speecht5_finetuned_storytel

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: 0.4020

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: 10000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.6213	0.1856	100	0.5374
0.5587	0.3711	200	0.4969
0.5245	0.5567	300	0.4817
0.5127	0.7423	400	0.4697
0.5044	0.9279	500	0.4636
0.5006	1.1141	600	0.4547
0.4924	1.2997	700	0.4512
0.4885	1.4853	800	0.4486
0.4805	1.6708	900	0.4509
0.4807	1.8564	1000	0.4429
0.4794	2.0427	1100	0.4414
0.4748	2.2283	1200	0.4390
0.4697	2.4138	1300	0.4357
0.4713	2.5994	1400	0.4365
0.466	2.7850	1500	0.4344
0.4688	2.9705	1600	0.4335
0.4627	3.1568	1700	0.4302
0.4583	3.3424	1800	0.4273
0.4614	3.5280	1900	0.4308
0.4599	3.7135	2000	0.4254
0.4607	3.8991	2100	0.4323
0.4576	4.0854	2200	0.4366
0.4526	4.2709	2300	0.4257
0.4545	4.4565	2400	0.4234
0.4582	4.6421	2500	0.4248
0.4509	4.8277	2600	0.4252
0.4458	5.0139	2700	0.4180
0.441	5.1995	2800	0.4222
0.4443	5.3851	2900	0.4172
0.4457	5.5706	3000	0.4193
0.444	5.7562	3100	0.4238
0.4414	5.9418	3200	0.4211
0.4392	6.1280	3300	0.4176
0.4446	6.3136	3400	0.4161
0.4428	6.4992	3500	0.4187
0.4409	6.6848	3600	0.4156
0.4363	6.8703	3700	0.4143
0.4426	7.0566	3800	0.4161
0.4344	7.2422	3900	0.4149
0.4327	7.4277	4000	0.4115
0.4318	7.6133	4100	0.4154
0.4401	7.7989	4200	0.4148
0.4328	7.9845	4300	0.4094
0.4336	8.1707	4400	0.4136
0.433	8.3563	4500	0.4089
0.438	8.5419	4600	0.4092
0.4336	8.7274	4700	0.4081
0.4335	8.9130	4800	0.4079
0.4277	9.0993	4900	0.4093
0.4268	9.2849	5000	0.4106
0.4269	9.4704	5100	0.4126
0.4247	9.6560	5200	0.4110
0.4194	9.8416	5300	0.4086
0.4226	10.0278	5400	0.4089
0.4249	10.2134	5500	0.4075
0.4225	10.3990	5600	0.4082
0.4214	10.5846	5700	0.4137
0.4245	10.7701	5800	0.4023
0.4243	10.9557	5900	0.4047
0.4196	11.1420	6000	0.4056
0.4176	11.3275	6100	0.4038
0.4216	11.5131	6200	0.4103
0.4201	11.6987	6300	0.4051
0.415	11.8842	6400	0.4061
0.4152	12.0705	6500	0.4053
0.4174	12.2561	6600	0.4060
0.4104	12.4417	6700	0.4038
0.4143	12.6272	6800	0.4091
0.4172	12.8128	6900	0.4028
0.4122	12.9984	7000	0.4078
0.4136	13.1846	7100	0.4055
0.4101	13.3702	7200	0.4047
0.4096	13.5558	7300	0.4051
0.4137	13.7414	7400	0.4047
0.4088	13.9269	7500	0.4051
0.4096	14.1132	7600	0.4037
0.4117	14.2988	7700	0.4043
0.4076	14.4843	7800	0.4042
0.4086	14.6699	7900	0.4052
0.4062	14.8555	8000	0.4006
0.409	15.0418	8100	0.4026
0.4056	15.2273	8200	0.4018
0.4048	15.4129	8300	0.4019
0.4084	15.5985	8400	0.4018
0.4056	15.7840	8500	0.4004
0.4157	15.9696	8600	0.4033
0.4024	16.1559	8700	0.4019
0.4064	16.3415	8800	0.4033
0.4027	16.5270	8900	0.4050
0.4041	16.7126	9000	0.4009
0.4028	16.8982	9100	0.4036
0.3979	17.0844	9200	0.4025
0.4008	17.2700	9300	0.4019
0.4012	17.4556	9400	0.4000
0.399	17.6412	9500	0.4001
0.3996	17.8267	9600	0.3999
0.4009	18.0130	9700	0.4017
0.4034	18.1986	9800	0.4019
0.3963	18.3841	9900	0.4017
0.3999	18.5697	10000	0.4020

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.3