Description
Pretrained Wav2vec2 pipeline, adapted from Hugging Face and curated to provide scalability and production-readiness using Spark NLP.asr_wav2vec2_large_voxrex_swedish_4gram is a Swedish model originally trained by viktor-enzell.
NOTE: This pipeline only works on a CPU, if you need to use this pipeline on a GPU device please use pipeline_asr_wav2vec2_large_voxrex_swedish_4gram_gpu
How to use
pipeline = PretrainedPipeline('pipeline_asr_wav2vec2_large_voxrex_swedish_4gram', lang = 'sv')
annotations = pipeline.transform(audioDF)
val pipeline = new PretrainedPipeline("pipeline_asr_wav2vec2_large_voxrex_swedish_4gram", lang = "sv")
val annotations = pipeline.transform(audioDF)
Model Information
| Model Name: | pipeline_asr_wav2vec2_large_voxrex_swedish_4gram |
| Type: | pipeline |
| Compatibility: | Spark NLP 4.2.0+ |
| License: | Open Source |
| Edition: | Official |
| Language: | sv |
| Size: | 757.4 MB |
Included Models
- AudioAssembler
- Wav2Vec2ForCTC