ceb_b32_le4_s12000

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: 0.4051

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: 2
• 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: 2000
• training_steps: 12000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.4691	9.9010	500	0.4229
0.4352	19.8020	1000	0.4041
0.424	29.7030	1500	0.4032
0.4091	39.6040	2000	0.4037
0.4117	49.5050	2500	0.4078
0.3884	59.4059	3000	0.4005
0.3826	69.3069	3500	0.4024
0.3766	79.2079	4000	0.4015
0.3712	89.1089	4500	0.4025
0.3571	99.0099	5000	0.4016
0.3671	108.9109	5500	0.4021
0.361	118.8119	6000	0.4025
0.3581	128.7129	6500	0.3989
0.3476	138.6139	7000	0.4029
0.3391	148.5149	7500	0.4026
0.3372	158.4158	8000	0.4037
0.3345	168.3168	8500	0.4045
0.3329	178.2178	9000	0.4067
0.331	188.1188	9500	0.4042
0.3366	198.0198	10000	0.4051
0.3276	207.9208	10500	0.4035
0.3297	217.8218	11000	0.4037
0.3298	227.7228	11500	0.4031
0.3241	237.6238	12000	0.4051

Framework versions

• Transformers 4.41.0.dev0
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1