README.md

metadata

license: cc-by-nc-4.0
tags:
  - generated_from_trainer
datasets:
  - common_voice_6_1
metrics:
  - wer
base_model: facebook/mms-1b-all
model-index:
  - name: wav2vec2-large-mms-1b-turkish-colab
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: common_voice_6_1
          type: common_voice_6_1
          config: tr
          split: test
          args: tr
        metrics:
          - type: wer
            value: 0.22275559187008478
            name: Wer

wav2vec2-large-mms-1b-turkish-colab

This model is a fine-tuned version of facebook/mms-1b-all on the common_voice_6_1 dataset. It achieves the following results on the evaluation set:

• Loss: 0.1556
• Wer: 0.2228

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: 0.001
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
4.905	0.92	100	0.2146	0.2796
0.2901	1.83	200	0.1673	0.2317
0.2659	2.75	300	0.1608	0.2293
0.2398	3.67	400	0.1556	0.2228

Framework versions

• Transformers 4.31.0.dev0
• Pytorch 2.0.1+cu118
• Datasets 2.13.0
• Tokenizers 0.13.3