You can deploy this model and start earning money today!

pneupane/
table-transformer-finetuned

Name: table-transformer-finetuned
Availability: InStock

Views

No views yet

table-transformer-finetuned

This model is a fine-tuned version of microsoft/table-structure-recognition-v1.1-all 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: 1e-06
train_batch_size: 1
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 8
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: 20
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
9.2523	1.0	14	9.3207
7.9949	2.0	28	9.0533
8.2292	3.0	42	8.6547
7.1495	4.0	56	8.0844
6.9055	5.0	70	7.3339
7.0771	6.0	84	7.2344
5.6554	7.0	98	6.6026
5.5633	8.0	112	6.3610
4.9506	9.0	126	5.9567
4.0856	10.0	140	5.9191
4.4286	11.0	154	5.6374
2.9043	12.0	168	4.9525
3.4755	13.0	182	4.6846
3.4152	14.0	196	4.3171
2.8456	15.0	210	3.5661
2.4149	16.0	224	3.6150
1.9328	17.0	238	3.2264
1.8503	18.0	252	2.8540
1.596	18.5981	260	2.2890