nyiha_nt-speecht5

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: 0.4470

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 3407
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 5
• training_steps: 20000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.4977	9.0090	1000	0.4676
0.4789	18.0180	2000	0.4541
0.4658	27.0270	3000	0.4502
0.458	36.0360	4000	0.4502
0.4573	45.0450	5000	0.4463
0.4533	54.0541	6000	0.4457
0.4499	63.0631	7000	0.4443
0.4428	72.0721	8000	0.4443
0.434	81.0811	9000	0.4464
0.4305	90.0901	10000	0.4457
0.4214	99.0991	11000	0.4426
0.4179	108.1081	12000	0.4461
0.4314	117.1171	13000	0.4473
0.4386	126.1261	14000	0.4457
0.4239	135.1351	15000	0.4450
0.4201	144.1441	16000	0.4464
0.4149	153.1532	17000	0.4456
0.4192	162.1622	18000	0.4449
0.4562	171.1712	19000	0.4465
0.4179	180.1802	20000	0.4470

Framework versions

• Transformers 4.53.3
• Pytorch 2.7.1+cu126
• Datasets 4.0.0
• Tokenizers 0.21.2