h6_f1536_l16_d384_trained_7

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0220
• Per Phonemes: 0.0082
• Per Hams Or Jahr: 0.0230
• Per Shidda Or Rakhawa: 0.0065
• Per Tafkheem Or Taqeeq: 0.0034
• Per Itbaq: 0.0049
• Per Safeer: 0.0037
• Per Qalqla: 0.0025
• Per Tikraar: 0.0101
• Per Tafashie: 0.0018
• Per Istitala: 0.0020
• Per Ghonna: 0.0038
• Average Per: 0.0064

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: 10
• eval_batch_size: 10
• 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: constant
• lr_scheduler_warmup_steps: 0.2
• num_epochs: 7

Training results

Training Loss	Epoch	Step	Average Per	Validation Loss	Per Ghonna	Per Hams Or Jahr	Per Istitala	Per Itbaq	Per Phonemes	Per Qalqla	Per Safeer	Per Shidda Or Rakhawa	Per Tafashie	Per Tafkheem Or Taqeeq	Per Tikraar
0.7398	1.0	19409	0.0273	0.1096	0.0175	0.0784	0.0093	0.0226	0.0397	0.0120	0.0296	0.0330	0.0089	0.0190	0.0307
0.0807	2.0	38818	0.0135	0.0588	0.0097	0.0417	0.0053	0.0095	0.0224	0.0065	0.0109	0.0169	0.0051	0.0094	0.0107
0.0501	3.0	58227	0.0103	0.0428	0.0075	0.0346	0.0038	0.0070	0.0166	0.0048	0.0067	0.0122	0.0035	0.0069	0.0102
0.0369	4.0	77636	0.0082	0.0340	0.0061	0.0258	0.0034	0.0060	0.0126	0.0043	0.0055	0.0101	0.0030	0.0056	0.0083
0.0294	5.0	97045	0.0090	0.0269	0.0053	0.0257	0.0030	0.0120	0.0104	0.0053	0.0076	0.0079	0.0027	0.0043	0.0142
0.0244	6.0	116454	0.0242	0.0092	0.0241	0.0072	0.0038	0.0121	0.0064	0.0054	0.0171	0.0023	0.0023	0.0043	0.0086
0.0212	7.0	135863	0.0220	0.0082	0.0230	0.0065	0.0034	0.0049	0.0037	0.0025	0.0101	0.0018	0.0020	0.0038	0.0064

Framework versions

• Transformers 5.9.0
• Pytorch 2.12.0+cu130
• Datasets 4.8.5
• Tokenizers 0.22.2