metadata

license: cc-by-sa-4.0
tags:
  - generated_from_trainer
datasets:
  - common_voice
metrics:
  - wer
model-index:
  - name: wav2vec2-detect-toxic-th
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: common_voice
          type: common_voice
          config: th
          split: validation
          args: th
        metrics:
          - name: Wer
            type: wer
            value: 0.4550641940085592

wav2vec2-detect-toxic-th

This model is a fine-tuned version of airesearch/wav2vec2-large-xlsr-53-th on the common_voice dataset. It achieves the following results on the evaluation set:

Loss: 1.5867
Wer: 0.4551

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: 0.0001
train_batch_size: 32
eval_batch_size: 16
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 30
num_epochs: 100
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
5.4333	3.23	100	3.3662	1.0
3.3254	6.45	200	3.2575	1.0
2.5091	9.68	300	1.2965	0.5571
1.1749	12.9	400	1.0687	0.5464
0.9091	16.13	500	1.0564	0.4872
0.756	19.35	600	1.0998	0.4757
0.6527	22.58	700	1.1492	0.4829
0.5879	25.81	800	1.1916	0.4786
0.5184	29.03	900	1.2662	0.4815
0.4688	32.26	1000	1.2109	0.4864
0.4587	35.48	1100	1.3144	0.4722
0.4005	38.71	1200	1.3111	0.4686
0.3851	41.94	1300	1.3420	0.4786
0.3563	45.16	1400	1.3679	0.4743
0.3591	48.39	1500	1.4444	0.4643
0.325	51.61	1600	1.4076	0.4722
0.3409	54.84	1700	1.4586	0.4629
0.3019	58.06	1800	1.4579	0.4529
0.292	61.29	1900	1.4887	0.4522
0.2729	64.52	2000	1.4966	0.4608
0.2656	67.74	2100	1.5232	0.4593
0.2575	70.97	2200	1.4984	0.4508
0.2532	74.19	2300	1.5332	0.4544
0.2474	77.42	2400	1.5301	0.4529
0.2539	80.65	2500	1.5214	0.4601
0.2526	83.87	2600	1.5413	0.4572
0.2601	87.1	2700	1.5553	0.4608
0.2315	90.32	2800	1.5768	0.4515
0.2477	93.55	2900	1.5787	0.4650
0.2363	96.77	3000	1.5900	0.4565
0.242	100.0	3100	1.5867	0.4551

Framework versions

Transformers 4.28.0
Pytorch 2.0.1+cu118
Datasets 1.16.1
Tokenizers 0.13.3