glpn-nyu-finetuned-diode-230530-193901

This model is a fine-tuned version of vinvino02/glpn-nyu on the diode-subset dataset. It achieves the following results on the evaluation set:

• Loss: 1.5356
• Mae: 3.1497
• Rmse: 3.6237
• Abs Rel: 6.0096
• Log Mae: 0.6926
• Log Rmse: 0.8186
• Delta1: 0.3020
• Delta2: 0.3077
• Delta3: 0.3094

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: 1e-05
• train_batch_size: 24
• eval_batch_size: 48
• seed: 2022
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Mae	Rmse	Abs Rel	Log Mae	Log Rmse	Delta1	Delta2	Delta3
No log	1.0	1	1.5604	3.2768	3.8048	6.3111	0.7037	0.8347	0.2996	0.3073	0.3091
No log	2.0	2	1.5559	3.2536	3.7731	6.2584	0.7017	0.8319	0.2998	0.3073	0.3092
No log	3.0	3	1.5513	3.2298	3.7401	6.2034	0.6997	0.8290	0.3002	0.3074	0.3092
No log	4.0	4	1.5469	3.2076	3.7083	6.1506	0.6977	0.8262	0.3006	0.3075	0.3093
No log	5.0	5	1.5434	3.1894	3.6815	6.1060	0.6961	0.8238	0.3011	0.3075	0.3093
No log	6.0	6	1.5407	3.1757	3.6614	6.0725	0.6949	0.8220	0.3015	0.3076	0.3094
No log	7.0	7	1.5387	3.1652	3.6460	6.0468	0.6940	0.8207	0.3017	0.3076	0.3094
No log	8.0	8	1.5371	3.1574	3.6348	6.0281	0.6933	0.8196	0.3019	0.3077	0.3094
No log	9.0	9	1.5361	3.1523	3.6273	6.0157	0.6928	0.8190	0.3020	0.3077	0.3094
No log	10.0	10	1.5356	3.1497	3.6237	6.0096	0.6926	0.8186	0.3020	0.3077	0.3094

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.1+cu118
• Tokenizers 0.13.3