bit-50

This model is a fine-tuned version of google/bit-50 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 7122385408.0
• Accuracy: 0.085

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: 0.0002
• 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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
5528748032.0	1.0	50	7122386944.0	0.085
5155133849.6	2.0	100	7122386944.0	0.085
5068722995.2	3.0	150	7122386944.0	0.085
5613660569.6	4.0	200	7122385408.0	0.085
7499937382.4	5.0	250	7122385408.0	0.085
5806654259.2	6.0	300	7122385408.0	0.085
5483250483.2	7.0	350	7122385408.0	0.085
6852667392.0	8.0	400	7122385408.0	0.085

Framework versions

• Transformers 4.40.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1