Layout-finetuned-fr-model-25instances25-100epochs-5e-05lr-GPU

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

• Loss: 0.0000
• Accuracy: 1.0
• Learning Rate: 5e-05

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: 8
• seed: 42
• 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: reduce_lr_on_plateau
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Learning Rate
5.3733	1.4286	10	4.1888	0.42	5e-05
3.7229	2.8571	20	2.2889	0.56	5e-05
2.06	4.2857	30	0.8765	0.74	5e-05
1.0064	5.7143	40	0.6363	0.76	5e-05
1.0577	7.1429	50	0.4538	0.88	5e-05
0.47	8.5714	60	0.3682	0.86	5e-05
0.6724	10.0	70	0.1981	0.9	5e-05
0.5865	11.4286	80	0.1792	0.9	5e-05
0.3017	12.8571	90	0.0972	0.96	5e-05
0.1234	14.2857	100	0.0420	0.98	5e-05
0.1442	15.7143	110	0.0221	1.0	5e-05
0.8793	17.1429	120	0.0764	0.98	5e-05
0.0422	18.5714	130	0.1742	0.96	5e-05
0.2005	20.0	140	0.0933	0.96	5e-05
0.0889	21.4286	150	0.0024	1.0	5e-05
0.007	22.8571	160	0.0014	1.0	5e-05
0.135	24.2857	170	0.0007	1.0	5e-05
0.0018	25.7143	180	0.0008	1.0	5e-05
0.0026	27.1429	190	0.0003	1.0	5e-05
0.0019	28.5714	200	0.0006	1.0	5e-05
0.0037	30.0	210	0.0003	1.0	5e-05
0.0008	31.4286	220	0.0029	1.0	5e-05
0.0128	32.8571	230	0.0002	1.0	5e-05
0.0695	34.2857	240	0.0001	1.0	5e-05
0.0003	35.7143	250	0.0002	1.0	5e-05
0.0011	37.1429	260	0.0001	1.0	5e-05
0.0002	38.5714	270	0.0001	1.0	5e-05
0.0002	40.0	280	0.0001	1.0	5e-05
0.0002	41.4286	290	0.0000	1.0	5e-05
0.0001	42.8571	300	0.0000	1.0	5e-05
0.0001	44.2857	310	0.0000	1.0	5e-05
0.0001	45.7143	320	0.0000	1.0	5e-05
0.0001	47.1429	330	0.0000	1.0	5e-05
0.0001	48.5714	340	0.0000	1.0	5e-05
0.0001	50.0	350	0.0000	1.0	5e-05
0.0001	51.4286	360	0.0000	1.0	5e-05
0.0001	52.8571	370	0.0000	1.0	5e-05
0.0001	54.2857	380	0.0000	1.0	5e-05
0.0001	55.7143	390	0.0000	1.0	5e-05
0.0001	57.1429	400	0.0000	1.0	5e-05
0.0001	58.5714	410	0.0000	1.0	5e-05
0.0001	60.0	420	0.0000	1.0	5e-05
0.0001	61.4286	430	0.0000	1.0	5e-05
0.0001	62.8571	440	0.0000	1.0	5e-05
0.0001	64.2857	450	0.0000	1.0	5e-05
0.0001	65.7143	460	0.0000	1.0	5e-05
0.0001	67.1429	470	0.0000	1.0	5e-05
0.0	68.5714	480	0.0000	1.0	5e-05
0.0001	70.0	490	0.0000	1.0	5e-05
0.0001	71.4286	500	0.0000	1.0	5e-05
0.0001	72.8571	510	0.0000	1.0	5e-05
0.0	74.2857	520	0.0000	1.0	5e-05
0.0001	75.7143	530	0.0000	1.0	5e-05
0.0001	77.1429	540	0.0000	1.0	5e-05
0.0001	78.5714	550	0.0000	1.0	5e-05
0.0001	80.0	560	0.0000	1.0	5e-05
0.0001	81.4286	570	0.0000	1.0	5e-05
0.0	82.8571	580	0.0000	1.0	5e-05
0.0	84.2857	590	0.0000	1.0	5e-05
0.0	85.7143	600	0.0000	1.0	5e-05
0.0	87.1429	610	0.0000	1.0	5e-05
0.0	88.5714	620	0.0000	1.0	5e-05
0.0	90.0	630	0.0000	1.0	5e-05
0.0	91.4286	640	0.0000	1.0	5e-05
0.0	92.8571	650	0.0000	1.0	5e-05
0.0	94.2857	660	0.0000	1.0	5e-05
0.0	95.7143	670	0.0000	1.0	5e-05
0.0001	97.1429	680	0.0000	1.0	5e-05
0.0	98.5714	690	0.0000	1.0	5e-05
0.0	100.0	700	0.0000	1.0	5e-05

Framework versions

• Transformers 4.48.1
• Pytorch 2.3.1.post300
• Datasets 3.2.0
• Tokenizers 0.21.0