test-model

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: 0.2455
• Accuracy: 0.9336

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: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 100
• training_steps: 600
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.9682	0.9639	60	0.8113	0.6177
0.7135	1.9277	120	0.6503	0.7425
0.585	2.8916	180	0.5173	0.8149
0.4603	3.8554	240	0.4556	0.8551
0.3855	4.8193	300	0.2616	0.9256
0.284	5.7831	360	0.3075	0.9235
0.2411	6.7470	420	0.2254	0.9376
0.1818	7.7108	480	0.2609	0.9356
0.1644	8.6747	540	0.2562	0.9336
0.1285	9.6386	600	0.2455	0.9336

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1