mega-base-wikitext-finetuned-KAGGLE

This model is a fine-tuned version of mnaylor/mega-base-wikitext on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.9914
• Accuracy: 0.8207

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-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	11	2.1216	0.4552
No log	2.0	22	1.9996	0.8207
No log	3.0	33	1.8810	0.8138
No log	4.0	44	1.7691	0.8138
No log	5.0	55	1.6643	0.8138
No log	6.0	66	1.5688	0.8138
No log	7.0	77	1.4820	0.8138
No log	8.0	88	1.4057	0.8138
No log	9.0	99	1.3398	0.8138
No log	10.0	110	1.2837	0.8138
No log	11.0	121	1.2361	0.8138
No log	12.0	132	1.1960	0.8138
No log	13.0	143	1.1628	0.8138
No log	14.0	154	1.1352	0.8138
No log	15.0	165	1.1129	0.8138
No log	16.0	176	1.0944	0.8138
No log	17.0	187	1.0790	0.8138
No log	18.0	198	1.0664	0.8138
No log	19.0	209	1.0561	0.8138
No log	20.0	220	1.0480	0.8138
No log	21.0	231	1.0416	0.8138
No log	22.0	242	1.0368	0.8138
No log	23.0	253	1.0334	0.8138
No log	24.0	264	1.0314	0.8138
No log	25.0	275	1.0307	0.8138

Framework versions

• Transformers 4.32.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3