grounding-dino-tiny-finetuned-cppe-5-10k-steps

danelcsb/grounding-dino-tiny-finetuned-cppe-5-10k-steps

grounding-dino-tiny-finetuned-cppe-5-10k-steps is a object detection model from danelcsb. Use it when you need objects located in an image. The card lists the license as apache-2.0.

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

Downloads · 30 days

3% of all-time downloads

All-time downloads

459

Public

Parameters

172M

689 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors689 MB · 100%

Parameter types

How the weights are stored.

F32172M · 100%

Base models

IDEA-Research/grounding-dino-tiny

Model card

From the Hugging Face model README

grounding-dino-tiny-finetuned-cppe-5-10k-steps

This model is a fine-tuned version of IDEA-Research/grounding-dino-tiny on the cppe-5 dataset. It achieves the following results on the evaluation set:

Loss: 5.5317
Map: 0.0151
Map 50: 0.0275
Map 75: 0.0157
Map Small: 0.0125
Map Medium: 0.0149
Map Large: 0.0236
Mar 1: 0.0202
Mar 10: 0.0902
Mar 100: 0.1127
Mar Small: 0.0815
Mar Medium: 0.0975
Mar Large: 0.1461
Map Coverall: 0.0755
Mar 100 Coverall: 0.5636
Map Face Shield: 0.0
Mar 100 Face Shield: 0.0
Map Gloves: 0.0
Mar 100 Gloves: 0.0
Map Goggles: 0.0
Mar 100 Goggles: 0.0
Map Mask: 0.0
Mar 100 Mask: 0.0

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: 1
eval_batch_size: 1
seed: 1337
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Small	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Small	Mar Medium	Mar Large	Map Coverall	Mar 100 Coverall
8355.9482	1.0	850	6.6137	0.014	0.0272	0.0134	0.0048	0.0111	0.0243	0.0149	0.0893	0.1073	0.0523	0.0889	0.1328	0.0702	0.5366
6.7523	2.0	1700	6.2357	0.0162	0.0302	0.0148	0.0106	0.0189	0.0192	0.0247	0.0894	0.107	0.0643	0.0968	0.1258	0.0809	0.535
6.5566	3.0	2550	6.0890	0.0158	0.0294	0.0134	0.01	0.0199	0.0215	0.0222	0.0876	0.1065	0.0671	0.0846	0.1324	0.0791	0.5323
6.2217	4.0	3400	5.9028	0.0144	0.0271	0.0134	0.0066	0.0096	0.0225	0.0232	0.0857	0.107	0.06	0.0823	0.1397	0.0721	0.5348
6.0963	5.0	4250	5.8411	0.0126	0.0215	0.014	0.0055	0.0138	0.0178	0.0201	0.0811	0.1052	0.044	0.0942	0.1377	0.0631	0.5258
5.996	6.0	5100	5.7244	0.0162	0.0311	0.0166	0.0059	0.0145	0.0221	0.0223	0.0869	0.1088	0.0667	0.0919	0.1328	0.0812	0.5437
5.8971	7.0	5950	5.5473	0.0154	0.027	0.016	0.01	0.014	0.0208	0.0244	0.0946	0.1154	0.084	0.106	0.1311	0.0769	0.5772
5.7451	8.0	6800	5.5231	0.0146	0.0267	0.0148	0.0021	0.0161	0.0183	0.0256	0.0905	0.1125	0.0325	0.1062	0.128	0.0731	0.5624
5.7931	9.0	7650	5.5190	0.0182	0.032	0.0195	0.0175	0.0147	0.0249	0.0299	0.1048	0.1138	0.08	0.0945	0.1309	0.091	0.5688
5.7435	10.0	8500	5.5317	0.0151	0.0275	0.0157	0.0125	0.0149	0.0236	0.0202	0.0902	0.1127	0.0815	0.0975	0.1461	0.0755	0.5636

Framework versions

Transformers 4.45.0.dev0
Pytorch 2.2.2
Datasets 2.20.0
Tokenizers 0.19.1

Task: Object Detection
Type: grounding-dino
License: apache-2.0
Created: Aug 8, 2024
Updated: Aug 9, 2024