trocr-chaghatay – AI Model by colorlessideas | AlphaNeural AI

You can deploy this model and start earning money today!

colorlessideas/
trocr-chaghatay

like

0

vision-encoder-decoder

generated_from_trainer

microsoft/trocr-base-handwritten

endpoints_compatible

Views

No views yet

trocr-chaghatay

This model is a fine-tuned version of microsoft/trocr-base-handwritten on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.3978
Cer: 0.8823

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: 4
eval_batch_size: 4
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 16
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
num_epochs: 8
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
1.7914	0.9948	95	0.4213	0.9869
0.3869	1.9948	190	0.4202	0.9739
0.3777	2.9948	285	0.3962	0.9933
0.3555	3.9948	380	0.3880	0.9999
0.3512	4.9948	475	0.3807	0.9573
0.3407	5.9948	570	0.3769	0.8869
0.3266	6.9948	665	0.3661	0.8941
0.326	7.9948	760	0.3639	0.8975

Framework versions

Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.1
Tokenizers 0.21.1