README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
base_model: facebook/wav2vec2-conformer-rope-large-960h-ft
datasets:
  - common_voice_17_0
metrics:
  - wer
model-index:
  - name: wav2vec2-conformer-rope-large-960h-ft-armenian-CV17.0
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: common_voice_17_0
          type: common_voice_17_0
          config: hy-AM
          split: None
          args: hy-AM
        metrics:
          - type: wer
            value: 0.990876791521137
            name: Wer

wav2vec2-conformer-rope-large-960h-ft-armenian-CV17.0

This model is a fine-tuned version of facebook/wav2vec2-conformer-rope-large-960h-ft on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:

• Loss: 3.1627
• Wer: 0.9909
• Cer: 0.8400

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
4.2764	1.0	325	3.1252	1.0	0.9984
2.9396	2.0	650	3.1627	0.9909	0.8400

Framework versions

• Transformers 4.41.0.dev0
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1