Xls R 300 Sv Cv7

🚀 XLS-R-300M - Swedish - CV7 - v2

This is a fine - tuned model for automatic speech recognition on the Swedish language, based on the facebook/wav2vec2 - xls - r - 300m model, achieving good results on the evaluation set.

🚀 Quick Start

This model is a fine - tuned version of facebook/wav2vec2-xls-r-300m on the MOZILLA - FOUNDATION/COMMON_VOICE_7_0 - SV - SE dataset. It achieves the following results on the evaluation set:

• Loss: 0.2604
• Wer: 0.2334

✨ Features

• Automatic Speech Recognition: Specialized for Swedish speech recognition tasks.
• Fine - Tuned: Based on a pre - trained model and fine - tuned on specific Swedish datasets.

📦 Installation

No installation steps are provided in the original document, so this section is skipped.

💻 Usage Examples

No code examples are provided in the original document, so this section is skipped.

📚 Documentation

Model Information

Model Index

• Name: XLS - R - 300M - Swedish - CV7 - v2
• Results:
  • Task 1:
    • Task Name: Automatic Speech Recognition
    • Dataset: Common Voice 7 (mozilla - foundation/common_voice_7_0, args: sv - SE)
    • Metrics:
      • Test WER: 15.99
      • Test CER: 5.2
  • Task 2:
    • Task Name: Automatic Speech Recognition
    • Dataset: Robust Speech Event - Dev Data (speech - recognition - community - v2/dev_data, args: sv)
    • Metrics:
      • Test WER: 24.41
      • Test CER: 11.88

Training Procedure

Training Hyperparameters

The following hyperparameters were used during training:

• learning_rate: 7.5e - 05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi - GPU
• num_devices: 8
• gradient_accumulation_steps: 1
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2000
• num_epochs: 50.0
• mixed_precision_training: Native AMP

Training Results

See Tensorboard

Evaluation Commands

1. To evaluate on mozilla - foundation/common_voice_7_0 with split test

python eval.py --model_id patrickvonplaten/xls - r - 300 - sv - cv7 --dataset mozilla - foundation/common_voice_7_0 --config sv - SE --split test

2. To evaluate on speech - recognition - community - v2/dev_data

python eval.py --model_id patrickvonplaten/xls - r - 300 - sv - cv7 --dataset speech - recognition - community - v2/dev_data --config sv --split validation --chunk_length_s 5.0 --stride_length_s 1.0

Framework Versions

• Transformers 4.17.0.dev0
• Pytorch 1.9.0+cu111
• Datasets 1.18.4.dev0
• Tokenizers 0.10.3

📄 License

This model is licensed under the Apache - 2.0 license.