speecht5_finetuned_as-krn_tr_v2

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: 0.3594

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
• training_steps: 500
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.7134	0.1575	25	0.5496
0.6033	0.3150	50	0.4918
0.5202	0.4724	75	0.4590
0.492	0.6299	100	0.4208
0.4701	0.7874	125	0.4153
0.451	0.9449	150	0.4076
0.4352	1.1008	175	0.3961
0.4328	1.2583	200	0.3925
0.4265	1.4157	225	0.3894
0.4263	1.5732	250	0.3935
0.4166	1.7307	275	0.3791
0.4156	1.8882	300	0.3735
0.4007	2.0441	325	0.3717
0.4002	2.2016	350	0.3686
0.4021	2.3591	375	0.3677
0.3945	2.5165	400	0.3651
0.3969	2.6740	425	0.3630
0.3936	2.8315	450	0.3617
0.3923	2.9890	475	0.3592
0.3839	3.1449	500	0.3594

Framework versions

• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2