You can deploy this model and start earning money today!

rikeshVertex/
paddleocr-nepali-stage2-hw-digital-neg-fullft

Name: paddleocr-nepali-stage2-hw-digital-neg-fullft
Availability: InStock

Views

No views yet

paddleocr-nepali-stage2-hw-digital-neg-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: 6
eval_batch_size: 6
seed: 42
gradient_accumulation_steps: 6
total_train_batch_size: 36
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: 261
training_steps: 2612

Training Loss	Epoch	Step	Validation Loss
0.2476	0.2298	150	0.4683
0.1441	0.4597	300	0.2510
0.0969	0.6895	450	0.1371
0.0926	0.9193	600	0.1173
0.0695	1.1486	750	0.1015
0.0597	1.3784	900	0.0959
0.0594	1.6083	1050	0.0847
0.0541	1.8381	1200	0.0768
0.0529	2.0674	1350	0.0749
0.0503	2.2972	1500	0.0772
0.0556	2.5271	1650	0.0748
0.0508	2.7569	1800	0.0701
0.0483	2.9867	1950	0.0630
0.0378	3.2160	2100	0.0693
0.0373	3.4459	2250	0.0636
0.0530	3.6757	2400	0.0642
0.0496	3.9055	2550	0.0647
0.0452	4.0	2612	0.0665