facebook/wav2vec2-large-xlsr-53-french

The Common Voice FR Test is based on a common voice test . The test was evaluated using the WAV2Vec2ForCTC, Wav2VEC2Processor and a WAVV2Fec2 Processor . The result was 25.2% [Evaluation] on Common VoiceFR Test . The code was created by torchaudio with the help of the torchaudioaudio.org and re-transformer.org . The data set was created from a dataset called load_dataset, load_metric.org, and a transformers. The code is compiled using torchaudio’s “common voice” toolkit.org,