speecht5_fine_tuned_dhivehi_tts_3

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7593

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: 5e-05
• 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: 20000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.2327	5.2632	200	1.1988
1.2184	10.5263	400	1.1711
1.2007	15.7895	600	1.1521
1.1662	21.0526	800	1.1317
1.143	26.3158	1000	1.1228
1.1281	31.5789	1200	1.0680
1.1014	36.8421	1400	1.0495
1.0777	42.1053	1600	1.0578
1.067	47.3684	1800	1.0724
1.0389	52.6316	2000	1.0126
1.03	57.8947	2200	1.0089
1.0187	63.1579	2400	1.0203
1.0155	68.4211	2600	0.9795
0.9996	73.6842	2800	0.9728
0.9747	78.9474	3000	0.9522
0.9631	84.2105	3200	0.9460
0.9469	89.4737	3400	0.9317
0.935	94.7368	3600	0.9173
0.928	100.0	3800	0.9068
0.915	105.2632	4000	0.9426
0.9155	110.5263	4200	0.9063
0.9015	115.7895	4400	0.8896
0.8918	121.0526	4600	0.8787
0.8794	126.3158	4800	0.8818
0.8822	131.5789	5000	0.8811
0.8641	136.8421	5200	0.8719
0.8589	142.1053	5400	0.8590
0.8477	147.3684	5600	0.8575
0.8322	152.6316	5800	0.8720
0.8233	157.8947	6000	0.8520
0.8267	163.1579	6200	0.8496
0.8208	168.4211	6400	0.8556
0.8038	173.6842	6600	0.8512
0.8072	178.9474	6800	0.8318
0.7964	184.2105	7000	0.8299
0.7862	189.4737	7200	0.8261
0.7876	194.7368	7400	0.8247
0.7793	200.0	7600	0.8236
0.7753	205.2632	7800	0.8190
0.7675	210.5263	8000	0.8118
0.763	215.7895	8200	0.8198
0.757	221.0526	8400	0.8156
0.7538	226.3158	8600	0.7988
0.7481	231.5789	8800	0.8002
0.7423	236.8421	9000	0.8164
0.7481	242.1053	9200	0.8059
0.7385	247.3684	9400	0.7962
0.7304	252.6316	9600	0.8026
0.7238	257.8947	9800	0.7994
0.7233	263.1579	10000	0.7929
0.7205	268.4211	10200	0.7854
0.7131	273.6842	10400	0.7943
0.7118	278.9474	10600	0.7919
0.7037	284.2105	10800	0.7921
0.7079	289.4737	11000	0.7955
0.7011	294.7368	11200	0.7978
0.7051	300.0	11400	0.7845
0.7019	305.2632	11600	0.7853
0.6954	310.5263	11800	0.7724
0.6922	315.7895	12000	0.7752
0.6862	321.0526	12200	0.7825
0.6888	326.3158	12400	0.7777
0.6886	331.5789	12600	0.7740
0.679	336.8421	12800	0.7879
0.6756	342.1053	13000	0.7712
0.6778	347.3684	13200	0.7746
0.6702	352.6316	13400	0.7624
0.6743	357.8947	13600	0.7661
0.6727	363.1579	13800	0.7642
0.6643	368.4211	14000	0.7719
0.6651	373.6842	14200	0.7622
0.6632	378.9474	14400	0.7655
0.6572	384.2105	14600	0.7652
0.6575	389.4737	14800	0.7624
0.6581	394.7368	15000	0.7692
0.6581	400.0	15200	0.7635
0.6553	405.2632	15400	0.7558
0.6567	410.5263	15600	0.7507
0.6511	415.7895	15800	0.7633
0.6491	421.0526	16000	0.7558
0.6526	426.3158	16200	0.7643
0.6511	431.5789	16400	0.7577
0.6507	436.8421	16600	0.7571
0.6475	442.1053	16800	0.7535
0.6543	447.3684	17000	0.7593

Framework versions

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