csikasote's picture
End of training
95e2b36 verified
metadata
library_name: transformers
license: cc-by-nc-4.0
base_model: mms-meta/mms-zeroshot-300m
tags:
  - automatic-speech-recognition
  - genbed
  - mms
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: mms-zeroshot-300m-genbed-m-model
    results: []

mms-zeroshot-300m-genbed-m-model

This model is a fine-tuned version of mms-meta/mms-zeroshot-300m on the GENBED - BEM dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3383
  • Wer: 0.4761

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
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 30.0
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
No log 0.5510 200 2.3107 1.0
No log 1.1019 400 0.4860 0.6380
2.8087 1.6529 600 0.4381 0.5988
2.8087 2.2039 800 0.4281 0.5855
0.6248 2.7548 1000 0.4056 0.5656
0.6248 3.3058 1200 0.4017 0.5513
0.6248 3.8567 1400 0.3904 0.5617
0.578 4.4077 1600 0.3776 0.5306
0.578 4.9587 1800 0.3722 0.5178
0.5343 5.5096 2000 0.3659 0.5152
0.5343 6.0606 2200 0.3614 0.5150
0.5343 6.6116 2400 0.3573 0.5016
0.5153 7.1625 2600 0.3625 0.5026
0.5153 7.7135 2800 0.3545 0.4867
0.4935 8.2645 3000 0.3506 0.4814
0.4935 8.8154 3200 0.3482 0.4922
0.4935 9.3664 3400 0.3383 0.4761
0.4731 9.9174 3600 0.3419 0.4645
0.4731 10.4683 3800 0.3391 0.4686
0.4619 11.0193 4000 0.3402 0.4576

Framework versions

  • Transformers 4.46.0.dev0
  • Pytorch 2.4.1+cu121
  • Datasets 3.0.1
  • Tokenizers 0.20.0