speecht5_dhivehi_tts_v4_from_scratch

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

• Loss: 0.6625

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: 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: 50000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.4206	2.8653	1000	1.3317
1.3805	5.7307	2000	1.2204
1.2854	8.5960	3000	1.2058
1.2263	11.4613	4000	1.1523
1.1856	14.3266	5000	1.1170
1.1581	17.1920	6000	1.0919
1.1352	20.0573	7000	1.0717
1.1101	22.9226	8000	1.0420
1.0812	25.7880	9000	1.0289
1.0617	28.6533	10000	1.0143
1.0408	31.5186	11000	0.9820
1.0162	34.3840	12000	0.9893
0.9929	37.2493	13000	0.9227
0.9724	40.1146	14000	0.9315
0.9545	42.9799	15000	0.9066
0.9334	45.8453	16000	0.8860
0.9181	48.7106	17000	0.8864
0.897	51.5759	18000	0.8532
0.8826	54.4413	19000	0.8283
0.868	57.3066	20000	0.8417
0.8535	60.1719	21000	0.7956
0.8407	63.0372	22000	0.8141
0.8262	65.9026	23000	0.7863
0.8127	68.7679	24000	0.7878
0.8023	71.6332	25000	0.7542
0.7907	74.4986	26000	0.7792
0.7788	77.3639	27000	0.7410
0.7713	80.2292	28000	0.7459
0.764	83.0946	29000	0.7247
0.7568	85.9599	30000	0.7291
0.7413	88.8252	31000	0.7108
0.737	91.6905	32000	0.7147
0.7317	94.5559	33000	0.6972
0.7253	97.4212	34000	0.6967
0.7205	100.2865	35000	0.6842
0.7114	103.1519	36000	0.6948
0.7053	106.0172	37000	0.6763
0.7027	108.8825	38000	0.6870
0.7002	111.7479	39000	0.6727
0.6965	114.6132	40000	0.6796
0.6928	117.4785	41000	0.6664
0.6948	120.3438	42000	0.6720
0.6883	123.2092	43000	0.6635
0.6857	126.0745	44000	0.6672
0.6822	128.9398	45000	0.6607
0.6847	131.8052	46000	0.6699
0.6843	134.6705	47000	0.6610
0.6825	137.5358	48000	0.6690
0.683	140.4011	49000	0.6604
0.682	143.2665	50000	0.6625

Framework versions

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