gpt2_m090_tiny-stories_1024

This model is a fine-tuned version of on the roneneldan/TinyStories dataset. It achieves the following results on the evaluation set:

• Loss: 1.1994
• Accuracy: 0.6811

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: 5e-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: 1.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.8751	0.0519	1000	2.4323	0.4504
1.9622	0.1037	2000	1.7853	0.5717
1.7125	0.1556	3000	1.6003	0.6046
1.5958	0.2074	4000	1.5009	0.6225
1.5199	0.2593	5000	1.4369	0.6347
1.4675	0.3112	6000	1.3928	0.6430
1.4297	0.3630	7000	1.3593	0.6495
1.3993	0.4149	8000	1.3303	0.6549
1.373	0.4668	9000	1.3077	0.6593
1.3537	0.5186	10000	1.2885	0.6631
1.3332	0.5705	11000	1.2709	0.6667
1.3207	0.6223	12000	1.2552	0.6697
1.3064	0.6742	13000	1.2452	0.6718
1.2972	0.7261	14000	1.2339	0.6740
1.2823	0.7779	15000	1.2240	0.6759
1.2703	0.8298	16000	1.2162	0.6775
1.2674	0.8817	17000	1.2090	0.6791
1.2591	0.9335	18000	1.2037	0.6802
1.2579	0.9854	19000	1.1997	0.6811

Framework versions

• Transformers 4.42.3
• Pytorch 2.2.2+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1