xls-r-300m-toratan-240

This model is a fine-tuned version of Benjaminpwh/xlsr-toratan-240-copt-base_K on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0082
• Cer: 0.0025

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• 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: 500
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
4.2511	2.0833	400	1.9040	0.5065
1.6576	4.1667	800	1.1012	0.3104
1.2488	6.25	1200	0.8374	0.2464
0.9964	8.3333	1600	0.6042	0.1834
0.8249	10.4167	2000	0.4650	0.1583
0.677	12.5	2400	0.3377	0.1236
0.5339	14.5833	2800	0.2268	0.0833
0.4255	16.6667	3200	0.1452	0.0540
0.3489	18.75	3600	0.0946	0.0386
0.2719	20.8333	4000	0.0551	0.0187
0.2151	22.9167	4400	0.0290	0.0112
0.1854	25.0	4800	0.0187	0.0069
0.1468	27.0833	5200	0.0127	0.0047
0.1284	29.1667	5600	0.0082	0.0025

Framework versions

• Transformers 4.52.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1