speecht5_kavinda1

This model is a fine-tuned version of microsoft/speecht5_tts on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4020

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: 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: 100
• num_epochs: 40
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.8309	1.0	14	0.5929
0.6544	2.0	28	0.5246
0.6079	3.0	42	0.4954
0.5618	4.0	56	0.4706
0.4976	5.0	70	0.4192
0.4688	6.0	84	0.4151
0.4628	7.0	98	0.4090
0.4365	8.0	112	0.4050
0.4674	9.0	126	0.4146
0.4548	10.0	140	0.3966
0.426	11.0	154	0.4067
0.4211	12.0	168	0.3890
0.4106	13.0	182	0.3842
0.4125	14.0	196	0.3863
0.4079	15.0	210	0.3810
0.4003	16.0	224	0.3731
0.396	17.0	238	0.3912
0.392	18.0	252	0.3810
0.387	19.0	266	0.3835
0.3908	20.0	280	0.3847
0.3846	21.0	294	0.3785
0.3834	22.0	308	0.3706
0.386	23.0	322	0.3826
0.3868	24.0	336	0.3899
0.3814	25.0	350	0.3859
0.3711	26.0	364	0.3980
0.3743	27.0	378	0.3855
0.3702	28.0	392	0.3896
0.3682	29.0	406	0.3976
0.3704	30.0	420	0.3898
0.3633	31.0	434	0.3868
0.364	32.0	448	0.3934
0.3674	33.0	462	0.3838
0.357	34.0	476	0.3909
0.3581	35.0	490	0.4014
0.3554	36.0	504	0.3981
0.361	37.0	518	0.3970
0.3531	38.0	532	0.3970
0.3486	39.0	546	0.3998
0.3499	40.0	560	0.4020

Framework versions

• Transformers 4.53.0
• Pytorch 2.6.0+cu124
• Datasets 2.14.4
• Tokenizers 0.21.2