donut-base-full_text_wt_val_1008

JandC/donut-base-full_text_wt_val_1008

donut-base-full_text_wt_val_1008 is a image-text-to-text model from JandC. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as mit.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

5% of all-time downloads

All-time downloads

201

Public

Parameters

202M

16.2 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors809 MB · 99%

Parameter types

How the weights are stored.

F32202M · 100%

Try a prompt

Base models

naver-clova-ix/donut-base

Datasets

imagefolder

Model card

From the Hugging Face model README

donut-base-full_text_wt_val_1008

This model is a fine-tuned version of naver-clova-ix/donut-base on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.1060

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: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.9227	0.2	100	0.6985
0.6852	0.4	200	0.4421
0.5102	0.6	300	0.3346
0.4178	0.79	400	0.2886
0.4476	0.99	500	0.2455
0.2931	1.19	600	0.2287
0.2647	1.39	700	0.2072
0.2418	1.59	800	0.1905
0.3031	1.79	900	0.1754
0.2306	1.98	1000	0.1667
0.2031	2.18	1100	0.1619
0.1918	2.38	1200	0.1536
0.1802	2.58	1300	0.1504
0.1646	2.78	1400	0.1436
0.1816	2.98	1500	0.1379
0.1344	3.17	1600	0.1395
0.1752	3.37	1700	0.1336
0.1388	3.57	1800	0.1306
0.1402	3.77	1900	0.1262
0.1123	3.97	2000	0.1277
0.144	4.17	2100	0.1248
0.1077	4.37	2200	0.1226
0.1134	4.56	2300	0.1186
0.1192	4.76	2400	0.1179
0.1142	4.96	2500	0.1194
0.1426	5.16	2600	0.1202
0.1022	5.36	2700	0.1165
0.0815	5.56	2800	0.1164
0.1096	5.75	2900	0.1166
0.0866	5.95	3000	0.1121
0.1148	6.15	3100	0.1122
0.0771	6.35	3200	0.1129
0.0996	6.55	3300	0.1096
0.0622	6.75	3400	0.1099
0.0985	6.94	3500	0.1092
0.0684	7.14	3600	0.1097
0.0669	7.34	3700	0.1086
0.0624	7.54	3800	0.1088
0.0763	7.74	3900	0.1069
0.0579	7.94	4000	0.1060
0.0623	8.13	4100	0.1083
0.0599	8.33	4200	0.1058
0.0625	8.53	4300	0.1073
0.0499	8.73	4400	0.1059
0.0628	8.93	4500	0.1059
0.0684	9.13	4600	0.1063
0.0472	9.33	4700	0.1056
0.068	9.52	4800	0.1057
0.06	9.72	4900	0.1062
0.0636	9.92	5000	0.1060

Framework versions

Transformers 4.38.0.dev0
Pytorch 2.1.0+cu121
Datasets 2.16.1
Tokenizers 0.15.1

Task: Image-Text-to-Text
Library: transformers
Type: vision-encoder-decoder
License: mit
Created: Jan 26, 2024
Updated: Jan 27, 2024