sentance_split_by_aoi_gpt_crossAttention

This model is a fine-tuned version of OFA-Sys/chinese-clip-vit-base-patch16 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 4.1585
• Accuracy: 0.0675

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: 25
• eval_batch_size: 20
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 200
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 60.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.2577	5.9676	276	2.9735	0.0719
1.194	11.9351	552	2.9341	0.0719
1.1008	17.9027	828	3.0206	0.0690
1.0173	23.8703	1104	3.2514	0.0667
0.9404	29.8378	1380	3.4461	0.0679
0.8841	35.8054	1656	3.6906	0.0698
0.8364	41.7730	1932	3.8565	0.0702
0.8136	47.7405	2208	4.0121	0.0697
0.7757	53.7081	2484	4.0667	0.0686
0.766	59.6757	2760	4.1585	0.0680

Framework versions

• Transformers 4.42.3
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1