mit-b0-finetuned-sidewalk-semantic

raks87/mit-b0-finetuned-sidewalk-semantic

mit-b0-finetuned-sidewalk-semantic is a machine learning model from raks87. 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 other.

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

Downloads · 30 days

6% of all-time downloads

All-time downloads

158

Public

Parameters

3.7M

91.7 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors14.9 MB · 87%

At a glance

Library: transformers
License: other
Model type: segformer
Access: Public
Created: Apr 12, 2024
Updated: Apr 13, 2024
SHA: 0fbccfaf

Base models

nvidia/mit-b0

Model card

From the Hugging Face model README

mit-b0-finetuned-sidewalk-semantic

This model is a fine-tuned version of nvidia/mit-b0 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.3201
Mean Iou: 0.3806
Mean Accuracy: 0.4846
Overall Accuracy: 0.6943
Accuracy Background: nan
Accuracy Hat: 0.0
Accuracy Hair: 0.8309
Accuracy Sunglasses: 0.0
Accuracy Upper-clothes: 0.8803
Accuracy Skirt: 0.5781
Accuracy Pants: 0.8338
Accuracy Dress: 0.4711
Accuracy Belt: 0.0
Accuracy Left-shoe: 0.1599
Accuracy Right-shoe: 0.3381
Accuracy Face: 0.8563
Accuracy Left-leg: 0.7194
Accuracy Right-leg: 0.7205
Accuracy Left-arm: 0.6508
Accuracy Right-arm: 0.6578
Accuracy Bag: 0.5406
Accuracy Scarf: 0.0
Iou Background: 0.0
Iou Hat: 0.0
Iou Hair: 0.7122
Iou Sunglasses: 0.0
Iou Upper-clothes: 0.6504
Iou Skirt: 0.4790
Iou Pants: 0.6587
Iou Dress: 0.3859
Iou Belt: 0.0
Iou Left-shoe: 0.1507
Iou Right-shoe: 0.2691
Iou Face: 0.7173
Iou Left-leg: 0.5748
Iou Right-leg: 0.5947
Iou Left-arm: 0.5816
Iou Right-arm: 0.5871
Iou Bag: 0.4893
Iou Scarf: 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: 2
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 8
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Hair	Accuracy Upper-clothes	Accuracy Skirt	Accuracy Pants	Accuracy Dress	Accuracy Left-shoe	Accuracy Right-shoe	Accuracy Face	Accuracy Left-leg	Accuracy Right-leg	Accuracy Left-arm	Accuracy Right-arm	Accuracy Bag	Iou Hair	Iou Upper-clothes	Iou Skirt	Iou Pants	Iou Dress	Iou Left-shoe	Iou Right-shoe	Iou Face	Iou Left-leg	Iou Right-leg	Iou Left-arm	Iou Right-arm	Iou Bag
1.5584	1.0	100	1.4751	0.1357	0.2382	0.4526	nan	0.8771	0.8883	0.0443	0.7221	0.0035	0.0187	0.0055	0.2572	0.5884	0.5612	0.0822	0.0013	0.0	0.5636	0.3813	0.0433	0.3814	0.0035	0.0182	0.0055	0.2523	0.3602	0.3582	0.0746	0.0013	0.0
1.1073	2.0	200	1.0997	0.2194	0.3308	0.5583	nan	0.9122	0.8933	0.5007	0.6982	0.1416	0.0076	0.0436	0.7573	0.6194	0.7115	0.2770	0.0608	0.0012	0.6610	0.4693	0.3429	0.5229	0.1246	0.0076	0.0416	0.6491	0.4038	0.4308	0.2338	0.0605	0.0012
0.805	3.0	300	0.7604	0.2466	0.3515	0.5861	nan	0.8500	0.8839	0.4934	0.8517	0.2381	0.0038	0.0406	0.8209	0.5776	0.7025	0.2485	0.2341	0.0298	0.6900	0.5378	0.3542	0.5424	0.2035	0.0038	0.0391	0.6827	0.4027	0.4848	0.2384	0.2289	0.0296
0.604	4.0	400	0.5498	0.2906	0.3944	0.6189	nan	0.8108	0.8788	0.6810	0.7835	0.2571	0.0016	0.1009	0.8612	0.6496	0.6929	0.4317	0.4043	0.1522	0.6910	0.5894	0.4338	0.6222	0.2234	0.0016	0.0918	0.6875	0.4402	0.5096	0.4032	0.3877	0.1492
0.4334	5.0	500	0.4440	0.3219	0.4196	0.6428	nan	0.8265	0.8612	0.4725	0.8254	0.4861	0.0033	0.1673	0.8410	0.6689	0.6548	0.5207	0.5088	0.2962	0.6959	0.6233	0.3809	0.6130	0.3510	0.0033	0.1437	0.7028	0.4987	0.5323	0.4820	0.4809	0.2858
0.4213	6.0	600	0.3817	0.3491	0.4549	0.6658	nan	0.8247	0.8762	0.7055	0.7855	0.3145	0.0273	0.2536	0.8611	0.6931	0.7257	0.6254	0.6281	0.4132	0.7044	0.6379	0.4727	0.6504	0.2752	0.0272	0.2056	0.7066	0.5298	0.5651	0.5557	0.5634	0.3902
0.3325	7.0	700	0.3484	0.3690	0.4758	0.6840	nan	0.8352	0.8333	0.6651	0.8321	0.4643	0.0780	0.3248	0.8554	0.6926	0.7224	0.6461	0.6486	0.4906	0.7079	0.6573	0.4848	0.6432	0.3743	0.0765	0.2516	0.7128	0.5528	0.5816	0.5693	0.5773	0.4521
0.2556	8.0	800	0.3384	0.3795	0.4845	0.6971	nan	0.8404	0.8723	0.6558	0.8311	0.4614	0.1270	0.3250	0.8533	0.6978	0.7209	0.6525	0.6619	0.5364	0.7130	0.6572	0.5012	0.6634	0.3790	0.1220	0.2599	0.7153	0.5627	0.5908	0.5849	0.5933	0.4873
0.3337	9.0	900	0.3201	0.3806	0.4846	0.6943	nan	0.8309	0.8803	0.5781	0.8338	0.4711	0.1599	0.3381	0.8563	0.7194	0.7205	0.6508	0.6578	0.5406	0.7122	0.6504	0.4790	0.6587	0.3859	0.1507	0.2691	0.7173	0.5748	0.5947	0.5816	0.5871	0.4893
0.2843	10.0	1000	0.3204	0.3879	0.4943	0.7036	nan	0.8304	0.8535	0.6956	0.8303	0.4990	0.1708	0.3445	0.8594	0.7149	0.7322	0.6598	0.6786	0.5344	0.7126	0.6681	0.5240	0.6700	0.4029	0.1600	0.2739	0.7169	0.5757	0.6008	0.5868	0.6012	0.4902

Framework versions

Transformers 4.38.2
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.15.2

Library: transformers
Type: segformer
License: other
Created: Apr 12, 2024
Updated: Apr 13, 2024