BEiT_beit-base-finetuned-ade-640-640_Clean-Set3-Grayscale_RGB

Hasano20/BEiT_beit-base-finetuned-ade-640-640_Clean-Set3-Grayscale_RGB

BEiT_beit-base-finetuned-ade-640-640_Clean-Set3-Grayscale_RGB is a machine learning model from Hasano20. Use it for the machine learning 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 apache-2.0.

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

Downloads · 30 days

17% of all-time downloads

All-time downloads

Public

Parameters

163M

653 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors653 MB · 100%

At a glance

Library: transformers
License: apache-2.0
Model type: beit
Access: Public
Created: Jul 10, 2024
Updated: Jul 10, 2024
SHA: c567d454

Base models

microsoft/beit-base-finetuned-ade-640-640

Model card

From the Hugging Face model README

BEiT_beit-base-finetuned-ade-640-640_Clean-Set3-Grayscale_RGB

This model is a fine-tuned version of microsoft/beit-base-finetuned-ade-640-640 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.0652
Mean Iou: 0.9446
Mean Accuracy: 0.9687
Overall Accuracy: 0.9855
Accuracy Background: 0.9893
Accuracy Melt: 0.9263
Accuracy Substrate: 0.9905
Iou Background: 0.9796
Iou Melt: 0.8757
Iou Substrate: 0.9785

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: 16
eval_batch_size: 16
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 200
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Melt	Accuracy Substrate	Iou Background	Iou Melt	Iou Substrate
0.1893	3.7037	100	0.1358	0.9180	0.9486	0.9815	0.9923	0.8651	0.9884	0.9821	0.7997	0.9723
0.0951	7.4074	200	0.0746	0.9483	0.9791	0.9876	0.9909	0.9575	0.9890	0.9860	0.8775	0.9814
0.0892	11.1111	300	0.0631	0.9489	0.9772	0.9866	0.9875	0.9536	0.9903	0.9817	0.8854	0.9796
0.0878	14.8148	400	0.0692	0.9375	0.9687	0.9829	0.9853	0.9332	0.9877	0.9742	0.8632	0.9750
0.07	18.5185	500	0.0631	0.9419	0.9700	0.9844	0.9865	0.9339	0.9897	0.9777	0.8714	0.9767
0.0507	22.2222	600	0.0646	0.9379	0.9659	0.9829	0.9879	0.9230	0.9869	0.9738	0.8653	0.9747
0.0523	25.9259	700	0.0569	0.9485	0.9724	0.9864	0.9903	0.9371	0.9898	0.9809	0.8855	0.9791
0.0402	29.6296	800	0.0622	0.9422	0.9686	0.9848	0.9881	0.9279	0.9897	0.9777	0.8711	0.9778
0.0406	33.3333	900	0.0631	0.9427	0.9683	0.9852	0.9894	0.9256	0.9899	0.9790	0.8707	0.9785
0.0389	37.0370	1000	0.0650	0.9436	0.9695	0.9852	0.9879	0.9303	0.9904	0.9784	0.8739	0.9784
0.0396	40.7407	1100	0.0634	0.9457	0.9700	0.9857	0.9900	0.9301	0.9899	0.9798	0.8787	0.9787
0.036	44.4444	1200	0.0653	0.9443	0.9684	0.9855	0.9894	0.9254	0.9906	0.9797	0.8748	0.9786
0.0224	48.1481	1300	0.0652	0.9446	0.9687	0.9855	0.9893	0.9263	0.9905	0.9796	0.8757	0.9785

Framework versions

Transformers 4.41.2
Pytorch 2.0.1+cu117
Datasets 2.19.2
Tokenizers 0.19.1

Library: transformers
Type: beit
License: apache-2.0
Created: Jul 10, 2024
Updated: Jul 10, 2024