batch_size_2

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.8600

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: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss
13.2945	1.69	200	3.6357
3.737	3.38	400	3.6624
3.707	5.06	600	3.6271
3.6038	6.75	800	3.5749
3.4502	8.44	1000	3.2693
3.1641	10.13	1200	2.9155
2.6047	11.81	1400	2.3544
2.0028	13.5	1600	2.0022
1.5415	15.19	1800	1.8024
1.1696	16.88	2000	1.7754
0.8809	18.57	2200	1.6793
0.7591	20.25	2400	1.8127
0.6056	21.94	2600	1.7993
0.5091	23.63	2800	1.7812
0.4569	25.32	3000	1.7781
0.4167	27.0	3200	1.8802
0.3432	28.69	3400	1.8600

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1
• Datasets 2.12.0
• Tokenizers 0.13.2