speecht5_dhivehi_tts_v7_from_scratch

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3932

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: 92
• 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: 2000
• training_steps: 70000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.4842	4.1667	1000	0.4482
0.485	8.3333	2000	0.4464
0.4862	12.5	3000	0.4476
0.482	16.6667	4000	0.4442
0.4832	20.8333	5000	0.4435
0.4803	25.0	6000	0.4433
0.4784	29.1667	7000	0.4421
0.4769	33.3333	8000	0.4416
0.4705	37.5	9000	0.4374
0.4711	41.6667	10000	0.4369
0.4705	45.8333	11000	0.4356
0.4694	50.0	12000	0.4331
0.4675	54.1667	13000	0.4327
0.4666	58.3333	14000	0.4325
0.4643	62.5	15000	0.4301
0.4608	66.6667	16000	0.4277
0.4585	70.8333	17000	0.4254
0.4592	75.0	18000	0.4255
0.4597	79.1667	19000	0.4232
0.4564	83.3333	20000	0.4247
0.4543	87.5	21000	0.4205
0.4528	91.6667	22000	0.4189
0.4507	95.8333	23000	0.4183
0.4489	100.0	24000	0.4167
0.4503	104.1667	25000	0.4168
0.447	108.3333	26000	0.4154
0.4421	112.5	27000	0.4150
0.4444	116.6667	28000	0.4149
0.4431	120.8333	29000	0.4116
0.4411	125.0	30000	0.4118
0.4431	129.1667	31000	0.4097
0.4376	133.3333	32000	0.4072
0.4343	137.5	33000	0.4091
0.4398	141.6667	34000	0.4061
0.4362	145.8333	35000	0.4063
0.4365	150.0	36000	0.4039
0.4363	154.1667	37000	0.4042
0.4341	158.3333	38000	0.4029
0.4302	162.5	39000	0.4030
0.4307	166.6667	40000	0.4015
0.4321	170.8333	41000	0.3996
0.4301	175.0	42000	0.4018
0.4305	179.1667	43000	0.4000
0.4287	183.3333	44000	0.4005
0.4325	187.5	45000	0.4001
0.4287	191.6667	46000	0.3981
0.4272	195.8333	47000	0.3972
0.4249	200.0	48000	0.3968
0.427	204.1667	49000	0.3970
0.4271	208.3333	50000	0.3967
0.4247	212.5	51000	0.3958
0.4233	216.6667	52000	0.3953
0.4245	220.8333	53000	0.3956
0.4251	225.0	54000	0.3940
0.4266	229.1667	55000	0.3951
0.4203	233.3333	56000	0.3940
0.4235	237.5	57000	0.3934
0.4276	241.6667	58000	0.3929
0.4215	245.8333	59000	0.3938
0.4285	250.0	60000	0.3936
0.425	254.1667	61000	0.3936
0.4237	258.3333	62000	0.3933
0.4243	262.5	63000	0.3932

Framework versions

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