learning_rate: 0.0001
train_batch_size: 8
eval_batch_size: 8
seed: 3407
gradient_accumulation_steps: 4
total_train_batch_size: 32
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: cosine
lr_scheduler_warmup_steps: 4000
training_steps: 40000
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
0.1265	111.1212	1000	0.1211
0.0925	222.2424	2000	0.1183
0.0851	333.3636	3000	0.1289
0.0675	444.4848	4000	0.1313
0.0621	555.6061	5000	0.1472
0.0535	666.7273	6000	0.1498
0.0594	777.8485	7000	0.1603
0.0463	888.9697	8000	0.1630
0.0431	1000.0	9000	0.1716
0.0488	1111.1212	10000	0.1718
0.0409	1222.2424	11000	0.1790
0.0393	1333.3636	12000	0.1840
0.0388	1444.4848	13000	0.1867
0.0384	1555.6061	14000	0.1926
0.0397	1666.7273	15000	0.1949
0.0331	1777.8485	16000	0.2052
0.0357	1888.9697	17000	0.1998
0.0332	2000.0	18000	0.2012
0.0311	2111.1212	19000	0.2058
0.0301	2222.2424	20000	0.2089
0.0285	2333.3636	21000	0.2130
0.0289	2444.4848	22000	0.2136
0.0299	2555.6061	23000	0.2148
0.0289	2666.7273	24000	0.2151
0.0278	2777.8485	25000	0.2170
0.0268	2888.9697	26000	0.2223
0.0262	3000.0	27000	0.2197
0.0373	3111.1212	28000	0.2234
0.0328	3222.2424	29000	0.2239
0.0268	3333.3636	30000	0.2254
0.0266	3444.4848	31000	0.2251
0.0264	3555.6061	32000	0.2280
0.0237	3666.7273	33000	0.2279
0.0291	3777.8485	34000	0.2260
0.0284	3888.9697	35000	0.2254
0.0241	4000.0	36000	0.2281
0.0262	4111.1212	37000	0.2263
0.0262	4222.2424	38000	0.2291
0.0254	4333.3636	39000	0.2276
0.0368	4444.4848	40000	0.2274

sil-ai/senga-nt-canon-mat-tts

senga-nt-canon-mat-tts