jimregan
/

wav2vec2-large-xlsr-latvian-cv

Automatic Speech Recognition Transformers PyTorch JAX

Latvian wav2vec2 audio speech xlsr-fine-tuning-week Eval Results Inference Endpoints

Model card Files Files and versions Community

jimregan commited on Mar 20, 2021

Commit

78d9d91

•

1 Parent(s): a8be5f0

fix more

Files changed (1) hide show

README.md +3 -3

README.md CHANGED Viewed

@@ -18,7 +18,7 @@ model-index:
     dataset:
       name: Common Voice lv
       type: common_voice
-      args: id
     metrics:
        - name: Test WER
          type: wer
@@ -26,7 +26,7 @@ model-index:
 ---
 # Wav2Vec2-Large-XLSR-Latvian
 Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53)
-on the [Indonesian Common Voice dataset](https://huggingface.co/datasets/common_voice).
 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:
@@ -35,7 +35,7 @@ import torch
 import torchaudio
 from datasets import load_dataset
 from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
-test_dataset = load_dataset("common_voice", "id", split="test[:2%]")
 processor = Wav2Vec2Processor.from_pretrained("jimregan/wav2vec2-large-xlsr-latvian-cv")
 model = Wav2Vec2ForCTC.from_pretrained("jimregan/wav2vec2-large-xlsr-latvian-cv")
 resampler = torchaudio.transforms.Resample(48_000, 16_000)

     dataset:
       name: Common Voice lv
       type: common_voice
+      args: lv
     metrics:
        - name: Test WER
          type: wer
 ---
 # Wav2Vec2-Large-XLSR-Latvian
 Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53)
+on the [Latvian Common Voice dataset](https://huggingface.co/datasets/common_voice).
 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:
 import torchaudio
 from datasets import load_dataset
 from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
+test_dataset = load_dataset("common_voice", "lv", split="test[:2%]")
 processor = Wav2Vec2Processor.from_pretrained("jimregan/wav2vec2-large-xlsr-latvian-cv")
 model = Wav2Vec2ForCTC.from_pretrained("jimregan/wav2vec2-large-xlsr-latvian-cv")
 resampler = torchaudio.transforms.Resample(48_000, 16_000)