ESP_TWICE_DATA

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: 0.4606

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• training_steps: 1500
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.5685	5.1613	100	0.5270
0.5244	10.3226	200	0.4922
0.5034	15.4839	300	0.4847
0.4875	20.6452	400	0.4703
0.4759	25.8065	500	0.4669
0.4718	30.9677	600	0.4664
0.4604	36.1290	700	0.4609
0.4526	41.2903	800	0.4637
0.4525	46.4516	900	0.4602
0.4452	51.6129	1000	0.4608
0.448	56.7742	1100	0.4622
0.4373	61.9355	1200	0.4580
0.4326	67.0968	1300	0.4580
0.4335	72.2581	1400	0.4594
0.4328	77.4194	1500	0.4606

Framework versions

• Transformers 4.46.0.dev0
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.0