metadata
license: apache-2.0
base_model: facebook/wav2vec2-base-960h
tags:
- generated_from_trainer
datasets:
- ami
metrics:
- wer
model-index:
- name: my_awesome_asr_mind_model6e-5
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: ami
type: ami
config: ihm
split: None
args: ihm
metrics:
- name: Wer
type: wer
value: 0.26252597552528284
my_awesome_asr_mind_model6e-5
This model is a fine-tuned version of facebook/wav2vec2-base-960h on the ami dataset. It achieves the following results on the evaluation set:
- Loss: 0.9652
- Wer: 0.2625
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: 6e-05
- 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: 1000
- training_steps: 3000
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
3.1601 | 15.1515 | 500 | 3.1815 | 1.0 |
3.0665 | 30.3030 | 1000 | 3.5100 | 1.0 |
2.1863 | 45.4545 | 1500 | 1.2838 | 0.3812 |
0.9609 | 60.6061 | 2000 | 0.9112 | 0.2863 |
0.6826 | 75.7576 | 2500 | 0.9450 | 0.2667 |
0.5687 | 90.9091 | 3000 | 0.9652 | 0.2625 |
Framework versions
- Transformers 4.42.4
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1