learning_rate: 1e-05
train_batch_size: 16
eval_batch_size: 2
seed: 42
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: 1000
training_steps: 100000
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer	Avg Precision Exact	Avg Recall Exact	Avg F1 Exact	Avg Precision Letter Shift	Avg Recall Letter Shift	Avg F1 Letter Shift	Avg Precision Word Level	Avg Recall Word Level	Avg F1 Word Level	Avg Precision Word Shift	Avg Recall Word Shift	Avg F1 Word Shift	Precision Median Exact	Recall Median Exact	F1 Median Exact	Precision Max Exact	Recall Max Exact	F1 Max Exact
0.0276	3.0030	20000	3.0853	98.7760	0.0336	0.0472	0.0380	0.0581	0.0831	0.0658	0.0801	0.1144	0.0904	0.2032	0.3029	0.2334	0.0213	0.0303	0.0267	0.6667	0.6667	0.6667
0.0026	6.0060	40000	3.8257	98.8227	0.0303	0.0497	0.0364	0.0528	0.0880	0.0637	0.0728	0.1208	0.0877	0.1812	0.3095	0.2204	0.0	0.0	0.0	0.75	0.75	0.75
0.0004	9.0090	60000	4.3074	99.0124	0.0283	0.0504	0.0347	0.0476	0.0872	0.0590	0.0641	0.1169	0.0792	0.1600	0.2986	0.1987	0.0	0.0	0.0	1.0	1.0	0.8
0.0001	12.0120	80000	4.6534	98.8315	0.0295	0.0516	0.0361	0.0502	0.0882	0.0611	0.0687	0.1192	0.0826	0.1715	0.3112	0.2103	0.0	0.0	0.0	1.0	1.0	1.0
0.0	15.0150	100000	4.7623	98.8781	0.0291	0.0505	0.0356	0.0493	0.0876	0.0605	0.0685	0.1200	0.0833	0.1707	0.3059	0.2088	0.0	0.0	0.0	0.4444	1.0	0.4286

cantillation/Teamim-tiny_WeightDecay-0.005_Augmented__date-10-04-2025

he-cantillation