git-large-r-coco-IDB_ADv1_COCO

ooliverz/git-large-r-coco-IDB_ADv1_COCO

git-large-r-coco-IDB_ADv1_COCO is a image-text-to-text model from ooliverz. 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

18% of all-time downloads

All-time downloads

Public

Parameters

394M

183 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors1.6 GB · 100%

At a glance

Task: Image-Text-to-Text
Library: transformers
License: mit
Model type: git
Access: Public
Created: Apr 8, 2025
Updated: Apr 8, 2025
SHA: 172abab9

Try a prompt

Base models

microsoft/git-large-r-coco

Datasets

imagefolder

Model card

From the Hugging Face model README

git-large-r-coco-IDB_ADv1_COCO

This model is a fine-tuned version of microsoft/git-large-r-coco on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.0718
Meteor Score: {'meteor': 0.666916447209016}

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: 3e-05
train_batch_size: 128
eval_batch_size: 128
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 1024
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
num_epochs: 280
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Meteor Score
0.1391	5.0	5	0.1066	{'meteor': 0.5464308622024807}
0.0701	10.0	10	0.0890	{'meteor': 0.5291953574710693}
0.0468	15.0	15	0.0789	{'meteor': 0.5496036528506629}
0.0325	20.0	20	0.0711	{'meteor': 0.5439550126260527}
0.0225	25.0	25	0.0672	{'meteor': 0.552830292712962}
0.0161	30.0	30	0.0679	{'meteor': 0.5428427974008356}
0.0123	35.0	35	0.0667	{'meteor': 0.5293045890136272}
0.0114	40.0	40	0.0650	{'meteor': 0.5470594759810721}
0.0084	45.0	45	0.0643	{'meteor': 0.5561520014815091}
0.0064	50.0	50	0.0651	{'meteor': 0.5610482415310057}
0.0051	55.0	55	0.0667	{'meteor': 0.5487886186676766}
0.0041	60.0	60	0.0685	{'meteor': 0.5616454798881054}
0.0037	65.0	65	0.0692	{'meteor': 0.5699718193151979}
0.0032	70.0	70	0.0685	{'meteor': 0.5580309821952526}
0.0029	75.0	75	0.0683	{'meteor': 0.5748316918656275}
0.0026	80.0	80	0.0684	{'meteor': 0.5895555620949836}
0.0024	85.0	85	0.0687	{'meteor': 0.5857286795606523}
0.0022	90.0	90	0.0695	{'meteor': 0.5846091691510659}
0.0021	95.0	95	0.0697	{'meteor': 0.594178001025322}
0.002	100.0	100	0.0694	{'meteor': 0.6081664555245014}
0.0019	105.0	105	0.0696	{'meteor': 0.6221380770749247}
0.0018	110.0	110	0.0697	{'meteor': 0.6033596220663302}
0.0017	115.0	115	0.0699	{'meteor': 0.5934573428451106}
0.0016	120.0	120	0.0697	{'meteor': 0.6100068434120042}
0.0016	125.0	125	0.0701	{'meteor': 0.6226574997552852}
0.0015	130.0	130	0.0704	{'meteor': 0.6266141282552855}
0.0015	135.0	135	0.0708	{'meteor': 0.6266624596822102}
0.0014	140.0	140	0.0713	{'meteor': 0.6253640811501537}
0.0014	145.0	145	0.0715	{'meteor': 0.6268835998646377}
0.0013	150.0	150	0.0716	{'meteor': 0.6391957882900023}
0.0013	155.0	155	0.0716	{'meteor': 0.6372778085384403}
0.0013	160.0	160	0.0714	{'meteor': 0.6420748904347257}
0.0012	165.0	165	0.0712	{'meteor': 0.6542694795600709}
0.0012	170.0	170	0.0711	{'meteor': 0.6640970636042774}
0.0012	175.0	175	0.0712	{'meteor': 0.6581755350563626}
0.0012	180.0	180	0.0715	{'meteor': 0.6563782816038787}
0.0012	185.0	185	0.0717	{'meteor': 0.6575711357356673}
0.0012	190.0	190	0.0719	{'meteor': 0.6615976015516674}
0.0011	195.0	195	0.0719	{'meteor': 0.6664084419111367}
0.0011	200.0	200	0.0719	{'meteor': 0.6714641579289897}
0.0011	205.0	205	0.0719	{'meteor': 0.6733748542723833}
0.0011	210.0	210	0.0718	{'meteor': 0.6716577609340998}
0.0011	215.0	215	0.0718	{'meteor': 0.6718891332503508}
0.0011	220.0	220	0.0718	{'meteor': 0.6705874088889952}
0.0011	225.0	225	0.0718	{'meteor': 0.6687440927433674}
0.0011	230.0	230	0.0718	{'meteor': 0.6683625041894395}
0.0011	235.0	235	0.0717	{'meteor': 0.667993183281943}
0.0011	240.0	240	0.0717	{'meteor': 0.6684321600001021}
0.0011	245.0	245	0.0718	{'meteor': 0.668594259646557}
0.0011	250.0	250	0.0718	{'meteor': 0.6675028779539088}
0.0011	255.0	255	0.0718	{'meteor': 0.6677654410135724}
0.0011	260.0	260	0.0718	{'meteor': 0.6664467133271365}
0.0011	265.0	265	0.0718	{'meteor': 0.6667295014161946}
0.0011	270.0	270	0.0718	{'meteor': 0.6671952775082015}
0.0011	275.0	275	0.0718	{'meteor': 0.6672252794210247}
0.0011	280.0	280	0.0718	{'meteor': 0.666916447209016}

Framework versions

Transformers 4.46.1
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.20.2

Task: Image-Text-to-Text
Library: transformers
Type: git
License: mit
Created: Apr 8, 2025
Updated: Apr 8, 2025