aoi_clip_high_resolution_crossAttenttionFusion_fusin_new_sampler

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: 4.5135
• Accuracy: 0.0583

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: 100.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.156	9.9880	3110	2.9957	0.0492
2.1411	19.9759	6220	3.0536	0.0537
1.9888	29.9639	9330	3.3324	0.0567
1.8759	39.9518	12440	3.6092	0.0571
1.8129	49.9398	15550	3.8091	0.0575
1.7708	59.9277	18660	3.9898	0.0578
1.7413	69.9157	21770	4.2735	0.0579
1.7172	79.9037	24880	4.3434	0.0580
1.7056	89.8916	27990	4.5120	0.0581
1.7018	99.8796	31100	4.5135	0.0582

Framework versions

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