonvoice60 with split test 2. To evaluate on speech-recognition-community-v2/devdata If you want to cite this model you can use this Open weights apache-2.0 transformers [View model](https://savrn.com/models/wav2vec2-large-xlsr-53-portuguese) ## Indonesian NLP [All models and datasets](https://savrn.com/model-publishers/indonesian-nlp) ## Versions - [e5e699fa5aa5](https://savrn.com/models/wav2vec2-indonesian-javanese-sundanese/versions/e5e699fa5aa5) · current 2026-10-25 ## Explore More - [All speech recognition models](https://savrn.com/models/tasks/speech-recognition) - [All models under apache-2.0](https://savrn.com/models/licenses/apache-2-0) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-19. - [Hugging Face record](https://huggingface.co/indonesian-nlp/wav2vec2-indonesian-javanese-sundanese) - [How the hub is built](https://savrn.com/model-hub/methodology)