speecht5_improved_data_less_steps

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: 0.6468

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 250
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.4475	0.1533	25	1.1070
1.3754	0.3065	50	1.0239
1.2957	0.4598	75	0.9667
1.1821	0.6130	100	0.9231
1.1414	0.7663	125	0.8946
1.0555	0.9195	150	0.8335
0.9565	1.0728	175	0.7764
0.9189	1.2261	200	0.7485
0.8751	1.3793	225	0.7332
0.8533	1.5326	250	0.7197
0.8219	1.6858	275	0.7133
0.8171	1.8391	300	0.7009
0.8088	1.9923	325	0.6926
0.785	2.1456	350	0.6877
0.7989	2.2989	375	0.6824
0.7755	2.4521	400	0.6784
0.7914	2.6054	425	0.6738
0.7696	2.7586	450	0.6694
0.7741	2.9119	475	0.6680
0.7613	3.0651	500	0.6659
0.7733	3.2184	525	0.6654
0.7605	3.3716	550	0.6623
0.7538	3.5249	575	0.6606
0.7626	3.6782	600	0.6596
0.7573	3.8314	625	0.6577
0.7469	3.9847	650	0.6556
0.7524	4.1379	675	0.6537
0.7342	4.2912	700	0.6491
0.7305	4.4444	725	0.6511
0.7433	4.5977	750	0.6486
0.7438	4.7510	775	0.6487
0.7505	4.9042	800	0.6481
0.7354	5.0575	825	0.6450
0.7354	5.2107	850	0.6439
0.7333	5.3640	875	0.6462
0.7246	5.5172	900	0.6444
0.7289	5.6705	925	0.6417
0.7436	5.8238	950	0.6474
0.741	5.9770	975	0.6428
0.7443	6.1303	1000	0.6468

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.17.0
• Tokenizers 0.19.1