You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this model content.

Visualize in Weights & Biases

w2v2-bert-Wolof-1-hour-kallaama-dataset

This model is a fine-tuned version of facebook/w2v-bert-2.0 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 5.1408
  • Wer: 0.8229
  • Cer: 0.4237

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_ratio: 0.1
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Wer Cer
2.39 14.8148 200 2.1002 0.8772 0.4412
0.9105 29.6296 400 3.2675 0.8088 0.4193
0.1322 44.4444 600 5.1408 0.8229 0.4237

Framework versions

  • Transformers 4.44.1
  • Pytorch 2.1.0+cu118
  • Datasets 2.17.0
  • Tokenizers 0.19.1
Downloads last month
0
Safetensors
Model size
606M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.

Model tree for asr-africa/w2v2-bert-Wolof-1-hour-kallaama-dataset

Finetuned
(255)
this model