speecht5_finetuned_sr_proj

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: 0.4312

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: 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: 200
• training_steps: 1400
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.7348	0.2093	100	0.6215
0.6423	0.4186	200	0.5610
0.5402	0.6279	300	0.4746
0.5147	0.8373	400	0.4594
0.5036	1.0466	500	0.4507
0.5033	1.2559	600	0.4477
0.4917	1.4652	700	0.4407
0.492	1.6745	800	0.4391
0.481	1.8838	900	0.4387
0.4826	2.0931	1000	0.4334
0.4842	2.3025	1100	0.4342
0.4796	2.5118	1200	0.4325
0.4735	2.7211	1300	0.4311
0.4771	2.9304	1400	0.4312

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1