glpn-nyu-finetuned

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: 1.5286
• Mae: 3.1196
• Rmse: 3.5796
• Abs Rel: 5.9353
• Log Mae: 0.6899
• Log Rmse: 0.8145
• Delta1: 0.3012
• Delta2: 0.3076
• Delta3: 0.3093

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	1.5476	3.2112	3.7133	6.1586	0.6980	0.8267	0.2998	0.3073	0.3091
No log	2.0	2	1.5441	3.1939	3.6889	6.1181	0.6965	0.8245	0.3001	0.3073	0.3091
No log	3.0	3	1.5410	3.1783	3.6668	6.0811	0.6951	0.8225	0.3003	0.3074	0.3092
No log	4.0	4	1.5381	3.1643	3.6465	6.0474	0.6939	0.8207	0.3005	0.3074	0.3092
No log	5.0	5	1.5355	3.1520	3.6285	6.0172	0.6928	0.8190	0.3007	0.3075	0.3092
No log	6.0	6	1.5333	3.1415	3.6128	5.9909	0.6918	0.8176	0.3009	0.3075	0.3092
No log	7.0	7	1.5315	3.1329	3.5999	5.9693	0.6911	0.8164	0.3010	0.3075	0.3093
No log	8.0	8	1.5301	3.1264	3.5901	5.9529	0.6905	0.8155	0.3011	0.3075	0.3093
No log	9.0	9	1.5291	3.1219	3.5832	5.9413	0.6901	0.8149	0.3012	0.3076	0.3093
No log	10.0	10	1.5286	3.1196	3.5796	5.9353	0.6899	0.8145	0.3012	0.3076	0.3093

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.1+cu118
• Tokenizers 0.13.3