speecht5_finetuned

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: 0.5052

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: 8
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: 500
num_epochs: 60
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.5807	1.1364	100	0.5336
0.5739	2.2727	200	0.5341
0.5683	3.4091	300	0.5317
0.5699	4.5455	400	0.5380
0.5803	5.6818	500	0.5312
0.5616	6.8182	600	0.5345
0.5611	7.9545	700	0.5211
0.5543	9.0909	800	0.5207
0.5546	10.2273	900	0.5216
0.5483	11.3636	1000	0.5167
0.544	12.5	1100	0.5161
0.5359	13.6364	1200	0.5157
0.5375	14.7727	1300	0.5093
0.539	15.9091	1400	0.5106
0.5385	17.0455	1500	0.5112
0.5359	18.1818	1600	0.5094
0.5283	19.3182	1700	0.5186
0.5257	20.4545	1800	0.5095
0.5226	21.5909	1900	0.5108
0.52	22.7273	2000	0.5070
0.5188	23.8636	2100	0.5086
0.5182	25.0	2200	0.5052
0.5148	26.1364	2300	0.5059
0.509	27.2727	2400	0.5063
0.511	28.4091	2500	0.5055
0.5051	29.5455	2600	0.5049
0.5022	30.6818	2700	0.5052
0.5015	31.8182	2800	0.5013
0.5012	32.9545	2900	0.5063
0.5049	34.0909	3000	0.5030
0.4962	35.2273	3100	0.5013
0.493	36.3636	3200	0.5031
0.5	37.5	3300	0.5031
0.4902	38.6364	3400	0.5035
0.4922	39.7727	3500	0.5038
0.4871	40.9091	3600	0.5043
0.4913	42.0455	3700	0.5030
0.4915	43.1818	3800	0.5044
0.4831	44.3182	3900	0.5067
0.4893	45.4545	4000	0.5055
0.4829	46.5909	4100	0.5038
0.4891	47.7273	4200	0.5044
0.4861	48.8636	4300	0.5039
0.4846	50.0	4400	0.5030
0.484	51.1364	4500	0.5032
0.4821	52.2727	4600	0.5052
0.4833	53.4091	4700	0.5047
0.4845	54.5455	4800	0.5041
0.4778	55.6818	4900	0.5056
0.4793	56.8182	5000	0.5054
0.4743	57.9545	5100	0.5053
0.4721	59.0909	5200	0.5052

Framework versions

Transformers 4.51.3
Pytorch 2.7.1+cu118
Datasets 3.6.0
Tokenizers 0.21.1

Loading…

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0001

train_batch_size: 4

eval_batch_size: 8

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: 500

num_epochs: 60

mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.5807	1.1364	100	0.5336
0.5739	2.2727	200	0.5341
0.5683	3.4091	300	0.5317
0.5699	4.5455	400	0.5380
0.5803	5.6818	500	0.5312
0.5616	6.8182	600	0.5345
0.5611	7.9545	700	0.5211
0.5543	9.0909	800	0.5207
0.5546	10.2273	900	0.5216
0.5483	11.3636	1000	0.5167
0.544	12.5	1100	0.5161
0.5359	13.6364	1200	0.5157
0.5375	14.7727	1300	0.5093
0.539	15.9091	1400	0.5106
0.5385	17.0455	1500	0.5112
0.5359	18.1818	1600	0.5094
0.5283	19.3182	1700	0.5186
0.5257	20.4545	1800	0.5095
0.5226	21.5909	1900	0.5108
0.52	22.7273	2000	0.5070
0.5188	23.8636	2100	0.5086
0.5182	25.0	2200	0.5052
0.5148	26.1364	2300	0.5059
0.509	27.2727	2400	0.5063
0.511	28.4091	2500	0.5055
0.5051	29.5455	2600	0.5049
0.5022	30.6818	2700	0.5052
0.5015	31.8182	2800	0.5013
0.5012	32.9545	2900	0.5063
0.5049	34.0909	3000	0.5030
0.4962	35.2273	3100	0.5013
0.493	36.3636	3200	0.5031
0.5	37.5	3300	0.5031
0.4902	38.6364	3400	0.5035
0.4922	39.7727	3500	0.5038
0.4871	40.9091	3600	0.5043
0.4913	42.0455	3700	0.5030
0.4915	43.1818	3800	0.5044
0.4831	44.3182	3900	0.5067
0.4893	45.4545	4000	0.5055
0.4829	46.5909	4100	0.5038
0.4891	47.7273	4200	0.5044
0.4861	48.8636	4300	0.5039
0.4846	50.0	4400	0.5030
0.484	51.1364	4500	0.5032
0.4821	52.2727	4600	0.5052
0.4833	53.4091	4700	0.5047
0.4845	54.5455	4800	0.5041
0.4778	55.6818	4900	0.5056
0.4793	56.8182	5000	0.5054
0.4743	57.9545	5100	0.5053
0.4721	59.0909	5200	0.5052

Framework versions

Transformers 4.51.3

Pytorch 2.7.1+cu118

Datasets 3.6.0

Tokenizers 0.21.1

schalor/speecht5_finetuned

speecht5_finetuned

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

speecht5_finetuned

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions