Edit on GitHub
  • John Snow Labs
  • Sep 25, 2022

German asr_wav2vec2_large_xlsr_53_german_with_lm TFWav2Vec2ForCTC from aware-ai

  • wav2vec2
  • de
  • audio
  • open_source
  • pipeline
  • asr

Description

Pretrained Wav2vec2 pipeline, adapted from Hugging Face and curated to provide scalability and production-readiness using Spark NLP.asr_wav2vec2_large_xlsr_53_german_with_lm is a German model originally trained by aware-ai.

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_xlsr_53_german_with_lm_gpu

Download Copy S3 URI

How to use


    pipeline = PretrainedPipeline('pipeline_asr_wav2vec2_large_xlsr_53_german_with_lm', lang = 'de')
    annotations =  pipeline.transform(audioDF)
    


    val pipeline = new PretrainedPipeline("pipeline_asr_wav2vec2_large_xlsr_53_german_with_lm", lang = "de")
    val annotations = pipeline.transform(audioDF)

Model Information

Model Name: pipeline_asr_wav2vec2_large_xlsr_53_german_with_lm
Type: pipeline
Compatibility: Spark NLP 4.2.0+
License: Open Source
Edition: Official
Language: de
Size: 1.2 GB

Included Models

  • AudioAssembler
  • Wav2Vec2ForCTC