beit-finetuned-pokemon

This model is a fine-tuned version of ydmeira/beit-finetuned-pokemon on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0219
• Mean Iou: 0.4955
• Mean Accuracy: 0.9910
• Overall Accuracy: 0.9910
• Per Category Iou: [0.0, 0.9909617791470107]
• Per Category Accuracy: [nan, 0.9909617791470107]

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: 6e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Per Category Iou	Per Category Accuracy
0.0354	0.21	1000	0.0347	0.4978	0.9955	0.9955	[0.0, 0.9955007125868244]	[nan, 0.9955007125868244]
0.0273	0.43	2000	0.0277	0.4951	0.9903	0.9903	[0.0, 0.9902709092544748]	[nan, 0.9902709092544748]
0.0307	0.64	3000	0.0788	0.4875	0.9751	0.9751	[0.0, 0.9750850921785902]	[nan, 0.9750850921785902]
0.0295	0.85	4000	0.0412	0.4939	0.9877	0.9877	[0.0, 0.9877162657609527]	[nan, 0.9877162657609527]
0.0255	1.07	5000	0.0842	0.4862	0.9723	0.9723	[0.0, 0.972304346385062]	[nan, 0.972304346385062]
0.0253	1.28	6000	0.0325	0.4950	0.9901	0.9901	[0.0, 0.9900621363084688]	[nan, 0.9900621363084688]
0.0239	1.49	7000	0.0440	0.4917	0.9835	0.9835	[0.0, 0.9834701005512881]	[nan, 0.9834701005512881]
0.0238	1.71	8000	0.0338	0.4950	0.9900	0.9900	[0.0, 0.9899977115151821]	[nan, 0.9899977115151821]
0.0223	1.92	9000	0.0319	0.4950	0.9900	0.9900	[0.0, 0.989994712810938]	[nan, 0.989994712810938]
0.0231	2.13	10000	0.0382	0.4921	0.9841	0.9841	[0.0, 0.984106425591889]	[nan, 0.984106425591889]
0.0205	2.35	11000	0.0450	0.4926	0.9851	0.9851	[0.0, 0.9851146530893756]	[nan, 0.9851146530893756]
0.0201	2.56	12000	0.0265	0.4954	0.9908	0.9908	[0.0, 0.9908277212846449]	[nan, 0.9908277212846449]
0.0188	2.77	13000	0.0377	0.4933	0.9866	0.9866	[0.0, 0.9865726862234793]	[nan, 0.9865726862234793]
0.0181	2.99	14000	0.0219	0.4955	0.9910	0.9910	[0.0, 0.9909617791470107]	[nan, 0.9909617791470107]

Framework versions

• Transformers 4.22.1
• Pytorch 1.12.1+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1