HopefulASR

This model is a fine-tuned version of microsoft/speecht5_tts on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4022

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: 4
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.4624	3.0769	100	0.4358
0.4523	6.1538	200	0.4288
0.4385	9.2308	300	0.4148
0.4288	12.3077	400	0.4126
0.4262	15.3846	500	0.4114
0.4236	18.4615	600	0.4128
0.416	21.5385	700	0.4105
0.4118	24.6154	800	0.4111
0.4091	27.6923	900	0.4074
0.4044	30.7692	1000	0.4085
0.4048	33.8462	1100	0.4080
0.402	36.9231	1200	0.4021
0.3957	40.0	1300	0.4034
0.3977	43.0769	1400	0.4027
0.3982	46.1538	1500	0.4041
0.395	49.2308	1600	0.4003
0.3913	52.3077	1700	0.4017
0.3969	55.3846	1800	0.4032
0.3918	58.4615	1900	0.4024
0.3898	61.5385	2000	0.4022

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1