Description
Pretrained Wav2Vec2ForCTC model, adapted from Hugging Face and curated to provide scalability and production-readiness using Spark NLP.w2v2_base_pretrained_lr1e_4_at0_8_da0_2 is a English model originally trained by MelanieKoe.
How to use
audioAssembler = AudioAssembler() \
.setInputCol("audio_content") \
.setOutputCol("audio_assembler")
speechToText = Wav2Vec2ForCTC.pretrained("w2v2_base_pretrained_lr1e_4_at0_8_da0_2","en") \
.setInputCols(["audio_assembler"]) \
.setOutputCol("text")
pipeline = Pipeline().setStages([audioAssembler, speechToText])
pipelineModel = pipeline.fit(data)
pipelineDF = pipelineModel.transform(data)
val audioAssembler = new DocumentAssembler()
.setInputCols("audio_content")
.setOutputCols("audio_assembler")
val speechToText = Wav2Vec2ForCTC.pretrained("w2v2_base_pretrained_lr1e_4_at0_8_da0_2", "en")
.setInputCols(Array("audio_assembler"))
.setOutputCol("text")
val pipeline = new Pipeline().setStages(Array(documentAssembler, speechToText))
val pipelineModel = pipeline.fit(data)
val pipelineDF = pipelineModel.transform(data)
Model Information
| Model Name: | w2v2_base_pretrained_lr1e_4_at0_8_da0_2 |
| Compatibility: | Spark NLP 5.5.1+ |
| License: | Open Source |
| Edition: | Official |
| Input Labels: | [audio_assembler] |
| Output Labels: | [text] |
| Language: | en |
| Size: | 348.7 MB |
References
https://huggingface.co/MelanieKoe/w2v2-base-pretrained_lr1e-4_at0.8_da0.2