learning_rate: 1e-05
train_batch_size: 1
eval_batch_size: 1
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 100
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Token Accuracy
11.1712	1.0	9	10.5461	0.0045	0.0476
7.915	2.0	18	8.6977	0.0045	0.3333
6.8327	3.0	27	6.2592	0.0045	0.3333
4.9657	4.0	36	4.7501	0.0045	0.3333
4.059	5.0	45	3.8129	0.0045	0.3333
2.8765	6.0	54	2.6799	0.0045	0.1905
2.3787	7.0	63	1.9647	0.0045	0.1429
1.9747	8.0	72	1.5757	0.0045	0.1905
1.9692	9.0	81	1.2219	0.0045	0.1429
1.5935	10.0	90	1.0530	0.0045	0.2381
1.4664	11.0	99	0.9725	0.0045	0.2381
1.3034	12.0	108	0.8732	0.0045	0.1905
1.1197	13.0	117	0.7470	0.0045	0.1905
1.0003	14.0	126	0.7166	0.0045	0.2857
0.9327	15.0	135	0.6241	0.0045	0.1905
0.8005	16.0	144	0.5790	0.0045	0.2857
0.8996	17.0	153	0.4649	0.0045	0.1905
0.7476	18.0	162	0.4327	0.0045	0.2381
0.6274	19.0	171	0.3926	0.0045	0.2381
0.5374	20.0	180	0.2948	0.0045	0.3810
0.5547	21.0	189	0.2403	0.0045	0.3333
0.47	22.0	198	0.2102	0.0045	0.3333
0.5181	23.0	207	0.2081	0.0045	0.3810
0.4835	24.0	216	0.1405	0.0045	0.3333
0.3301	25.0	225	0.1212	0.0045	0.3333
0.2952	26.0	234	0.0825	0.0045	0.2381
0.2618	27.0	243	0.0712	0.0045	0.2857
0.2666	28.0	252	0.1157	0.0045	0.2857
0.1654	29.0	261	0.0490	0.0045	0.2381
0.1697	30.0	270	0.0452	0.0045	0.3333
0.1484	31.0	279	0.0448	0.0045	0.2857
0.1946	32.0	288	0.0322	0.0045	0.3810
0.2329	33.0	297	0.0477	0.0045	0.3333
0.195	34.0	306	0.0279	0.0045	0.3810
0.1196	35.0	315	0.0209	0.0045	0.3333
0.0658	36.0	324	0.0219	0.0045	0.3810
0.0812	37.0	333	0.0171	0.0045	0.3333
0.136	38.0	342	0.0196	0.0045	0.3810
0.1589	39.0	351	0.0281	0.0045	0.3333
0.0814	40.0	360	0.0187	0.0045	0.3333
0.0547	41.0	369	0.0149	0.0045	0.3333
0.0711	42.0	378	0.0133	0.0045	0.2857
0.0965	43.0	387	0.0123	0.0045	0.2381
0.0523	44.0	396	0.0134	0.0045	0.2381
0.0341	45.0	405	0.0107	0.0045	0.2381
0.0576	46.0	414	0.0102	0.0045	0.2857
0.048	47.0	423	0.0133	0.0045	0.2857
0.0668	48.0	432	0.0120	0.0045	0.2857
0.0594	49.0	441	0.0127	0.0045	0.1905
0.0383	50.0	450	0.0099	0.0045	0.2381
0.0367	51.0	459	0.0100	0.0045	0.3810
0.0224	52.0	468	0.0094	0.0045	0.3810
0.036	53.0	477	0.0084	0.0045	0.3333
0.0428	54.0	486	0.0099	0.0045	0.3333
0.0563	55.0	495	0.0064	0.0045	0.3333
0.0262	56.0	504	0.0077	0.0045	0.3333
0.017	57.0	513	0.0055	0.0045	0.3333
0.0393	58.0	522	0.0111	0.0045	0.3333
0.0202	59.0	531	0.0045	0.0045	0.3333
0.0172	60.0	540	0.0055	0.0045	0.3333
0.0339	61.0	549	0.0121	0.0045	0.2857
0.0181	62.0	558	0.0039	0.0045	0.2857
0.0054	63.0	567	0.0041	0.0045	0.2381
0.0155	64.0	576	0.0039	0.0045	0.3333
0.0068	65.0	585	0.0040	0.0045	0.2857
0.0352	66.0	594	0.0019	0.0045	0.2381
0.0213	67.0	603	0.0041	0.0045	0.2381
0.0184	68.0	612	0.0015	0.0045	0.3333
0.008	69.0	621	0.0019	0.0045	0.3333
0.0188	70.0	630	0.0060	0.0045	0.3333
0.0147	71.0	639	0.0020	0.0045	0.2857
0.0093	72.0	648	0.0008	0.0045	0.2381
0.0172	73.0	657	0.0015	0.0045	0.2381
0.0166	74.0	666	0.0016	0.0045	0.3333
0.0136	75.0	675	0.0005	0.0045	0.3333
0.0178	76.0	684	0.0006	0.0045	0.2381
0.0054	77.0	693	0.0005	0.0045	0.2857
0.0205	78.0	702	0.0029	0.0045	0.2857
0.0046	79.0	711	0.0009	0.0045	0.2857
0.0239	80.0	720	0.0006	0.0045	0.2381
0.0121	81.0	729	0.0011	0.0045	0.2381
0.0168	82.0	738	0.0014	0.0045	0.2857
0.0088	83.0	747	0.0005	0.0045	0.2857
0.0073	84.0	756	0.0004	0.0045	0.2381
0.011	85.0	765	0.0006	0.0045	0.2381
0.01	86.0	774	0.0004	0.0045	0.3333
0.0031	87.0	783	0.0005	0.0045	0.2381
0.0136	88.0	792	0.0009	0.0045	0.2381
0.0105	89.0	801	0.0013	0.0045	0.2381
0.004	90.0	810	0.0009	0.0045	0.3333
0.0039	91.0	819	0.0006	0.0045	0.3333
0.0069	92.0	828	0.0004	0.0045	0.3333
0.0075	93.0	837	0.0003	0.0045	0.3333
0.0058	94.0	846	0.0003	0.0045	0.3333
0.0125	95.0	855	0.0004	0.0045	0.3333
0.0158	96.0	864	0.0003	0.0045	0.3333
0.016	97.0	873	0.0003	0.0045	0.3333
0.0037	98.0	882	0.0003	0.0045	0.3333
0.0059	99.0	891	0.0003	0.0045	0.3333
0.0049	100.0	900	0.0003	0.0045	0.3333

Framework versions

Transformers 4.56.2
Pytorch 2.8.0+cu128
Datasets 4.1.1
Tokenizers 0.22.1

Cyber-Zhaba/donut-metric-tables-boxes

donut-metric-tables-boxes

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Cyber-Zhaba/
donut-metric-tables-boxes