mobilebert_1000exs_20timesteps_run

This model is a fine-tuned version of google/mobilebert-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8330
• Accuracy: 0.51

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: 2e-06
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 16

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	16	3052559.25	0.51
No log	2.0	32	2330149.0	0.51
No log	3.0	48	1716418.75	0.51
No log	4.0	64	1113477.0	0.51
No log	5.0	80	524712.25	0.51
No log	6.0	96	36429.3633	0.51
No log	7.0	112	5025.1558	0.51
No log	8.0	128	2793.7087	0.51
No log	9.0	144	173.9886	0.51
No log	10.0	160	1.6353	0.49
No log	11.0	176	1.2341	0.49
No log	12.0	192	0.9290	0.49
No log	13.0	208	0.7376	0.49
No log	14.0	224	0.7311	0.51
No log	15.0	240	0.8214	0.51
No log	16.0	256	0.8330	0.51

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.0+cu118
• Datasets 2.17.1
• Tokenizers 0.15.2