git-base-coco-dummy-temp12

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

• Loss: 0.3199
• Wer Score: 2.1457

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Score
7.2365	3.23	50	4.6009	2.7286
2.7457	6.45	100	0.9830	2.0444
0.5898	9.68	150	0.4162	1.9788
0.2845	12.9	200	0.3463	2.3103
0.1764	16.13	250	0.3227	2.3805
0.1162	19.35	300	0.3167	2.3759
0.0856	22.58	350	0.3144	2.2456
0.065	25.81	400	0.3138	2.3201
0.0513	29.03	450	0.3149	2.1917
0.0435	32.26	500	0.3163	2.1561
0.0363	35.48	550	0.3162	2.1466
0.0312	38.71	600	0.3179	2.1793
0.0286	41.94	650	0.3200	2.2413
0.0267	45.16	700	0.3204	2.1061
0.0255	48.39	750	0.3199	2.1457

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.3
• Tokenizers 0.13.3