mt5_hieroglyph

This model is a fine-tuned version of Asiif/mt5_hieroglyph on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.3105
• Bleu: 71.4581
• Chrf: 77.3596

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• 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_ratio: 0.08
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Chrf
0.077	10.0	6470	2.1542	70.7322	76.6222
0.0653	20.0	12940	2.3105	71.4581	77.3596
0.0593	30.0	19410	2.4402	71.1786	77.0790
0.097	40.0	25880	2.0266	70.9717	77.2137
0.1197	50.0	32350	1.7680	71.1443	77.4270

Framework versions

• Transformers 4.57.6
• Pytorch 2.9.0+cu126
• Datasets 3.1.0
• Tokenizers 0.22.2