2024-01-09_one_stage_subgraphs_weighted_entropyreg_txt_vis_conc_1_4_8_12_gate

This model is a fine-tuned version of microsoft/layoutlmv3-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0736
• Accuracy: 0.755
• Exit 0 Accuracy: 0.0625
• Exit 1 Accuracy: 0.0725
• Exit 2 Accuracy: 0.0625
• Exit 3 Accuracy: 0.0625
• Exit 4 Accuracy: 0.755

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: 2e-05
• train_batch_size: 2
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 24
• total_train_batch_size: 48
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 60

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Exit 0 Accuracy	Exit 1 Accuracy	Exit 2 Accuracy	Exit 3 Accuracy	Exit 4 Accuracy
No log	0.96	16	2.6720	0.1675	0.06	0.08	0.0625	0.0625	0.1675
No log	1.98	33	2.5560	0.235	0.06	0.0725	0.0625	0.0625	0.235
No log	3.0	50	2.4719	0.2475	0.0675	0.0775	0.0625	0.0625	0.2475
No log	3.96	66	2.3389	0.31	0.0675	0.08	0.0625	0.0625	0.31
No log	4.98	83	2.2557	0.34	0.07	0.0775	0.0625	0.0625	0.34
No log	6.0	100	2.2581	0.3075	0.07	0.0675	0.0625	0.0625	0.3075
No log	6.96	116	2.0395	0.4175	0.065	0.0675	0.0625	0.0625	0.4175
No log	7.98	133	1.8828	0.5025	0.0625	0.0725	0.0625	0.0625	0.5025
No log	9.0	150	1.7408	0.545	0.0625	0.0775	0.0625	0.0625	0.545
No log	9.96	166	1.5987	0.5975	0.0625	0.07	0.0625	0.0625	0.5975
No log	10.98	183	1.4925	0.63	0.0625	0.07	0.0625	0.0625	0.63
No log	12.0	200	1.2624	0.705	0.0625	0.07	0.0625	0.0625	0.705
No log	12.96	216	1.2142	0.7125	0.0625	0.0675	0.0625	0.0625	0.7125
No log	13.98	233	1.1393	0.7225	0.0625	0.0675	0.0625	0.0625	0.7225
No log	15.0	250	1.1158	0.7125	0.0625	0.0675	0.0625	0.0625	0.7125
No log	15.96	266	1.0664	0.7075	0.0625	0.0625	0.0625	0.0625	0.7075
No log	16.98	283	1.0757	0.7325	0.0625	0.0625	0.0625	0.0625	0.7325
No log	18.0	300	1.0319	0.73	0.0625	0.0675	0.0625	0.0625	0.73
No log	18.96	316	1.0531	0.735	0.0625	0.07	0.0625	0.0625	0.735
No log	19.98	333	1.0393	0.72	0.0625	0.065	0.0625	0.0625	0.72
No log	21.0	350	1.0517	0.7225	0.0625	0.0675	0.0625	0.0625	0.7225
No log	21.96	366	1.0466	0.73	0.0625	0.0625	0.0625	0.0625	0.73
No log	22.98	383	1.0608	0.7275	0.0625	0.07	0.0625	0.0625	0.7275
No log	24.0	400	1.0289	0.7325	0.0625	0.0625	0.0625	0.0625	0.7325
No log	24.96	416	1.0302	0.7375	0.0625	0.0675	0.0625	0.0625	0.7375
No log	25.98	433	1.0541	0.7325	0.0625	0.065	0.0625	0.0625	0.7325
No log	27.0	450	1.0866	0.7375	0.0625	0.0625	0.0625	0.0625	0.7375
No log	27.96	466	1.0708	0.725	0.0625	0.065	0.0625	0.0625	0.725
No log	28.98	483	1.0817	0.735	0.0625	0.065	0.0625	0.0625	0.735
0.5648	30.0	500	1.0515	0.74	0.0625	0.065	0.0625	0.0625	0.74
0.5648	30.96	516	1.0553	0.735	0.0625	0.065	0.0625	0.0625	0.735
0.5648	31.98	533	1.0368	0.7475	0.0625	0.0625	0.0625	0.0625	0.7475
0.5648	33.0	550	1.0459	0.75	0.0625	0.0725	0.0625	0.0625	0.75
0.5648	33.96	566	1.0581	0.745	0.0625	0.07	0.0625	0.0625	0.745
0.5648	34.98	583	1.0532	0.7525	0.0625	0.07	0.0625	0.0625	0.7525
0.5648	36.0	600	1.0679	0.75	0.0625	0.0725	0.0625	0.0625	0.75
0.5648	36.96	616	1.0850	0.7425	0.0625	0.0725	0.0625	0.0625	0.7425
0.5648	37.98	633	1.0780	0.755	0.0625	0.0725	0.0625	0.0625	0.755
0.5648	39.0	650	1.0592	0.75	0.0625	0.07	0.0625	0.0625	0.75
0.5648	39.96	666	1.0790	0.75	0.0625	0.07	0.0625	0.0625	0.75
0.5648	40.98	683	1.0661	0.7525	0.0625	0.07	0.0625	0.0625	0.7525
0.5648	42.0	700	1.0841	0.7475	0.0625	0.07	0.0625	0.0625	0.7475
0.5648	42.96	716	1.0792	0.7475	0.0625	0.07	0.0625	0.0625	0.7475
0.5648	43.98	733	1.0659	0.7525	0.0625	0.07	0.0625	0.0625	0.7525
0.5648	45.0	750	1.0794	0.75	0.0625	0.07	0.0625	0.0625	0.75
0.5648	45.96	766	1.0712	0.745	0.0625	0.07	0.0625	0.0625	0.745
0.5648	46.98	783	1.0714	0.745	0.0625	0.07	0.0625	0.0625	0.745
0.5648	48.0	800	1.0710	0.75	0.0625	0.07	0.0625	0.0625	0.75
0.5648	48.96	816	1.0711	0.75	0.0625	0.07	0.0625	0.0625	0.75
0.5648	49.98	833	1.0669	0.7525	0.0625	0.0725	0.0625	0.0625	0.7525
0.5648	51.0	850	1.0751	0.7525	0.0625	0.07	0.0625	0.0625	0.7525
0.5648	51.96	866	1.0712	0.7575	0.0625	0.07	0.0625	0.0625	0.7575
0.5648	52.98	883	1.0700	0.7525	0.0625	0.07	0.0625	0.0625	0.7525
0.5648	54.0	900	1.0707	0.7525	0.0625	0.0725	0.0625	0.0625	0.7525
0.5648	54.96	916	1.0718	0.755	0.0625	0.0725	0.0625	0.0625	0.755
0.5648	55.98	933	1.0732	0.755	0.0625	0.0725	0.0625	0.0625	0.755
0.5648	57.0	950	1.0736	0.755	0.0625	0.0725	0.0625	0.0625	0.755
0.5648	57.6	960	1.0736	0.755	0.0625	0.0725	0.0625	0.0625	0.755

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu117
• Datasets 2.13.1
• Tokenizers 0.13.3