STT_Model_5

This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1410
• Wer: 0.1808

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: 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: 1000
• num_epochs: 80

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.882	5.68	500	0.5476	0.9998
0.4219	11.36	1000	0.2672	0.7620
0.1972	17.05	1500	0.1670	0.4849
0.1313	22.73	2000	0.1425	0.3606
0.1002	28.41	2500	0.1259	0.2846
0.0752	34.09	3000	0.1307	0.2460
0.0552	39.77	3500	0.1193	0.2192
0.0426	45.45	4000	0.1264	0.2025
0.0319	51.14	4500	0.1237	0.1974
0.025	56.82	5000	0.1327	0.1904
0.0207	62.5	5500	0.1493	0.1921
0.0178	68.18	6000	0.1416	0.1865
0.0142	73.86	6500	0.1382	0.1802
0.0123	79.55	7000	0.1410	0.1808

Framework versions

• Transformers 4.26.0
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2