speecht5_finetuned_binisha

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: 0.6198

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
0.5825	88.8889	100	0.5775
0.4984	177.7778	200	0.6064
0.4252	266.6667	300	0.5800
0.3997	355.5556	400	0.5745
0.366	444.4444	500	0.5863
0.3521	533.3333	600	0.5969
0.3308	622.2222	700	0.5716
0.32	711.1111	800	0.5757
0.3088	800.0	900	0.6095
0.3006	888.8889	1000	0.6352
0.2911	977.7778	1100	0.6207
0.2849	1066.6667	1200	0.6181
0.2869	1155.5556	1300	0.6321
0.2853	1244.4444	1400	0.6271
0.285	1333.3333	1500	0.6198

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1