blip-image-captioning-base-blip-dummy-temp-1

This model is a fine-tuned version of Salesforce/blip-image-captioning-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.7117
• Wer Score: 0.4618

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
7.1762	3.23	50	4.4355	0.8626
2.7292	6.45	100	1.8212	0.7393
1.6499	9.68	150	1.7030	0.5773
1.4923	12.9	200	1.6767	0.5157
1.4285	16.13	250	1.6746	0.4752
1.3994	19.35	300	1.6786	0.4720
1.3862	22.58	350	1.6862	0.4611
1.3801	25.81	400	1.6925	0.4572
1.3764	29.03	450	1.6972	0.4590
1.3742	32.26	500	1.7026	0.4583
1.3727	35.48	550	1.7066	0.4604
1.3718	38.71	600	1.7087	0.4600
1.3712	41.94	650	1.7102	0.4597
1.3708	45.16	700	1.7113	0.4614
1.3707	48.39	750	1.7117	0.4618

Framework versions

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