urdu-tts-phonemes-finetuned-extended

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8014

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-06
• train_batch_size: 6
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• gradient_accumulation_steps: 4
• total_train_batch_size: 48
• total_eval_batch_size: 4
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 500
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
4.6038	0.9149	500	0.9382
3.9723	1.8289	1000	0.8982
3.8796	2.7429	1500	0.8837
3.7655	3.6569	2000	0.8691
3.7279	4.5709	2500	0.8611
3.6874	5.4849	3000	0.8507
3.6253	6.3989	3500	0.8450
3.5747	7.3129	4000	0.8403
3.5732	8.2269	4500	0.8419
3.5595	9.1409	5000	0.8367
3.5520	10.0549	5500	0.8339
3.5297	10.9698	6000	0.8357
3.5303	11.8838	6500	0.8333
3.5060	12.7978	7000	0.8260
3.4967	13.7118	7500	0.8256
3.5028	14.6258	8000	0.8239
3.4786	15.5398	8500	0.8211
3.4918	16.4538	9000	0.8209
3.4901	17.3678	9500	0.8193
3.4815	18.2818	10000	0.8204
3.4688	19.1958	10500	0.8163
3.4426	20.1098	11000	0.8160
3.4519	21.0238	11500	0.8157
3.4514	21.9387	12000	0.8138
3.4341	22.8527	12500	0.8148
3.4341	23.7667	13000	0.8114
3.4261	24.6807	13500	0.8129
3.4276	25.5947	14000	0.8111
3.4234	26.5087	14500	0.8086
3.4379	27.4227	15000	0.8083
3.4171	28.3367	15500	0.8055
3.4083	29.2507	16000	0.8067
3.4227	30.1647	16500	0.8090
3.4050	31.0787	17000	0.8063
3.4144	31.9936	17500	0.8072
3.3968	32.9076	18000	0.8028
3.4001	33.8216	18500	0.8056
3.3873	34.7356	19000	0.8032
3.4032	35.6496	19500	0.8057
3.3936	36.5636	20000	0.8040
3.3829	37.4776	20500	0.8036
3.3927	38.3916	21000	0.8034
3.3895	39.3056	21500	0.8037
3.3789	40.2196	22000	0.8027
3.3938	41.1336	22500	0.8020
3.3907	42.0476	23000	0.8018
3.3752	42.9625	23500	0.8004
3.3759	43.8765	24000	0.8009
3.3807	44.7905	24500	0.8015
3.3808	45.7045	25000	0.8030
3.3720	46.6185	25500	0.8016
3.3851	47.5325	26000	0.8018
3.3733	48.4465	26500	0.8007
3.3703	49.3605	27000	0.8014

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.3
• Tokenizers 0.22.2