You can deploy this model and start earning money today!

rikeshVertex/
paddleocr-nepali-stage2-handwritten-fullft

Name: paddleocr-nepali-stage2-handwritten-fullft
Availability: InStock

Views

No views yet

paddleocr-nepali-stage2-handwritten-fullft

This model is a fine-tuned version of rikeshVertex/paddleocr-nepali-stage1-merged on an unknown dataset. It achieves the following results on the evaluation set:

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 2e-05
train_batch_size: 4
eval_batch_size: 4
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 32
optimizer: Use OptimizerNames.PAGED_ADAMW with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 182
training_steps: 1820

Training Loss	Epoch	Step	Validation Loss
1.3594	0.2199	100	0.8548
0.4657	0.4398	200	0.5001
0.3191	0.6597	300	0.3355
0.3135	0.8796	400	0.2572
0.2543	1.0990	500	0.2040
0.1428	1.3189	600	0.1798
0.1596	1.5388	700	0.1569
0.0850	1.7587	800	0.1463
0.0741	1.9786	900	0.1291
0.1349	2.1979	1000	0.1216
0.1149	2.4178	1100	0.1147
0.0787	2.6377	1200	0.1133
0.0954	2.8576	1300	0.1069
0.0598	3.0770	1400	0.1033
0.0809	3.2969	1500	0.1009
0.0613	3.5168	1600	0.1003
0.0596	3.7367	1700	0.0999
0.0581	3.9566	1800	0.0998
0.0616	4.0	1820	0.0998