speecht5_dhivehi_tts_v1

This model is a fine-tuned version of on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3982

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: 128
• eval_batch_size: 16
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_steps: 500
• training_steps: 30000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.1121	6.5789	500	1.0624
1.0902	13.1579	1000	1.0063
1.0398	19.7368	1500	0.9727
0.9906	26.3158	2000	0.9334
0.9517	32.8947	2500	0.8930
0.9112	39.4737	3000	0.8556
0.8754	46.0526	3500	0.8299
0.8455	52.6316	4000	0.7964
0.8212	59.2105	4500	0.7761
0.7956	65.7895	5000	0.7534
0.782	72.3684	5500	0.7660
0.7518	78.9474	6000	0.7337
0.7403	85.5263	6500	0.7190
0.7262	92.1053	7000	0.7095
0.7112	98.6842	7500	0.6916
0.7086	105.2632	8000	0.6946
0.69	111.8421	8500	0.6727
0.6711	118.4211	9000	0.6532
0.6614	125.0	9500	0.6574
0.6538	131.5789	10000	0.6357
0.6307	138.1579	10500	0.6166
0.622	144.7368	11000	0.6139
0.6064	151.3158	11500	0.5978
0.5955	157.8947	12000	0.5843
0.5761	164.4737	12500	0.5681
0.5633	171.0526	13000	0.5688
0.5546	177.6316	13500	0.5483
0.5468	184.2105	14000	0.5336
0.535	190.7895	14500	0.5205
0.5279	197.3684	15000	0.5134
0.5172	203.9474	15500	0.5072
0.5094	210.5263	16000	0.4902
0.4969	217.1053	16500	0.4830
0.4893	223.6842	17000	0.4691
0.4821	230.2632	17500	0.4686
0.4727	236.8421	18000	0.4618
0.4696	243.4211	18500	0.4529
0.4659	250.0	19000	0.4477
0.4609	256.5789	19500	0.4421
0.456	263.1579	20000	0.4383
0.441	269.7368	20500	0.4305
0.4393	276.3158	21000	0.4277
0.435	282.8947	21500	0.4229
0.4326	289.4737	22000	0.4189
0.4283	296.0526	22500	0.4157
0.4277	302.6316	23000	0.4117
0.425	309.2105	23500	0.4113
0.4273	315.7895	24000	0.4076
0.4234	322.3684	24500	0.4034
0.4221	328.9474	25000	0.4052
0.4208	335.5263	25500	0.4034
0.4173	342.1053	26000	0.4000
0.42	348.6842	26500	0.4008
0.4188	355.2632	27000	0.3995
0.4224	361.8421	27500	0.4002
0.4144	368.4211	28000	0.3986
0.4182	375.0	28500	0.3979
0.4157	381.5789	29000	0.3984
0.4166	388.1579	29500	0.3987
0.4127	394.7368	30000	0.3982

Framework versions

• Transformers 4.48.0.dev0
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0